METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 708,  1998 
FINNISH FOREST RESEARCH  INSTITUTE, RESEARCH  PAPERS 708, 1998 
Demand for  Finnish exports of 
forest  products: 
Econometric  analyses  using  time series  data 
Riitta Hänninen  
HELSINGIN  TUTKIMUSKESKUS  -  HELSINKI  RESEARCH  CENTRE  

METSÄNTUTKIMUSLAITOKSEN  TIEDONANTOJA 708,1998  
FINNISH  FOREST  RESEARCH  INSTITUTE,  RESEARCH  PAPERS, 708, 1998 
DEMAND  FOR  FINNISH  EXPORTS  OF  FOREST  PRODUCTS:  
Econometric  analyses  using  time series  data 
Riitta Hänninen 
Doctoral dissertation 
To be presented,  by  the permission  of the Faculty  of  Agriculture  and Forestry  of the 
University  of Helsinki,  for  public  examination in Auditorium XII of  the main building  of  the 
University,  Unioninkatu 34,  on  November 13, 1998 at 12 o'clock  noon. 
HELSINGIN TUTKIMUSKESKUS -  HELSINKI  RESEARCH CENTRE 
Hänninen,  Riitta. 1998. Demand for Finnish  exports  of forest products:  
Econometric analyses  using  time series  data. 
Metsäntutkimuslaitoksen tiedonantoja  708.  Finnish Forest  Research Institute,  Research 
papers 708. ISBN 951-40 -1650 -5,  ISSN 0358 -  4283. 60 p.  + 5  original  papers. 
Publisher: Finnish Forest Research Institute 
Approved:  Matti Kärkkäinen,  Research Director,  25.9.1998 
Author's address:  
Riitta Hänninen, Finnish  Forest Research  Institute,  Helsinki Research  Center,  
Unioninkatu 40 A, 00170 Helsinki,  Finland, 
tel:  +358 9  857057 46,  fax:  +358 9  85705717 
e-mail: riitta.hanninen@metla.fi 
Doctoral dissertation at  the Department  of Forest Economics,  University  of  Helsinki 
Supervised  by: Professor Jari Kuuluvainen,  
Department  of  Forest  Economics,  
University  of Helsinki 
Pre-examiners: Ph. D. Jussi Uusi  vuori 
Finnish Forest Research Institute 
Helsinki 
and 
Dr. Jouko Vilmunen 
Research  Department,  
Bank  of  Finland 
Helsinki 
Opponent: Professor  Joseph  Buongiorno  
Forest  Ecology  and Management,  
University  of  Wisconsin 
Madison, USA 
ABSTRACT  
The study  analyses  Finland's and its  main competitor  countries' export  price  formation in 
British (UK)  and German (BDR)  forest  product  markets using  time series  data. The results 
of  the study  are  useful in modeling  forest  product  trade flows. They  also  yield information 
for policy  makers  and industrial  agents,  for  example,  in the adjustment  to  the third  stage of  
European  Economic and Monetary  Union (EMU). Johansen's cointegration  method, thusfar  
infreaqvently  applied  to the Finnish forest sector,  is used in the estimation. 
Export  price  formation is  examined in three ways.  First,  price  substitution between 
Finland's,  and  its competitor's  sawnwood is  analysed  in the UK  market. Second,  the law of  
one price (LOP)  related to  arbitrage  is  tested for  prices  of sawnwood  and newsprint  
originating  from Finland,  and its  competitor  countries. Finally,  the effects of  exchange  rate  
changes  on Finnish sawnwood, newsprint  and chemical pulp  export  prices  are  analysed  in 
both destination countries. 
The estimation results  indicated imperfect  competition  in the export markets.  The  
price  elasticities of substitution between Finnish,  Swedish and Canadian sawnwoods were  
relatively  low and  the LOP  was  rejected  in the UK.  The LOP  was  also  rejected  between the 
main newsprint supplying  countries in German and British markets except between Sweden  
and Canada in the UK. The resulting  exchange  rate  pass-through  (PT) elasticities of Finnish 
export prices  indicate the importance  of exchange  rates  on exports  of forest products.  
Higher  PT for sawnwood than  for paper suggests  that Finnish sawnwood exporters have 
made use  of  depreciations  and devaluations of  the Finnish  markka  (FIM) to improve  their  
price competitiveness  and to maintain and increase market share. In paper markets,  Finnish 
exporters  have pursued  a midway  pricing  strategy  aimed at both  maintaining  market shares 
and profitability.  
Key  words: 
Forest  products,  sawnwood,  newsprint,  export  price  formation,  imperfect  competition,  
elasticity  of substitution,  law  of one price, exchange  rate  pass-through,  Johansen 
cointegration  method,  United Kingdom,  Germany  

Dedicated to my son Tapio  

Preface 
My research interest  on price  determination and exchange  rate  effects  on Finnish  exports of 
forest  products  is  related to  Finland's participation  in the integration  of Europe.  The 
realization of  European  Monetary  Union  (EMU)  on 1.1.1999 is  the next  step  in this  process  
and it will intensely  affect the  market environment for Finnish forest industry.  Although  it is 
difficult to  asses  the future by  analysing  historical data,  the  results  of  the present  study  give  
insights  into  operating  within the impending  EMU. The present  study  is  part  of  the project  
"Short-term Forecasting  System  for the Finnish Forest Sector" (MESU) at the Finnish 
Forest  Research  Institute  (Metla)  between 1994 and 1998. The research  has  been mainly  
financed by  Metla,  but  research  grants from Metsämiesten Säätiö are  also gratefully 
acknowledged.  
My special  gratitude  belongs  to  Jari Kuuluvainen,  who  has  supervised  my work  from  the 
beginning  with encouragement and  guidance.  Moreover,  as  my  previous  project  leader and 
collaborator,  Jari has  been an  essential  support  throughout  all  of  my  scientific  work.  I  also 
want  to thank Heikki  Juslin for his encouragement of  my  studies and Mikko  Tervo for his 
comments  on  this  summarizing  report.  My  project  leader,  Lauri Hetemäki,  has  been  
important  throughout  my  research and  his  constructive  criticism has  greatly  increased my  
skills  as  a  researcher.  I  am also  grateful  to  Anne Toppinen  for  her  friendship  and  her  
participation  in this project,  as  well  as  her  collaboration on  other projects.  Part  of this  
dissertation is  a  result  from joint  research  with Anne  as  well  as  Pertti  Ruuska.  Antti Ripatti  
helped  me  with methodological  questions  and  the comments  of both Ville Ovaskainen  and 
Ashley  Selby  were  also  helpful  in organizing  this  summarizing  report.  I  also  want  to  thank, 
warmly,  my  pre-examiners,  Jussi  Uusivuori  and Jouko Vilmunen,  for  their valuable 
comments and suggestions  to  improve  the summary report.  Additionally,  Robert Craig  has 
been helpful  in improving my English  in the final stages of  the summary report.  
I have been fortunate to share  a  researcher's  life  with my beloved husband,  Harri,  to whom 
I am profoundly  grateful  for his support  and understanding.  
Kerava,  October 9, 1998 Riitta Hänninen 

LIST OF  SEPARATE STUDIES 10 
1. INTRODUCTION 11 
1.1. Background  and  justification 11 
1.2 Aim and outline 13 
2. EXPORTERS' MARKET SHARES  AND UNIT PRICES  IN THE UK 
AND GERMANY 15 
3. THEORETICAL FRAMEWORK AND EARLIER LITERATURE 20 
3.1 Trade models and market  competition 20 
3.2. Modeling  export markets  using  Armington  approach 22 
3.3 Testing  the  law of one price  in forest products  markets 25 
3.4 Exchange  rate  pass-through  on  forest  products  export  prices 28 
4. DATA 32  
5. COINTEGRATION METHOD 35 
51. Johansen's multivariate cointegration  method 35 
52. Testable hypotheses 39  
6. RESULTS 43 
6.1  Price  elasticities  of substitution  between Finland and  the other  supplier  
countries in the demand of the United Kingdom  sawnwood imports 43 
6.2  The law of  one  price in United Kingdom  soft  sawnwood  imports  
-  a  cointegration  approach 44 
6.3  Testing  arbitrage  in  newsprint  imports  to  United Kingdom  and  Germany 45 
6.4  Exchange  rate  changes  and the Finnish  sawnwood demand and  price  
in the UK market 46 
6.5 Long-run  price  effects  of  exchange  rate  changes  on Finnish pulp  and paper 
exports 47 
7. CONCLUSIONS 48 
REFERENCES 51 
10 
LIST  OF SEPARATE  STUDIES  
The  dissertation includes the following  separate studies: 
Hänninen,  R. 1994. Price  elasticities of  substitution between  Finland and the other supplier  
countries in the demand of  the United Kingdom  sawnwood imports.  In Proceedings  of the 
Biennial meeting  of the Scandinavian  Forest  Economics,  Helles,  F.,  and  Linddal,  M. (eds.).  
Scandinavian Forest Economics 35:204-217. 
Hänninen,  R.,  Toppinen,  A.,  and  Ruuska,  P.  1997. Testing  arbitrage  in newsprint imports to 
United Kingdom  and Germany.  Canadian Journal of Forest  Research  27:1946-1952. 
Hänninen, R. H. 1998  a. The law  of one price  in United Kingdom  soft  sawnwood imports -  a 
cointegration  approach.  Forest  Science 44(1):  17-23 
Hänninen, R. H.  1998b. Exchange  rate  changes  and the Finnish sawnwood demand and 
price  in the UK  market.  Silva  Fennica  32(l):61-73.  
Hänninen, R. and Toppinen,  A. 1998. Long-run  price  effects of  exchange  rate  changes  in 
Finnish pulp  and paper exports.  Forthcoming  in Applied  Economics. 
The first  four articles are  reprinted  with permission. 
11 
1. INTRODUCTION  
1.1. Background  and justification  
Substantial changes  in the export market environment produce  challenges  for  Finnish forest  
industry.  Competition  in world markets  and Finland's main  market area,  Europe,  is  likely  to  
increase in the future due to  new producers  and low production  costs in,  e.g., Southeast 
Asia (Hetemäki*  1997). The realization of the third phase  of European  Economic and 
Monetary  Union (EMU)  on 1.1.1999 removes  the exchange  rate  buffer for Finnish forest 
industry  trade with the other  EMU countries. Adjustment  difficulties to  the new market 
environment may arise,  if exchange  rate  has  an important effect on Finnish exports.  A 
common currency  will facilitate price  comparisons  in the EMU area,  which  may increase 
competition  between producers  in different countries. This study  produces  information 
about price formation and competition  in  export  markets,  which  is  useful for  policy  makers 
and for industry in adjusting  to future market changes.  For those modeling  Finnish  forest 
industry  exports,  the information is  also useful in choosing  a  suitable modeling  strategy.  
The study focuses  on  the two  most  important export  countries for Finland in the European  
Union, the United Kingdom  (UK) and Germany  (BDR).  They  accounted for  32  percent of 
sawnwood exports,  42 percent  of newsprint  exports  and 60 percent  of chemical pulp  
exports  from Finland in 1996 (Statistical  Yearbook...  1997). Finnish sawnwood and 
newsprint  industries are clearly export  industries;  the greatest  part  of  their production  is  
exported.  In the Finnish chemical pulp  industry,  the export  share of  total production  is, 
however, only  about 25 percent.  Finland's main competitors in paper and pulp  markets are 
Sweden,  Canada and the USA, and in sawnwood markets Sweden,  Canada and Russia. 
Sawnwood,  newsprint  and chemical pulp  were  chosen for  this  analysis,  because  they  form 
relatively  homogenous  product  groups, a  prerequisite  for  an analysis  based on  unit prices.  
In  the present  study  of  price  formation three different aspects  are examined. First,  the study  
examines price  substitution. Second,  it tests the existence of price  differences (the  law of 
one  price,  LOP)  between  Finland's and its competitor's  products.  Third, the magnitude  of 
exchange  rate  changes  on Finnish  export demand and prices  expressed  in terms of foreign  
12 
currency (pass-through  effect)  is  examined. The three topics  have implications  as  to the  
degree of  competition  in the market. High  substitution elasticities,  the existence  of LOP  
and a low pass-through  (PT) effect would imply perfectly  competitive  markets,  whereas,  
relatively low  substitution elasticities, the rejection  of the LOP and a high PT would 
indicate imperfect  competition.  
Previous studies on export  price  formation in the whole of Finnish industry suggest  
imperfect  competition at the aggregate level (Honkatukia  1994). For paper industry  
exports,  the results  gave clear support to a  judgement  of  imperfect competition,  but  for  
wood  products  the evidence was less  clear-cut (Sukselainen  1986, Honkatukia  1994 and 
1995).  Because  forest industry  produces  many different products  and product  groups, it is  
natural that each of these may show varying  degrees  of competition  in export markets. 
Thus, it is difficult to  obtain a uniform view over  the degree of competition  in their 
international trade. Traded products  may be relatively  homogenous,  suggesting  more 
perfect  than imperfect  competition.  However,  even  in this  case  market imperfections  remain 
possible.  
Studies modeling  the trade in different forest  products  have used different assumptions  
concerning  competition. In some studies the starting  point  has  been perfect  competition,  
while the others  assume  imperfect  competition.  The  assumptions  are  not, however, usually  
explicitly  tested. In particular,  the differences between  Finland's and its  competitor's  prices  
have  not  been statistically  tested in earlier works,  although  there do exist  a  few studies on 
the effect of exchange  rate  on Finnish forest  product  exports  (e.g.  Uusivuori and 
Buongiorno  1990,  Vesala 1992,  Laaksonen 1998).  In the estimation of exchange  rate  
effects,  the present  study  uses  Johansen's method (1988),  which takes account  of  time 
series  properties  and cointegration  of the data. This  has  been done only  in the most  recent  
studies of  the Finnish  forest sector  (e.g., Ripatti  1990,  Toppinen  and Kuuluvainen 1997, 
Laaksonen 1998, and Toppinen 1998). Using new estimation methods,  additional 
information is generated on  export  price  formation, thus  increasing  the understanding  of  the 
functioning  of  forest products  export  markets. 
13 
1.2 Aim and outline  
The study examines forest products  price  formation in Finnish export markets  by  applying  
partial  equilibrium  models and  the econometric method. The analysis  includes five  separate 
studies  concerning  soft  sawnwood, newsprint  and chemical  pulp  exports  to  Finland's most 
important  export countries, the United  Kingdom  and  Germany.  The analysis  can  be divided 
into three specific  sub-topics:  
1. Price  substitution in the UK  sawnwood imports  from Finland,  Sweden,  Canada and 
Russia  (Hänninen  1994). 
2. The existence of long-term  price  differences in UK sawnwood imports  and in British and 
German  newsprint  imports  from major  suppliers  (Hänninen  1998 a,  Hänninen et ai. 
1997). 
3. Exchange  rate  effects  on  Finnish  sawnwood export  demand and price  in  the UK  market 
and  on  Finnish  newsprint and  chemical pulp  prices  in the UK  and  Germany  (Hänninen  
1998b, Hänninen and Toppinen  1998). 
The  study of  the first topic  aimed to  estimate own-price  and  substitution elasticities between 
the four supplier  countries' sawnwoods  in the UK  market.  For this  purpose, an econometric 
model for UK  sawnwood demand from the supplier  countries generalizing the original  
export demand model of Armington (1969)  was  specified  and estimated using annual data 
from the period  1961-1990. The studies of  the second topic examined  differences between 
Finland's,  and its  competitor's  export prices  for  sawnwood and newsprint  in the UK  and 
Germany.  The  analyses  were done by  testing  the  law of one price  (LOP) in these markets 
using  quarterly data from the period  1978-1992 for sawnwood and 1980-1994 for 
newsprint.  The studies  of  the third topic  aimed to  examine the effects  of  exchange  rate  
changes  on Finnish sawnwood, newsprint  and chemical  pulp  export  prices.  For  newsprint  
and  pulp  exchange  rate  PT was  estimated from  a  markup  price  equation  (e.g.,  Hung  et  al. 
1993) and for sawnwood,  the model was  enlarged  to include an export demand equation  
assuming  constant  elasticity  of substitution between supplier  countries.  The estimations 
were  made using  quarterly  data from the period  1978-1994 for sawnwood and 1980-1994 
for newsprint.  
14 
The outline of  the present  summary paper is  as  follows. Chapter  2 presents  the development  
of Finland's and its competitor's  market shares and the prices of  their exports  in the UK and 
Germany.  Chapter  3  reviews  the related literature and presents  the theoretical framework of 
the studies. Data  is  summarized in Chapter  4 and Johansen's cointegration  method and the 
testable hypothesis  of the study  are presented  in Chapter  5. Short summaries of each study  
are  presented  in Chapter  6  followed by  conclusions with suggestions  for  further research in 
Chapter  7. 
15 
2.  EXPORTERS'  MARKET  SHARES  AND UNIT  PRICES  IN  THE UK  
AND GERMANY  
The main exporters  to the UK sawnwood market,  Finland,  Sweden,  Canada and Russia,  
account  for  about 80 percent  of  total UK  sawnwood imports.  The UK  is highly  dependent  
on imports,  even though  it doubled domestic sawnwood production  between 1980 and 
1994. The percentage  ratio share of  domestic production  to imports  has  increased during  
the past  14 years  from 18 percent  in 1980 to  25  percent  in 1994  (FAO  1991 and 1994). 
In  the newsprint  markets  of  the UK and Germany,  the main suppliers  are  Finland,  Sweden 
and Canada.  These three countries account  for about 85 per cent  of newsprint  imports  into 
the UK and 68  percent  into Germany.  Both  countries have doubled their newsprint 
production  between 1980 and 1994 and increased the percentage ratio of domestic 
production  to imports. In 1980 the  ratio was 34 percent in the UK and 70 percent in 
Germany.  In 1994, the respective  ratios were 45 percent  and 116 percent.  (FAO  1991 and 
1994).  The UK is  far more dependent on imported  newsprint than  Germany. 
The  import  shares'  of suppliers  show  relatively  large  business  fluctuations  in  UK  sawnwood 
and newsprint  imports  over  time, while in Germany,  market shares have  been  rather stable 
(Figures  1-2). The  market shares  of suppliers  (see Table 1 in: Hänninen 1998 a  and  
Hänninen et ai.  1997)  have also  been relatively  similar indicating  that the supplier  countries 
have not  differed much from each other in terms of their potential  market  power. 
The main events  in the UK  sawnwood market were  the decrease of the  Russian share after 
the collapse  of the Soviet Union and large  fluctuations in the Canadian and Swedish  shares  
(Figure  1). Finland's share,  however, has  been rather stable.  In UK newsprint  imports,  
Finland has lost somewhat of its share,  while Swedish market share has  increased (Figure  
2).  The decline of the Finnish share in the UK has partially  been offset by  the acquisition  of 
1 Suppliers'  percentage  import shares  are  calculated by  dividing  the sawnwood and newsprint  import 
quantity of  each  supplier  by  the respective  total imports  to the  UK and Germany.  The  data for 
Figures  land 2  are  obtained from the British and German national customs  statistics (CSO,  
Statistiches Bundesamt). 
16 
production  capacity  by  Finnish  forest  industries in the UK market during  the 1980 s.  In the 
19905, Finnish firms  have also established newsprint  production  capacity  in Germany.  
Figure  1. Supplier  countries' percentage shares  of  total sawnwood imports  into the United 
Kingdom,  1980-1994. 
Figure  2.  Supplier  countries' percentage shares  of  total newsprint  imports  into the United 
Kingdom  and Germany,  1980-1994. 
17 
Supplier  countries' export  prices  for  newsprint  and  sawnwood are  presented  in Figures  3-4. 
If  suppliers'  prices  in the eport  markets  differ,  suppliers  must have at least some market 
power. 
Figure  3.  Logarithms  of  quarterly  real  unit prices  of newsprint  imports  into the United 
Kingdom and Germany  (GBP/metric  ton  and DEM/ton),  1980-1994. 
18 
From the  graphs  of  the  price  series
2
,
 it  could be concluded that  the  suppliers'  real  unit prices  
have developed,  relatively  uniformly,  in newsprint  markets.  After  a  stable price  development  
during the 1980  s,  there has been a fall in real  newsprint  prices  in both importing  countries  
during the economic recession of the early  1990 s  (Figure 3). 
Price development  in the sawnwood market has  been different. Graphs  of  real  sawnwood 
prices reveal that,  especially  in the second half of the period,  price  series have clearly  
deviated from each other in the UK  (Figure  4).  Therefore, the LOP is  more probably  valid 
for newsprint  than  for  sawnwood. However, explicit  tests  taking  into account  time series 
properties  of the data are required  to obtain statistically  accurate information about the 
validity  of the LOP.  
Figure  4. Logarithms  of  quarterly  real unit prices  of sawnwood  imports  to the United 
Kingdom  (GBP/m
3
),  1978-1992. 
2
 The time series  of  prices  are  suppliers'  quarterly  import  unit values into the UK and Germany 
calculated by dividing the value (CIF,  including cost,  insurance and freight) of the suppliers'  exports  
by  the respective  quantity. The  sources  of  the data are  British and German national customs  
statistics (CSO,  Statistiches Bundesamt). Unit prices  were  deflated by the UK and German producer  
price  indexes  obtained from the Main Economic  Indicators (OECD).  More information on the data 
can be found in Hänninen 1998 a and Hänninen et ai. 1997. 
19 
Graphs  of Finnish newsprint  prices,  and exchange  rates  with the destination countries (see  
Hänninen and  Toppinen  1998: Figure  1) indicate that these series are negatively  correlated 
with each other. This suggests  some degree  of exchange  rate  PT in both markets;  FIM 
depreciations  have lowered  the price  of  exports  from Finland and FIM  appreciations  have 
raised it. This  also  seems  to  hold for  Finnish sawnwood exports  (Hänninen  1998b: Figures  
1C and ID).  
During  the period  studied,  the FIM fluctuated widely  with respect  to  GBP  and Deutsche  
mark (DEM)  (Hänninen  and Toppinen  1998, Figure  1A). Moreover, the exchange  rate  
regime  was revised several times. During  the period  1978-1991,  the FIM was expressed  
with  a  currency basket  index which  was  a  trade weighted  average of  the exchange  rates of 
Finland's  most  important  trade partners.  In 1984,  the peg  was  changed  when the Soviet 
currency was  removed  from the index. The currency basket  index was  supposed  to be  kept  
within a band or  a target zone. Under this system the devaluations and revaluations were 
made by  specific  decision of Finnish government. The FIM was  revalued  in 1979,  1980,  
1989 and devalued in  1978, 1982. In June 1991, the FIM was  linked to  ECU basket  upon 
application  of  EU membership.  However, as  a result  of  speculative  attacks  the FIM was  
devalued in November 1991. During  1992, the FIM with several  other European  currencies 
came again  under massive speculative  attacks  and the Bank of  Finland decided to  abandon 
the peg  to  ECU.  In  September  1992,  the FIM was  devalued and it was  allowed to  float until 
October 1996, when Finland decided to  join the European  Exchange  Rate Mechanism 
(ERM).  
20 
3. THEORETICAL  FRAMEWORK AND EARLIER  LITERATURE  
3.1 Trade models and market  competition  
The traditional theory  of foreign  trade explains  trade entirely  with differences between 
countries, especially  differences in their factor endowments. It assumes  comparative  
advantage,  perfectly  competitive  markets,  homogeneity  of goods  and constant  returns-to  
scale  technology  (e.g.  Helpman  and Krugman  1985). However, traditional trade theory  
cannot  explain  why  industrial  countries have large  exports  and imports  of relatively  similar 
products.  In reality, many industries  cannot  be characterized  by  constant  returns-to-scale. 
Modern approaches  to  trade theory  explain  that  economies of  scale  (i.e.,  increasing  returns  
to-scale)  in industrial production  technologies  will  lead countries to  specialize  in  production  
and then trade with one another. It also assumes  imperfect  competition  in the market. 
Large  forest  sector  models are commonly  based on  the assumption  of  perfect  competition
3
.  
For  example,  ETTS V presents  forecasts  for different countries' forest  products  exports  
and imports in this framework (Brooks et  al. 1995). The Swedish forest sector  model 
assumes  perfect  competition  in West European  markets (Byström  and Eriksson 1991). The  
same assumption  is  also  used in partial  equilibrium  model for the global  forest sector,  The 
General Trade Model, GTM, (Dykstra  and Kallio 1987) and in the annual medium-term 
econometric model for  the Finnish  economy (KESSU)  (Hetemäki  and Kaski  1992). The 
quarterly model of Finnish economy (BOF4)  assumes  perfect  competition  for Finnish 
industry  exports  in the long-run,  but  imperfect  competition in the short run (The  BOF4, 
Quarterly  Model... 1990). Perfect competition is also assumed in some small partial  
equilibrium  models of trade in individual forest products. For example,  Buongiorno  and 
Gilless (1984),  who modelled newsprint  trade, Andersen (1993),  who studied European  
newsprint  markets,  and Brännlund et al. (1982),  who modeled Swedish exports  of forest  
products  to Western Europe,  assumed perfect  competition.  
3
 In  a perfectly  competitive  market  there are  many buyers  and  sellers  and none of  them has  a  large  
market share. Sellers are price-takers and  they  can only  adjust  their quantity  at  the current world 
market price.  Free  entrance  to the  market and  free exit  also  characterize the  existence  of  perfect  
competition. 
21 
On the other hand,  several other studies have assumed imperfect  competition
4
.
 Some 
previous  models of the newsprint  industry  in North  America  (e.g., Dagenais  1976, Booth et 
al. 1991)  assumed  oligopoly  with dominant firm price  leadership.  Brännlund and Löfgren  
(1995)  assumed that Canadian pulp  and newsprint  industry  is  a price  setter in North 
America,  but a  quantity setter  in its  European  exports.  In the study  of  Wiberg  (1987),  
Swedish  pulp  exports  to Germany  are assumed  to  follow price-setting  behavior. Oligopoly  
was  assumed  by  Halonen (1990),  who  studied Finnish  and Swedish pulp  and paper exports  
and Ronnila (1995),  who analyzed  Finnish paper exports  and tested Cournot oligopoly  in 
Europe.  Overall,  the results indicated the possibility  of noncompetitive  behavior in the 
Finnish paper industry.  In a  number  of  studies  that  have applied  the theory  of  Armington  
(1969),  imperfect  competition  is assumed in the trade of individual forest products.  
Imperfect  competition  was  assumed,  for  example,  in US  softwood lumber imports  from 
Canada (Hseu  and Buongiorno  1993), in Finnish sawnwood exports  to Europe  (Hänninen 
1986) and the UK (Tervo et al. 1988), in Finnish paper exports  to  the UK  (e.g.,  Laaksonen 
et al. 1997)  and in Finnish paper exports  to the UK and Germany  (Volk  1983). 
In  sum,  some previous  studies  of  forest  product  exports  have assumed  perfect  competition,  
while others have  assumed imperfect  competition.  Forest product  markets include many 
different products  and product  groups and each of  these may show varying degrees  of 
competition  in different  countries and  in different time periods.  Thus,  it is  difficult to  obtain 
a  uniform view  as  to  the degree  of  competition  in their international trade. Even  though  the 
traded products  in question  were  homogenous,  thus  suggesting  more  perfect  than imperfect  
competition,  the existence  of  market imperfections  remains a  possibility.  
In forest industry,  there are  attributes that  would suggest perfect  competition.  In pulp  and 
paper industry,  product  standards and specifications  are  fairly  universal,  exporters'  foreign  
investment in capacity  has increased and mergers and acquisitions  are common (e.g.,  
4
 Imperfect  market  competition  is  usually characteristic  of  markets  in  which there are  only  a few 
major  producers  as  well  as  markets,  where each  producer's  product  is  seen by consumers as  
differentiated from those of  rivals.  In imperfect  competition,  each  producer  may be  a  price-setter,  
unlike in perfect  competition.  In imperfect  competition, a  producer  faces  a  downward sloping  
demand curve  indicating that the quantities sold  increase when prices  fall. 
22 
Canadian Forest Service 1996). In the sawnwood industry,  the products of different 
suppliers  are  also relatively  homogenous.  
However,  there are  also features suggesting  market imperfections.  The  concentration 
development  of forest  industry,  especially  in the paper sector,  does not  necessarily  promote  
competition.  It has also been common to forest industry  that firms voluntarily  cut  
production  to be  able to maintain certain price levels,  instead  of competing  on price 
(Ronnila  1995).  In  the paper industry,  the investment  costs  of  new plants  are high,  which  is 
likely  to deter new firms from entering  the market.  Also, exit  from the market may  be 
complicated.  Due to the high investment costs, it is not  easy  to eliminate the excess  
capacity,  if it exists.  Exit from the market requires  that high-priced production  plants  are 
sold  to  competitors.  In  sawnwood industry,  imperfections  in competition  can  be  assumed to 
arise from regionally  concentrated trade and differences in building  codes, product 
dimensions and grades  between countries (Canadian  Forest  Service 1996). 
According  to Buongiorno  et al. (1979,  p.  643)  systematic  distortions are  also  possible  in 
commodity  markets in the case where products  are  technically  relatively  homogenous.  For 
example,  preferences  for  lumber quality,  even of the same species  and grade,  business 
practices  and habits as well as  attachments to a particular  supplier  country or  even to a 
specific  supplier  may  differ depending  on  the importing  country.  These effects are  related to 
switching  costs that consumers  or  other  end-users may face when switching  between brands 
or  suppliers  of  a  product  (e.g.,  Froot and Klemperer  1989). Okun (1975)  stresses  the costs 
of breaking  personal  sales relationships  in industrial transactions. Supplying  firms may 
create  relationships  with customers  by,  e.g., using repeat-purchase  discounts or making  
services  incompatible  with those of other firms.  
3.2. Modeling  export  markets using  the  Armington  approach  
The traditional export demand theory  of Armington  (1969)  assumes  that  products from 
different countries of  origin  are  imperfect  substitutes for  each  other. The model has  been  
widely applied in studies of international trade of various commodities,  for example,  US 
23 
cotton  exports  (Babula  1987, Duffy  et  al. 1990),  world  trade in rice  (Ito  et  al. 1990)  and 
fruits  and vegetables  (Sarris 1983)  and  for  Finland's foreign  trade in services  (Miikkulainen  
1989).  
Armington's  approach  has  been applied  to the modeling  of forest products  trade by,  e.g., 
Chou and  Buongiorno  (1983),  who studied the US  demand for  plywood  imports by  country  
of  origin.  Blatner (1989)  estimated the  price  elasticities for Western European  countries' 
imports  of paper products  by  supplier  country.  In Finnish studies,  the model has been 
applied  in modeling  sawnwood exports  to western  European  countries (Hänninen  1986),  
and to the UK market (Tervo et al. 1988)  as  well as  in paper exports  (Volk  1983 and 
Laaksonen et. al 1997).  
The popularity  of the Armington model derives from the simplicity  with which it estimates 
complicated  substitutability  relationships.  The model assumes  that  the  consumers' utility 
function is homogeneous  of degree one in prices  and weakly  separable  so  that the 
consumer's decision process  occurs  in two  stages.  First,  the total quantity  of imports  of a  
product  is  determined,  and  then second,  the  quantity  to be imported  is  allocated among 
competing  suppliers  using  constant  elasticity  of substitution (CES)  demand functions. Weak 
separability  in consumption  (or  production)  has  the advantage  that  it saves degrees  of  
freedom in the estimation. This makes  it possible  to  reduce  the number of  variables in the 
analysis.  It is  an essential advantage  in models of international trade that invariably concern  
many commodities and several  exporting  and  importing countries. Too many commodities 
in the consumption  function  and  too  many separate inputs  in the production  function would 
not  lead to a meaningful statistical analysis.  
The  restrictive assumptions  of the model, related to  the weak separability, are not  usually  
tested in applications.  This  has  been considered to  be problematic  in the interpretation  of  the 
results  (e.g.,  Davis  and Kruse 1993). However, testing  is not  always  possible  because of a 
lack  of  data. In many studies,  when testing  is made, the underlying  assumptions  of  the 
model have not  always  been satisfied ( e.g., Alston et al. 1990). This is why  some of the 
applications  have refined the original  assumptions  of  the Armington  model,  for example,  by  
relaxing  the homogeneity  condition (Ito  et al. 1990). Also, more general functional forms 
24 
and models which account  for both non-homogeneity  and the existence  of variable 
elasticities of  substitution have been presented  (e.g.  Henning  and  Martin 1989).  
A  more  flexible functional form, the translog  form,  presented  by  Christensen et  ai.  (1973),  is  
applied  in the present  study  (Hänninen 1994). This generalizes  the original  model from two 
to many factors and relaxes  the restrictive assumption  of constant  elasticity  of substitution 
applied  by  the Armington model and  most  of  its  applications.  Thus, it  was  possible  to  obtain  
substitution elasticities  for all possible  country-pairs  separately.  Translog  cost  functions  
have been common in evaluating  factor substitution and economies of scale in forest (and  
other)  industries (e.g.,  Fuss  1977, Martinello 1985, Puttock and Prescott 1992, Hetemäki 
1990). 
The present study  formulates an export  demand model system to estimate substitution 
elasticities between Finnish,  Swedish, Canadian and Russian sawnwoods in the  UK market 
(Hänninen 1994). Because sawnwood is  an intermediate product,  the model is  derived from 
a cost  function of the construction industry  and not  from the consumer's utility, as in 
Armington  model. The two-stage demand model for sawnwood implies  that a representative  
end-user in the UK  first decides how much sawnwood and other (aggregate)  inputs  are 
required  to produce  a certain output.  After this,  he chooses between sawnwood from 
different countries of  origin,  including domestically  produced  sawnwood. 
Assuming  cost-minimizing  behavior for the sawnwood end user  in British construction 
industry,  the optimal sawnwood demands from different origins  can be derived from the end 
user's  cost  function using  Shephard's  Lemma. The cost  function is homogenous  of degree 
one in prices  and it is  assumed  to  be translogarithmic  (Christensen  et ai.  1973).  Thus,  the 
demand functions in terms of shares  can be presented  as 
(3.1) (ainC/ain  Pi)  =PiXi/C= Si  = Pi  +ZsijlnPj,  
25 
where 5,  = P,X,/C  = the cost share of the /'th  (z'=1,...,5)  supplier  country  and i denotes 
Finland,  Sweden, Canada,  Russia  and the rest  of  the world. Pj  is  the respective  price  of  the 
jth (j=  1,...,5)  supplier.  The estimation of  the system  using the iterative Zellner estimation 
method (e.g.,  Pindyck  and  Rubinfeld 1988)  (with  the adding  up criterion i.e.,lSj  =l,  and the 
homogeneity  and  symmetry  conditions imposed)  provides  parameters from which the own  
price  and substitution elasticities can be calculated for each observation. 
The  translogarithmic  model is  more suitable in examining  substitution between many 
countries than the earlier models which assumed constant  elasticity  of substitution. 
However,  the problems  of  nonstationary  (i.e., trended)  data  makes  the interpretation  of the 
results difficult. As  a drawback in using a translogarithmic  model, cross-equation  
restrictions are  problematic  in  the  estimation if  time series  properties  are  taken into  account.  
For  example,  in the Johansen method (Johansen  1988) coefficient restrictions between 
cointegration  vectors  are  not  possible.  
3.3 Testing  the law  of one  price  in forest products  markets 
The Armington  (Armington  1969)  model assumes  imperfect  substitution  between products  
from different countries of  origin.  This implies  that, in the presence  of  price differences, 
substitution between supplier  countries' products  is  measurable. The existence of  price  
differences is  explicitly  studied by  testing  the law of one price  (LOP)  in the UK  and  German  
markets  (Hänninen  1998 a  and  Hänninen et  ai 1997).  The  law of one price  is  a  variant of 
Purchasing  Power Parity  (PPP).  PPP is  a  long-run  relationship  and states that, once  
converted to a common currency, national price  levels  should be equal.  A large  amount  of 
empirical  evidence on  PPP shows  that the real  exchange  rates  (nominal  exchange  rates  
adjusted  for  differences in national price  levels) tend toward PPP  only  in the very  long  run,  
the speed  of  adjustment  being extremely  slow, and  that short-run  deviations from PPP are 
large  and  volatile (Rogoff  1996). 
The origins of  empirical  PPP parity  studies trace  to  the debate on how to restore  the world 
financial  system after its collapse  during  World War I,  when gold  standard  was  abandoned 
26 
(Rogoff  1996).  Prior to  the war,  the exchange  rate  between two  currencies  tied to  a  gold  
standard simply  reflected their relative  gold values. After the war, countries faced the 
problem of deciding  how to reset  exchange  rates  with  minimal disruption to  prices  and 
governement finances. Cassel  (1921)  was  the first  to  study  the subject  empirically  proposing  
to use of inflation differentials between countries to  calculate the exchange  rate  changes  
needed to maintain PPP. Various versions of  PPP are  used in  a  vast number  of applications,  
e.g., to  choose the right  initial exchange  rate  for a  newly independent  country,  to  forecast  
real  exchange  rates  or  to try to adjust  price  differences in international comparisons  of  
income (Rogoff  1996). 
The law of  one price  is  a  building  block  of  PPP,  and it states that for any good  i :  
where p,  is  the domestic-currency  price  of good  i,  pc  is  the foreign  currency price  and  er  is  
the exchange  rate  defined as  the home-currency  price  of foreign  currency. Thus, LOP 
implies  that once  prices are  converted to  a  common  currency,  the same  good  should sell  for  
the same price  in different countries. 
The failure of  the LOP is  found in  large  number of  studies.  Of  the early  studies  most  notable 
were Isard (1977),  who found deviations form the LOP between U.S., German,  Canadian 
and Japanese  exports  of  goods,  such  as  industrial chemicals,  paper and glass.  Richardson 
(1978)  found only  some  evidence  of  commodity  price  arbitrage  in the exports  of  industrial 
products  between U.S. and Canada. According to Giovannini (1988) price  differentials 
existed  in  manufacturing  goods.  He  also  found, as  Isard  (1977)  and  Richardson (1978),  that 
LOP deviations are highly correlated with exchange rate  movements. Also, studies 
investigating  export unit values from a  single  country  to multiple  destinations have found 
deviations from the LOP (e.g.,  Knetter  1989, 1993). 
Several reasons  for  LOP  devations in traded goods  can also be found in the literature. For 
example,  tariffs,  nontariff barriers  and transportation  costs  may  cause  differences between 
prices  in different countries. For some classes  of goods, differing national standards may 
(3.2) pi=erpc,  
27 
weaken arbitrage. These different adjustment  costs form a buffer within which nominal 
exchange  rates  can  move without producing  an immediate proportional  response in relative 
domestic prices.  Deviations from LOP  in traded goods may  also  be  explained  by  "pricing  to 
market" theory  (Krugman,  1987 and Dornbush,  1987). In this framework,  oligopolistic  
suppliers  can charge  different prices  for the same good  in different countries. 
Conventionally,  the LOP has  been tested by  regressing  a  commodity  price  in one country  on 
the  same commodity  price  in another  country  in common currency. Then, the slope  
coefficient is  restricted  to  equal  one and the intercept  to  equal  zero.  If  the restrictions  are  
not  rejected,  the law is  concluded to hold. The flaw  in these tests was  the failure to  take  
account  of the time series properties  of the variables. This was  especially  criticized  in the 
1980's and,  of the early  studies  taking the nonstationarity  of  the variables into account,  can  
be mentioned, e.g., Ardeni (1989),  who used a  bivariate Engle  and Granger cointegration  
method (1987)  to test  the LOP. 
Single  equation  methods,  however,  inefficiently  model the interaction between testable price  
variables and fail to  take  account  of  possible  dynamics  between them (e.g.,  Edison et al. 
1997).  More recently,  systems  methods have been developed,  e.g., by  Johansen (1988).  
With regard to LOP  tests with more than two price variables,  single  equation  methods are 
unable to find more than  one cointegration  vector, while according  to Goodwin and 
Grennes (1994),  full  integration  of  the markets  requires  n-1 cointegration  vectors  among n 
variables. 
Although  there is a  wide literature on testing  the law of  one  price for  different commodities,  
there are  few results  for forest products.  For example,  Buongiorno  and Uusivuori  (1992),  
tested price  differences in US pulp  and  paper exports  to  western  European  countries and 
Japan  and  found evidence for  the LOP.  They  used the cointegration  approach  and  a  Dickey-  
Fuller type bivariate method. Johansen's cointegration  method  was  used by  Alavapati  et  al. 
(1997),  who examined the price  formation of  Canadian pulp  exports.  They  found support 
for  the LOP between Canadian and  American pulp  prices.  Jung  and  Doroodian (1994)  used 
the same method in testing  the law of one price  between four  regional  softwood lumber 
markets  in the United States. They  interpreted  the result  as  supportive  of  the LOP, but  the 
28 
result  is  problematic,  because  it does not  reveal  for  which  of  the four prices  the law  holds. 
In the present study,  the Jung and  Doroodian analysis  was  extended by  conducting  
additional tests. These tests help  to determine to which specific  prices  the law could 
possibly  apply.  
3.4 Exchange  rate pass-through  on  forest  product  export  prices 
Pass-through  (PT)  effects of  exchange  rates  on  prices  have been analyzed  for  trade in many 
commodities and for aggregate imports  and exports  (see,  e.g., the survey by  Menon 1995) 
and the effects  are found to  vary widely by industry  (e.g., Feenstra et al. 1996). The 
variation is  related to  industry characteristics,  such as the degree  of competition,  product  
substitutability  and the relative (domestic  and foreign)  market shares (see,  e.g., Officer  
1986, Menon 1995, Yang  1997). 
The large  literature on PT arises from the discussion  of  fixed versus  flexible  exchange  rates  
in international economics. After the breakdown of the Bretton Woods  system  of 1971-73, 
explanations  were  demanded for  the question  wether floating exchange  rates  would play  an 
equilibrating  role in trade balances. In this  context, the underlying  relationship  between 
exchange  rates  and the prices  of internationally  traded goods  have been widely  examined. 
Following  Dornbusch (1987), two  extreme  models  to measure  price  relationships  in the 
open economy  literature can  be presented.  One  assumes  that the law of one price  holds in 
the market  (equation  3.1),  in which case LOP belongs  to PPP  literature. The alternative 
model assumes  that products  from different origins  are  imperfect  substitutes. Then, the 
relation between domestic and foreign  price  levels (real  exchange  rate)  can be presented:  
where p ,■ is  domestic price  level,  er  is  nominal exchange  rate  and pc is  the foreign  price  
level. Under imperfect  competition,  price  will not  equal  marginal  cost  as  firms have the 
possibility  to  charge  a  markup  on  production  costs.  If  the markup  is  constant, exchange  rate  
(3.3) §  =  pi/erpc,  
29 
movements  change  relative prices one-for-one,  but, if the markup  is  variable,  the issue  
becomes how the markup  might  vary  in response  to  an exchange  rate  change.  The reaction 
of markups  is affected by,  for  example,  the degree  of  substitutability  between domestic and 
foreign  goods  and the degree  of  market integration  or  separation.  The lower  the degree  of 
substitutability  between goods,  and  the lower the degree  of market integration  (the  presence  
of segmented  markets),  the greater will be  the market power  of suppliers  (Menon  1995).  
The  degree  of  the exchange  rate  effect on prices  can  be measured by  the exchange  rate  
pass-through  (PT) coefficient,  i.e., the percentage  change  in export prices  (denoted  in 
foreign  currency)  associated  with a  one  percent change  in the exchange  rate.  The  magnitude  
of the PT may be O<PT<l  and it is  related  to the market structure  as follows:  If PT is  close 
to zero,  competition  is  perfect  in the foreign  market and exporting  firms have no market 
power  to  affect their foreign  currency  prices  as a  result  of  a  change  in the exchange  rate.  If 
PT is  unity,  pass-through  of  exchange  rate changes  to foreign currency export  prices  is  
complete  and competition  is imperfect.  PT  may be O<PT<l, indicating  incomplete pass  
through and imperfect  competition.  It can  be concluded that the closer PT  is to unity, the 
more imperfect is market  competition.  
However, the interpretation  of  the relationship  between  PT  and  the degree of  competition  is  
not  always  straightforward  due to problems  in measuring  PT.  For  example,  it is  possible  
that PT  is  zero  or  near  zero  in imperfect  competition  if there are long-term  contracts  that  
fix  foreign  currency prices  for  months  ahead. In this case  exchange  rate  changes  can  pass  
through  to  foreign  currency prices  only  with lag.  On  the other  hand,  PT may  be higher  than 
zero  (incomplete)  in perfect  competition.  For  example,  producers  may increase  their export 
supply  as  a  result of a  growth  in markups  caused by  devaluation. The increment of  total 
supply  in the export  markets  usually  presses  down the foreign  price  level and the magnitude  
of  PT elasticity  becomes higher  than zero.  This difficulty  in interpreting  the PT elasticities 
must  be remembered when interpreting  the results of the present  study.  
Unlike other commodities, PT has  not  been extensively  studied for forest products  and the 
few  previous  studies  deal mostly  with  the USA. Earlier results  indicate that PT is  higher  for 
wood industry  products  than  for paper products,  but  conclusions are difficult to draw 
30 
because  of  the scarcity  of  results.  In a study  by  Vesala (1992), PT for  export  prices  of  
Finnish paper products  was  found to  be between 0.66  and 0.69 for western  Europe,  and 
between 0.16 and 0.30 for  the USA. The smaller PT for the USA was  assumed to result 
from the large  domestic market in the USA,  in which the Finnish  share is  much smaller than 
it is  in Europe.  Alavapati  et  al.  (1997)  found  that the exchange  rate  effect  on  Canadian pulp  
price  was  small,  suggesting  that  Canadian exporters  do not  have significant  market power  in 
exports.  Uusivuori and Buongiorno  (1991) estimated the PT for  US forest  products  exports  
to  Europe  and Japan.  Pass-through  was  incomplete  in most  of the product  categories.  In 
lumber  exports  to  Japan,  PT  was high:  from 0.79 to 1.04,  depending  on the species.  Menon 
(1993)  also estimated a  relatively  high  PT (0.80)  for Australian imports  of wood products,  
but for paper and board,  the PT was  0.45. 
Many of the previous studies have examined exchange  rate  effects by  estimating  the 
bivariate relationship  between the destination currency  price  of exports  and the exchange  
rate  (e.g.,  Knetter 1989 and 1993, Pick  and Park 1991, Uusivuori and Buongiorno  1991).  
In  these studies,  as  in most  other  previous  studies  on  PT,  estimation is  made using  Ordinary  
Least Squares  procedures  (OLS).  Also, OLS in differences has  been applied  (e.g.,  
Athukolara 1991),  and dynamics captured,  using  polynomial  distributed lags  (e.g.  Hooper  
and Mann 1989). 
In addition to bivariate models,  multivariate models have  also been applied  assuming  that  
the markup  is  affected by  the exchange  rate,  foreign price  and demand pressure.  Time  series 
properties  of  data have usually  been taken account  in the studies  in 1990  s  (e.g.,  Athukolara 
1991, Athukolara and Menon 1994, 1995, Menon 1993, Hung et  al. 1993, Naug and 
Nymoen  1996, Alavapati  et  al. 1997).  In  earlier studies,  based on the markup concept  and 
applying  multivariate approaches,  single  equation  models have commonly  been estimated 
(see  however,  e.g.,  Kongsted  1996).  In  contrast,  the present  study  formulates a  demand and 
supply  model  for estimating  PT for sawnwood export  price  (Hänninen  1998b)  that takes 
into account  time series properties  and estimates a  multiple  equation  model. 
The  PT  coefficients of  the present  study  were  estimated from a  markup price  equation  that 
is commonly applied  in the estimation of PT (e.g., Hung  et al. 1993). For sawnwood 
31 
(Hänninen  1998  b)  the model also  includes an  export  demand equation  derived by  assuming  
a  two-stage decision making process,  as  in the case of  Equation  3.1. The  functional form  
assumes  constant  elasticity of substitution between supplier  countries' sawnwoods and the 
demand and price equations  are  formulated as  
where,  Xf  and pj  are the Finnish quantity  and nominal unit price  in FIM to the destination 
market,  x„ and p„ are  the respective  quantity  and price  (in  the importer's currency)  of  
competitors'  products,  er  is the nominal exchange  rate (FIM/GPB,  FIM/DEM),  c  and 8  are 
constant  terms and cs is  the Finnish  unit production  cost, e  and u are  disturbance terms that 
describe all the other factors. Parameter 77 is  the constant  elasticity of substitution (i.e.,  the 
price  elasticity  of demand)  and y (0<  y  < 1) indicates the PT  coefficient (PT  =  ~(d(p/  -  er)/9 
er)  =  y).  The demand equation  (3.4)  is assumed  to  be homogenous  of degree  zero  in pj,  p„  
and er  and the price  equation  (3.5)  homogenous  of degree  one in er, p„  and c/.  The price  
relation allows the markup  to  vary. A constant  markup  is  related to a  large  PT and a 
variable markup  to a  small PT.  In the case  of a  constant markup,  exchange  rate  changes  
affect foreign  currency-denoted  prices,  while in the case of a variable markup  home 
currency  prices  are  affected. It  is  assumed  that  the exchange  rate  is  an  exogenous variable in 
the model. 
(3.4) Export  demand: x/-x„ = -r] (p/  - p„  -  er) +c+ e, 
(3.5)  Export  price: pf =S +  (1-  y)(er  +  /?„)  +  jc/ +u ,  
32 
4. DATA  
The main body  of  the data consists  of quantities  and unit prices  of  the supplier  countries' 
sawnwood and newprint imports  into the UK  and  Germany. The estimation of the 
substitution elasticities between the main supplier  countries'  sawnwood in the UK  market is 
based on yearly  data from 1961-1990 (Hänninen  1994). The observation period  was  
selected based on  the availability  of  the data. In the rest  of  the analysis,  concerning  LOP and 
PT-elasticities,  quarterly  (unadjusted)  data are  used due to  the requirements  of Johansen's 
cointegration  method (Johansen  1995). The observation period  for  quarterly  analysis  was  
chosen  to begin  after the years affected by  the oil crises  in  the middle of  1970's so  as to 
avoid the estimation problems  caused by  large  amounts  of  turbulence in the development  of 
prices.  In the following section, the  data from all five studies is  briefly  described. 
For  the study concerning  substitution between supplier  countries in British  soft  sawnwood 
imports (Hänninen  1994), yearly  data  on import  quantities and values  were gathered  from 
Overseas  Trade Statistics of the United Kingdom  (CSO)  using  product  code SITC 248.2-3. 
Sawnwood prices  for the four  suppliers (Finland,  Sweden, Canada and Russia) were 
described by  import  unit values.  They  were  calculated by  dividing  the import  value (CIF,  
including  cost, insurance and freight) by  the respective  import  quantity,  m 3).  For  the group 
comprised  of  the rest  of  the suppliers  of  imports  the import  quantity  and unit price  were  
calculated as  a  residual from the figures  of total sawnwood imports.  
The studies  testing  LOP  in the UK  sawnwood market and  in the UK  and German  newsprint 
markets  also  used unit values of  imports  to describe suppliers'  prices  (Hänninen  1998  a, 
Hänninen et ai. 1997). The  observation period  was 1978-1992 for sawnwood and 1980- 
1994 for  newsprint.  The quarterly  data on  import  quantities  (m
3
 and metric  ton) and  values 
(CIF in the destination countries' currencies)  were obtained form the national customs 
statistics:  Overseas  Trade Statistics  of the United Kingdom  (CSO)  and Aussenhandel Nach 
Waren und Ländern (Statistiches  Bundesamt),  where quarterly  figures  were calculated as  
sums of the respective  months. The product  code, for  sawnwood,  is  SITC  248.2-3 and,  for 
newsprint,  is SITC 641.1 in the UK data. In the German data, the product  code for 
newsprint  is  CN 4801 0010 (including  item CN 4801 650  for year  1980, item CN 4801 790 
33 
for  years 1981-87 and  item CN  4801 0090 for  1988 onwards).  LOP  analyses  were  made by  
using real  price series.  The deflators for  the nominal unit  values were  producer  price  indexes 
(1985=100)  for the UK and Germany obtained from the Main Economic Indicators 
(OECD).  
PT effects  were  also  analysed  using  unit  values to  describe prices.  For  sawnwood,  the 
observation period  was  1978-1994. The variables  describing  sawnwood  quantities and 
prices  in British imports  were  obtained from the same  source,  using  the same product  code, 
as  in the study  by  Hänninen (1998  a).  However,  the data for the years  1991-94 were  
gathered  from EU-trade statistics  (European  Comission).  In order to obtain the Finnish 
price  in FIM  (pj), the Finnish unit  price  (in  GBP)  was  converted into FIM at the quarterly  
average nominal exchange  rate,  er  (FIM/GBP),  obtained from International Monetary  Fund 
(IMF). The  competitors'  price  (p„) is  the (quantity  share weighted)  unit price  of British 
imports  from Sweden,  Canada,  and  Russia expressed  in GBP.  The respective  quantity  (x„)  is  
the sum of imports  from these same three countries. Production costs of Finnish sawnwood, 
cj,  were  described by  the production  price  index (1990=100,  manufacture of  wood  and 
furniture),  obtained from the Bulletin of  Statistics  (Statistics  Finland).  
In the study  concerning  PT  for newsprint  and pulp  (Hänninen  and  Toppinen 1998),  the 
observation period  was  1980-1994. The product  codes for newsprint data are  the same as  
mentioned for the LOP study.  The codes  for  pulp  (bleached  sulphate  pulp)  data are  SITC 
251.51 and SITC 251.52 for the UK,  and CN 470321 and CN 470329 for Germany.  Finnish 
unit prices  of  newsprint  and pulp  (/?/)  in FIM were  calculated using export  values (FOB,  free 
on board)  and quantities,  which  were  obtained from the customs  statistics  of Finland  (The  
Board of Customs).  The foreign  competitors'  prices  (/?„) are the average CIF (including  
cost, insurance and freight) unit values (GBP/ton,  DEM/ton)  of total newsprint  or pulp  
imports  (excluding  imports  from Finland)  into  the UK and Germany.  The  source  of 
competitor data were the customs statistics  of the UK (CSO)  and Germany  (Statistisches  
Bundesamt).  Exchange  rates  (er)  between the Finnish markka and pound  sterling  and 
Deutsche mark are  quarterly averages obtained from the Bank of Finland. The producer  
price index (1990=100),  for the commodity  group (manufacture  of paper and paper 
products)  obtained from the Bulletin of Statistics (Statistics Finland)  is used to describe 
34 
Finnish production  costs  (c/)  for  the pulp  and newsprint  industry.  Because  separate indexes 
are  not  available for pulp  and newsprint, costs of these industries were described with the 
above mentioned common-cost  variable. 
The choice of  the empirical  variables for  the study  was  based on  the  availability  of  the data. 
This means  that  they  do not  necessarily  describe the theoretical variables of  the models well. 
For  example,  the prices  were  described by  average unit values because  proper  data on 
prices  or  more  disaggregated  data on the qualities  of each supplier's  products  were  not  
available. Thus,  it was assumed  that the suppliers'  sawnwood,  newsprint  and pulp  was 
homogenous  enough  to  allow reliable estimation results.  Newsprint  and  pulp  are,  in fact,  
relatively  homogenous  grades among different paper products.  The homogeneity  
assumption  is  also justified  for Finland,  Sweden and Canada because  these countries have 
relatively  similar  raw  material basis,  i.e.,  coniferous wood. For Finnish,  Swedish and 
Russian  sawnwood,  the assumption  is also  well  justified  as  producers  in these countries 
have much  the same raw  material and end-use sectors. Only  Canadian sawnwood may not  
be  exactly  comparable  to  the other suppliers'  sawnwood. 
The use  of  the production  price  indexes  as a  proxy  for  the domestic cost  of  the sawnwood,  
newsprint  and pulp  industry  may  cause uncertainty  in the  testing  of  PT,  but  they are  used 
because  proper cost  variables were  not  available. Another possibility  would have  been to 
describe the production  costs  by the price  of  raw  material. Quarterly  time series  for  prices 
of  sawlogs  and pulpwood  are  however available  only  from 1985, which  would have made 
the estimation period  too  short. 
35 
5. COINTEGRATION  METHOD 
5.1 Johansen's multivariate  cointegration  method 
In the Ordinary  Least  Squares  estimations of  many earlier bivariate models,  the  time series 
were  implicitly  assumed  to be constant  in mean and  variance over  time or  they  were  
assumed to be  stationary.  However, prices  of commodities are  usually  not  stationary,  which 
may  lead to  a  "spurious  regression"  problem  (Granger  and Newbold 1974). In  this  case,  the 
interpretation of the results  is  difficult, because the conventional t or F statistics  calculated 
for such  regressors  do  not  follow standard distributions  (e.g.,  Banerjee  et  al. 1993).  
To avoid the problem,  time series  were  often  differenced to  make them stationary,  but this  
led to  another problem;  important  long-run  information of  the data was  completely  ignored.  
When the idea of cointegration  between  the variables was  developed  it became possible to  
capture long-term effects  in dynamic modeling.  Granger  (1981)  showed that, even in the 
case  that all the variables in a vector  are  not  stationary,  there may be linear combinations 
between them that are  stationary,  i.e., the variables are  cointegrated.  The  cointegration  
concept  is  important  in economic modeling;  if  there is  an equilibrium  relationship  between 
the variables,  these variables should not  diverge  from each other to a great extent in the 
long  run.  
In the present  study,  conventional methods, such as  bivariate OLS and multivariate single  
equation  approaches  (Engle  and Granger  1987) were replaced  by  Johansen's (1988)  
cointegration  method. It  is a  convenient way  to  estimate  multiple  cointegration  vectors  and 
test  hypotheses  in a  simultaneous multivariate framework. This  method  is  further developed  
by  Johansen and Juselius (e.g., 1990 and 1992). Before estimation,  the variables of the 
present  study  were  tested using  ADF  tests,  the results  of  which indicated that the data for 
the study  are  nonstationary  (Hänninen  1998 a: Table 2,  Hänninen et. al. 1997: Table 2,  
Hänninen and  Toppinen  1998:  Table 1  and Appendix  1  of the present  summary paper).  
The basic statistical model of Johansen (1988)  is  the p -  dimensional vector  autoregressive  
model with Gaussian errors:  
36 
where x,  is  apx  1 column vector  of  empirical  variables,  |i is  apx  1  vector of  constant  terms, 
D, is  a  vector  of  non-stochastic variables (e.g., seasonal  dummies),  k  is  the lag  length  (k =  
1,...,N) and e,  is  a  vector  of error  terms assumed to  be NID(O,Q). It is  assumed  that x, is  at 
most 1(1). The purpose of the analysis  is  to distinguish  between stationarity  by  linear 
combinations and by differencing. This is  why  model (5.1) is reformulated in the error  
correction form (e.g.,  Johansen 1995,  p.  89): 
where Ax, is  an 1(0) vector.  The  restriction of ji to cointegration  space  means that no linear 
trend is assumed in the data. F;,...,  Fw  and FI  =- 1 +IT +...+ 11* are coefficient matrices. F 
describes  short-term dynamics  of the process  and II is  the matrix of long-run  coefficients. 
The  proper  lag  length,  k,  for (5.2)  can  be chosen by  diagnostic  testing  of  the residuals  of 
individual equations  and by  using  the Schwarz  (SC)  and Hannan-Quinn  (HQ) information 
criteria (see Doornik and Hendry  1994). The reformulated model (5.2.)  is a traditional first  
differenced VAR model except for the  term ILt/.*.  
The  matrix II  has a reduced rank in the case of cointegration  (rank  n  <  p) and II can 
therefore be  decomposed  into  a matrix of loading  vectors,  a,  and a  matrix of cointegrating  
vectors,  (3,  such that n=a[3'.  The rank of the matrix n determines the number of 
cointegration  vectors.  When all  p  components (variables)  of  x,  are  stationary,  matrix n has 
the full rank  p.  When no cointegration  exists  between the variables in xt,  the matrix  n  has 
zero rank.  In this case, the variables are integrated  of order 1 and the VAR model in 
differences would be a proper framework for empirical  analysis.  
The Johansen test generalizes error  correction models to a multivariate framework and 
examines  the coefficient matrix II of the lagged levels to  extract information about long-run  
relationships.  Johansen (1988)  formulates two likelihood ratio tests', the trace test and the 
(5.1) Xi =Aix,.i  +...+ Ak  Xt-k  +n + ODi + £,, t =1,...,T, 
(5.2)  Axi Ti ÄXt-i r k-iAxt-k+i  + ri*/.*  + ji. + OZ)/  + St 
37 
maximum eigenvalue  test,  for the rank  of  n, which gives  the number of  cointegration  
vectors  in the set of  p  variables. The null hypothesis  is: 
where a  and (J  are  pxr  matrices. The  likelihood ratio test  statistic for the hypothesis  that 
there are  at most  r  cointegration  vectors  is: 
where X\  s  are  the squared  canonical correlations,  i.e., eigenvectors  (k'i  > >...>  X'
P
 ) 
between the two  sets  of residual vectors  RO,  and Ri,.  Residual Rol is  obtained from regressing  
(Ax,  )on the lagged  differences and the deterministic variables  (Ax,.i,...,Axt .k+i D,)  and 
residual R t, is obtained from regressing (x,.k  ) on  (Ax,.i,...,Axt.k+i  D, ). The resulting  
eigenvectors  V" = [v'/ (...,v'p] are  usually  normalized as  V"S uV= Ip.  The estimators of 
cointegration  vectors  (3 are  P =  [v';,...,v'r], the eigenvectors  corresponding  to  the r  largest  
eigenvalues.  The ML estimator of  the cointegration  space  P reflects  the r  largest  canonical 
correlations between the stationary  and  nonstationary  parts  of  the system.  It  is  important  to  
note  that the estimation  method  produces  only  the cointegration  space. Thus,  it  is  possible  
to give  economic interpretation to the cointegration  vectors  only  after identification. 
In testing  the rank by  the trace  test,  the null  hypothesis  is  tested in each  stage,  against  the 
general  alternative  that  there are  at  most  r  cointegration  vectors.  The maximum eigenvalue  
test is given by the maximal likelhood statistics.  The null  hypothesis,  that there are r  
cointegration  vectors,  is  tested against  the alternative that the rank  is  r  :  H0 :  rank(n)  =r  + 
1. The test statistic  is Lmax(r)  = -7"  ln(  1  -A.' r+i). 
The asymptotic  distributions of  these likelihood ratio tests  can  be presented  as  multivariate 
versions of the Dickey-Fuller  distribution and the distributions are also  provided  by 
Johansen and Juselius (1990).  The present study  applies the critical  values at five percent 
(5.3)  Ho: rank(n)  <r  or  n=af3',  
p 
Trace(r)  = -T ) 
i=r+l 
38 
significance  level  obtained in the connection of  the estimation by  the FIML econometric 
program (Doornik and Hendry  1994). 
When the number of  cointegration  vectors, r,  has  been determined,  it is  possible  to  test  
hypotheses  under r  by  imposing  linear restrictions  on  the matrix of  cointegration  vectors,  (3  
(=((3/ pr))  and loadings,  a (Johansen  and  Juselius (1990,  1992).  a,;/measures  the weight  
with which the yth cointegrating  vector  ((3,)  enters  the ith equation  of the system.  
Hypothesis  testing  is  done by  estimating  the model (Equation  5.2)  with restrictions  and 
comparing  the results to  those of the unrestricted model. The tests  for these linear 
2 
restrictions  are  asymptotically  %  distributed. The critical values  used in this  study  were  
obtained in the estimation by  the FTML econometric program at 5  percent significance  level  
(Doornik  and Hendry  1994). Under r,  the Johansen method (Johansen  and Juselius 1990,  
1992)  formulates  different types  of  restrictions on  P or  a: 
The alternative hypothesis  for (5.4-5.6)  is  (5.3).  Symbol  p  is  the dimension of the model 
and ris  the number of cointegration  vectors.  Matrices H|(pxs),  H2(  pxs) and A(pxm) are 
the design  matrices,  where  the restrictions on the parameters  P{pxr)  and a(pxr)  are 
formulated by  the researcher. The  restrictions reduce  the parameters to (p(sxr)  and \|/(mxr), 
where r<s  <p  and r<  m < p. 
In  the  hypothesis  (5.4)  the (p-s)  restrictions  are  placed  on  r  cointegration  vectors.  The test  
statistic  is  asymptotically  %
2
 distributed with r(p-s)  degrees  of freedom (Johansen  1988). 
The  hypothesis  (5.4) is applied  here  to  test the exclusion of a variable from the 
cointegration  relation and price homogeneity.  Hypothesis  (5.5)  is used to test the 
stationarity of variables. In this  test  the null hypothesis  is  stationarity,  unlike in the Dickey-  
Fuller (1979)  test. In (5.5),  r/  of  the r  cointegrationvectors  are  known a priori. Restrictions 
for these known vectors are presented  in the design matrix H2 and the remaining  r2 
(5.4) (3 =  Hi(p,  where Hi(pxs),  (p(iXr)  and  r<s<p  
(5.5)  P=  (H 2,V|/),  where H2(px n),  i|/{px  r2)  and r =r t  +r2 
(5.6) a  = A\|/,  where A(px m), \|i(mx r)  and r<m<p. 
39 
cointegration  vectors (included  in matrix \|/) are unrestricted. The test statistic is 
asymptotically  %
2
 distributed with (p  -  r)ri degrees  of freedom. 
Weak exogeneity  of  variables  with respect  to  long  run  parameters  a  and (3 can  be tested 
using  hypothesis  (5.6).  Weak exogeneity  is required  for  efficient  inference  (estimation)  of a  
conditional model. Whether a  variable  is  exogenous depends  upon whether it can  be taken 
as  "given" without losing  information for  the purpose  at hand. Weak exogeneity  of  the Zth 
variable for the long-run  parameters P  implies that  all the elements of  the Zth  row  of  a  are  
zero, i.e., no lagged levels appear  in that equation.  The test  statistic  is  %
2
 distributed with 
r(p-m) degrees  of  freedom. 
5.2  Testable hypotheses  
In the UK sawnwood market,  the  LOP was tested between Finnish  (pf), Swedish (ps),  
Canadian (pc) and Russian (pr) prices  of exports  (p  = 4) and in the newsprint market 
between Finnish,  Swedish and Canadian prices  (p  = 3).  For  example,  in the case of four 
price variables, LOP can be written pf  = ps  =pc= pr  in equilibrium.  This  can also  be 
presented  as pf  = ps,  ps  = pc,  pc  = pr.  This  implies  econometrically  that the long-run  
LOP requires  three stationary  relationships  for sawnwood (r  = 3)  and two  for newsprint  
(r= 2):  
To test (5.7), model (5.2),  with three seasonals  and a constant  term was estimated and 
Johansen's hypothesis  (5.4.)  was  used. The second stage  of  the testing  was  the pairwise  
test, in the cases  where the rank  test supported  r  =l. Under one cointegration  vector, the 
law cannot  hold for four (three)  prices  simultaneously.  Instead,  it can only  hold for  two of 
the prices  To examine for  which  two prices  LOP  holds it was  tested whether the following 
restriction on  the coefficients  of  the cointegration  vector  ((30  holds under r  =  1 
(5.7) sawnwood: pf-ps  = 1(0), ps  -pc =  1(0), pc-pr  = 1(0), 
newsprint:  pf-ps  =  1(0), ps  -pc = 1(0) 
40 
(Johansen's  hypothesis,  5.4). For example,  for sawnwood prices  {pf  ps,  pc,  pr)  six  design 
matrices (Hi) of the form (1,  -1, 0,0), (1,0,-1,0), (1,0,0,-1),  etc. were formulated to test 
the restrictions under r  =1 (Hänninen  1998). The rejection  of the restriction (by  %
2
 test)  
would suggest  the rejection  of  the LOP  for  the price  pair  in question.  
Johansen's method was also used in to  test the stationarity  of sawnwood price variables  
(Hänninen  1998  a) and the weak exogeneity  of newsprint  price  variables (Hänninen  et  ai. 
1997). In testing  stationarity,  Johansen's hypothesis  (5.5)  was  used. The H2 design matrices 
to restrict  the p, ;  coefficients were  of  the form (1,0,0,0),  (0,1,0,0),  (0,0,1,0) and (0,0,0,1)  
under r  =l. Weak  exogeneity  (hypothesis  5.6) of the supplier  countries' newsprint  prices  in 
the UK  market  were  tested so  as  to  have an indication whether any  of  the three prices  can 
be taken as a driving force behind the others The test  was  done by  restricting the row  of  the  
a,i loadings  of the respective  variable as  (0,0,0). 
The exchange  rate  PT estimation for sawnwood (Hänninen  1998b),  newsprint and pulp  
(Hänninen  and  Toppinen  1998)  was  based on  a  markup price  equation  (3.5).  For  sawnwood 
the model also  included an export  demand equation  (3.4).  In the case  that the data is  
cointegrated,  it is  possible  to identify  cointegration  vectors,  the coefficients of which 
describe the long  run  equilibrium  relations (3.4)  and (3.5).  The starting  point  in the 
estimation was  again  to  test  cointegration  rank  using  model (5.2)  with three seasonals  and  a  
constant, the data  vectors  (x,)  consisting  of  the following  variables: 
where  x/ is  Finnish quantity,  p{  is  Finnish  price  of  exports  (FIM),  er is  exchange  rate  
(FIM/GBP, FIM/DEM),  ct is  Finnish  unit cost, p„ is  competitors'  price  (in  GBP or  in 
DEM)  and  x„  is  competitors'  quantity.  The results  of  the rank  test  showed that r-  2  for  
sawnwood indicating  two  cointegration  vectors  in the  data with coefficients (3,y.  Let 
(5.8) p,v  =  -  (3  ji  (i'=1,...,4  for  sawnwood and  i  = for  newsprint,  i*j)  
sawnwood: x,=[ xf,  pf ,  er,  cr ,  p„,  x„  ], 
newsprint  and pulp: x,=[pf,  er,  cf ,  p„, ], 
41 
denote the two  vectors.  For  newsprint  and  pulp  the result  of the rank  test was  r  = 1 
indicating  one  cointegration  vector  that  can  be  desribed by  P2  of  the above system  (5.9).  
Next,  Py and p2  were  tested by  restricting  their coefficients  to  identify  the model for 
sawnwood (relations  3.4 and  3.5),  and  newsprint and pulp  (relation  3.5).  For  this  pupose, 
Johansen's hypothesis  (5.4)  was  applied.  For  sawnwood (Hänninen  1998b),  the following  
restrictions  were  first tested simultaneously  for the two cointegration  vectors  normalized 
with respect  to  P//(x/)  and P22(^l).  
Restrictions =oin  P;  and p /2  =  0  and p62 =oin  p  2  mean  that c/  is excluded in P;  and  xf 
and x„  in p 2.  In the demand relation  (3.4)  (pj  -p„  -  er)  implies  that the coefficients of  p„ and 
er  are  equal  (i.e.  P57  =  P_?/)  and that  demand should be  homogenous  of degree  zero in pj, p„  
,
 and  er  (i.e.  Pj/ =  -  p 2/). The  linear  homogeneity  of  the price  relation (3.5)  in the nominal 
variables pf,  c/and  (p„  + er)  implies  that  the  coefficients of  p„  and  er should be  equal  (p J2  = 
p.i  2)  and  that the coefficient of  c/  should equal  the difference between the coefficients of pf 
and  er  (p«  =  -  (P22  +P.i2  ))•  The model for  sawnwood was  rejected  with these restrictions  by  
X  2  test  at  5  percent  significance  level and  testing  was  continued by  relaxing  homogeneity  in 
prices  in P; .  The second  stage testable  restrictions  were  
Structure (5.11)  was  not  rejected  by  the %
2
 test  and  the  process  was continued by  testing  
whether markup can  be  restricted to  unity  in p2:  
(5.9) Pi:[P//(x/)  p 2y(p/)  P  3,{er) p,y( C/ )  psy(p„) P„(*,)]  
P  2:  [Pi2(x/)  P22(Pf) P  .12(c) P«(c/)  Pj  2(/?«)  Pö2(*»)]  
(5.10) P/: Pyy —1  j  P<//  —O, P.?y  -  -  P2y  ,  P  5-  P.;/,  P«y  -" 1, 
P  2  :  (3/2  =O,  P  22  =l,  P<s2  =O,P«  =  "  (P  22  +  P.)  2  ),  Pj  2  =  P.!  2
(5.11) P/-. P.( ;  *  -  p 2/,  others  as  in (5.10)  
p  2: restrictions as in (5.10)  
42 
In the study  of  Finnish newsprint  and pulp  exports  only  the price relation was  estimated 
(Hänninen  and Toppinen  1998)  and the restrictions were  tested in two  stages using the  
formulations (5.10)  and  (5.12)  for  p2 .  
(5.12) p; : restrictions  as  in (5.11)  
P2: P«  =  -  (p22+P.?2)  =-1  others  as  in (5.10)  
43 
6. RESULTS  
6.1  Price  elasticities  of substitution  between Finland and the other  supplier  
countries  in  the demand of  the United Kingdom  sawnwood imports  
(Hänninen,  R.  1994,  Scandinavian Forest  Economics  35:204-217.)  
The purpose of the study  was  to  estimate own-price and substitution elasticities between 
Finnish,  Swedish,  Canadian and  Russian  sawnwoods. Armington's  export demand model 
(Armington  1969)  commonly  applied  in earlier  studies of  trade flows was  generalized  by  
assuming  variable elasticity  of substitution between supplier  countries. Thus,  a 
translogarithmic  (Christensen  et a/.,  1973) cost share equation  system imposing  
homogeneity  and symmetry  and using yearly  observations between 1961-1990 was  
estimated. The results  indicate that  the UK demand for Finnish and  Canadian sawnwood 
has been price  elastic.  The own-price  elasticities were -1.62 for Finland and -1.58 for 
Canada. The elasticity  for Sweden was -0.62 and for Russia about zero. The  results for 
Finland are  in line  with earlier estimates (e.g.,  Hänninen 1986,  Tervo et ai 1988). 
The  Morishima-elasticities of substitution indicate that the supplier  countries' sawnwoods 
are  substitutes in British sawnwood  imports.  The estimated substitution elasticities between 
Finnish, Swedish and Canadian sawnwood vary between 1.70 and 2.38 depending  on the 
country  pair.  The fact  that the short  run  dynamics  could not  be  taken into account  in the 
estimation,  however, raises  some cause  for concern.  
44 
6.2  The law of  one  price  in United Kingdom  soft sawnwood imports  -  a 
cointegration  approach  
(Hänninen,  R.H. 1998  a, Forest  Science 44(1):  17-23.) 
In this  study,  the law of  one  price  (LOP)  was  tested for  imports  of soft  sawnwood into  the 
United Kingdom  from Finland,  Sweden, Canada and Russia using  the concept of 
cointegration.  The  data are  quarterly  import  unit values from the supplier  countries and 
cover  the period 1978-1992. The  study  used the multivariate cointegration  method of  
Johansen (1988,  Johansen and Juselius 1990 and 1992)  instead of  the bivariate and single 
equation  methods often used in earlier studies. 
The existence of the LOP  was initially  tested simultaneously  for all four price  series.  
Because only  one cointegration  vector  was found in  the data, the LOP  was also tested 
separately  for pairs  of prices  under cointegration.  The results  do not  support the LOP. 
Instead, they  indicate the existence  of  differences over  the long  run  between the suppliers'  
sawnwood prices.  The rejection  of the LOP  differs from the traditional assumption  that 
commodity  prices  are  perfectly arbitraged,  with instantaneous exchange-rate  pass-through,  
at least in  the long  run.  It also differs  from the results  of Buongiorno  and Uusivuori (1992), 
who tested the law  for US paper product  exports  across  destinations. However, the results 
are in  accordance, for example,  with Isard (1977),  who compared  prices  of different 
commodities,  including  paper products,  across  exporters  (USA  and Germany).  The results  
also  raise an  important  issue  concerning  the assumptions  to  be  made regarding  competion  in 
sawnwood markets. In studies of forest products  markets it would be useful to statistically  
test the type of  competition  before  modeling.  
45 
6.3  Testing arbitrage  in  newsprint  imports  to  United Kingdom and Germany  
(Hänninen,  R.,  Toppinen,  A. and Ruuska,  P,  1997. Canadian Journal of Forest Research 
27:1946-1952.)  
The long-run  law  of  one  price  (LOP)  was  tested for  British and German newsprint  imports  
from Finland, Sweden and Canada by  using  the theory  of cointegration. Johansen's 
multivariate  cointegration  method (Johansen  1988,  Johansen and Juselius  1990 and 1992)  
was  applied  in the empirical  analysis.  The data consisted of  quarterly  import  unit values 
from the supplier  countries for  the period  1980-94. One  cointegration  vector  was  found in 
the UK data, while the prices  were not  cointegrated  in Germany.  This indicates that the 
simultaneous LOP  is not  valid in  either market. In the UK LOP  was also tested for price 
pairs.  The results indicated the rejection  of the law except  for a pair  of Swedish and 
Canadian prices.  Differences in the results  of  the two destination markets may  be due to fact  
that Germany  is  less  dependent  on the imported  newsprint  than  the UK.  Therefore,  the price  
of domestic newsprint  production  in Germany,  which was  not  included, may affect the price  
structure  of the German market. 
In the UK  market,  the Swedish and Finnish  newsprint  prices  were found to  be weakly  
exogenous, but  this was  not  true  for  the Canadian price.  This  implies  that  the producers  of 
Scandinavia would not respond  if Canadian producers  change  prices away from 
Scandinavian prices. But if the Scandinavians change  their prices,  the Canadians  would 
probably  follow. Thus,  Scandinavian producers  dominate the European  newsprint  market, 
while Canadian imports  to the area are the residual of the market demand. The LOP 
between Swedish  and  Canadian prices  and  the weak  exogeneity  of  Swedish price in the UK 
can  be interpreted  to  mean that Canadian price  is  the adjusting  part, and the Swedish  price 
is  the determining  part,  of the relationship.  The cointegration  tests of the present  study  
provide  only  weak support to  the law of  one price,  indicating  that price  differences should 
be assumed in modeling the UK and German newsprint  markets. 
46 
6.4 Exchange  rate changes  and the Finnish  sawnwood demand and price in 
the UK market 
(Hänninen,  R.H. 1998 b. Silva Fennica 32(l):61-73.)  
The  study  examines long-run  exchange  rate  pass-through  (PT)  for  the Finnish price  of 
sawnwood in the UK  market by  estimating  an  export  demand and supply  model. The data 
are  quarterly  and cover  the period  1978-1994. Johansen's (1988,  Johansen and Juselius 
1990 and 1992) cointegration  method,  that allows for the estimation and identification of  
system  models, is  applied  in the estimation. 
Two cointegration  vectors  representing  the demand and price  equations  were  found in the 
data. However,  in the export  demand relation,  homogeneity  of degree  zero in prices  was 
rejected,  indicating  that the relation does not  follow the assumptions  of  the economic 
theory.  Thus,  the coefficients  of  the estimated long-run  demand equation  are  difficult to 
interpret.  The PT elasticity  obtained from the price  equation  was  close to  unity  (0.92)  and  it 
could be  further restricted  to  equal  one. The result  is  inline with the earlier studies  that have  
estimated relatively  large  PTs  for  wood products  (e.g.,  Uusivuori and Buongiorno  1991, 
Menon 1993).  The exchange  rate elasticity  of  export  quantity,  estimated from the demand 
equation,  was  also  relatively  large  (6.56). This  result  does not  agree with  earlier studies,  
which indicate  that the trade flows  of  forest  products  are  not  very  sensitive  to exchange  rate  
variations (Buongiorno  et al. 1988 and  Uusivuori and Buongiorno  1990). The previous  
results,  however, only concern  exports  to the US market. The results  of the  present  study  
indicate that exchange  rate  changes  have had an important  effect on Finnish  sawnwood 
export prices, measured in foreign  currency,  and on  sawnwood export  quantities.  A large  
PT also  indicates imperfect  competition  in the export  market. 
47 
6.5 Long-run  price  effects  of  exchange  rate changes  in Finnish  pulp  and 
paper exports  
(Hänninen,  R.,  and Toppinen A. 1998. Forthcoming  in Applied  Economics)  
The effects  of exchange-rate  changes  on  Finnish prices  of newsprint  and pulp  exports  were  
examined in British and  German markets. The pass-though  (PT) elasticities were  estimated 
from a markup price equation using  Johansen's cointegration  method (Johansen  1988,  
Johansen and  Juselius 1990 and 1992) and  applying  quarterly  data from 1980 -1994. 
The results  of  the study  suggest  that devaluations have boosted Finnish exports  of  paper  to  
the British and German markets.  According  to  the magnitudes  of  the  PT elasticities,  pass  
through  of  exchange  rate changes  to foreign currency  export  prices  has  been incomplete.  
Previous studies  have also estimated incomplete  PT for  paper industry  (e.g., Vesala 1992,  
Uusivuori  and Buongiorno  1991,  Alavapati  et al. 1997).  According  to  the results  of the 
present study,  the PT  elasticity  for newsprint  was  0.60 in the UK  and 0.46 in Germany.  
Thus,  about one half  of the changes  in the exchange  rates  has  passed  through  to Finnish 
foreign  currency  export  prices (GBP  and DEM) and about one half  to  Finnish  home  
currency export prices  (FIM). This indicates that Finnish paper exporters  appear to have 
pursued  a  midway  pricing  strategy  aimed at maintaining market shares  and profitability  as  
the exchange  rate  changes.  In pulp exports,  the PT elasticity  was  small in the UK  market 
(0.07)  as  compared  to in Germany  (0.68),  implying  that Finnish pulp  exporters'  pricing  
behaviour has  been more clearly  destination specific  as  compared  to that of newsprint  
exporters.  
48 
7. CONCLUSIONS  
The purpose of  the present  study  was  to  produce  new  information about the functioning  of 
forest  product  export  markets.  The empirical  analysis  focused on the price  responsiveness  
of the demand for Finland's and competitor's  forest products  in the UK and Germany.  The 
examination of  price  substitution between supplier  countries' sawnwood in the UK  market 
was  based on  the yearly  data from 1961 to 1990  (Hänninen  1994).  The tests for  the law of 
one  price  (Hänninen  1998  a, Hänninen et ai  1997)  and the analyses  of  the exchange  rate  
pass-through  (Hänninen  1998b, Hänninen and Toppinen  1998)  used quarterly  data from 
1978 or  1980 to the early  1990'5. 
The  implication,  of the results  is  evidence for  the existence  of imperfect,  rather than perfect,  
competition  between supplier  countries in sawnwood and newsprint  markets.  The results  
also  elaborate the important  effects that exchange  rate changes  have had on Finnish forest 
product  export  pricing.  The elasticities  estimated  from the econometric model indicate  that 
substitution between competitors'  sawnwoods in British imports  were  not  very  high.  In 
perfect  competition,  substitution elasticities  would be infinite. The failure of the LOP 
assumption  suggests price differences between  supplier  countries in British  sawnwood 
market and  in British  and  German newsprint  markets.  Exchange  rate  pass-through  (PT)  was  
large  in Finnish  sawnwood exports  and  relatively  large  in newsprint  and pulp  exports.  Only  
in British  market  was  the PT  to  pulp  export  price  low. In  perfect  competition,  PT  elasticity  
would be  close to, or  equal  to, zero.  
The rejection  of  the law of  one price  (LOP)  differs from the traditional assumption  that 
commodity  prices  are perfectly  arbitraged.  Newsprint  and sawnwood are relatively 
homogenous  product  groups by  product  composition,  so  arbitrage  could be expected  to 
hold. In  reality,  there may  be several  reasons  for  the deviations from the LOP.  For  example,  
deviations may  occur  if competition  is  not  perfect,  if products  are  differentiated and if  the 
price  series describing  the good  differ by  their commodity  compositions  in different 
countries (Officer  1986).  In addition, the costs of  arbitrage  may be high  in sawnwood and 
newsprint  markets,  with established customer  relationships  playing  an important  role.  
49 
The large  exchange  rate  pass-through  (PT) for Finnish  sawnwood  export  price indicates 
that the exporters  have adjusted  their  prices  measured in foreign  currency  by  almost the full 
change  in the exchange  rate.  In newsprint  exports,  currency  changes  are  partly reflected in 
Finnish export  prices  in foreign  currency  and partly  in the markup  over  production  costs. 
Thus,  the estimated PT coefficients have implications  concerning  the pricing  strategies  of 
Finnish exporters.  Sawnwood exporters  have tried to maintain their market shares but not  
necessarily  their markups as  the exchange  rate  changes.  Newsprint  exporters  appear to  have 
pursued  a  midway  pricing  strategy  aimed at maintaining  market shares  and profitability  as 
the exchange  rate  changes.  The higher  PT estimate for sawnwood than for paper  implies  
that Finnish sawnwood exporters  have more efficiently made use of depreciations  and 
devaluations of the FIM to maintain their market shares. This also indicates that the 
sawnwood industry  may have more difficulty  than the paper industry  in adjusting  to 
competition  for market  share  in the EMU (European  Monetary  Union)  environment. 
Under the current  Exchange  Rate Mechanism (ERM) of the European  Monetary  System, 
the possibility  of using  exchange  rate  changes  to improve  Finnish competitiveness  and 
market share  has more or  less  disappeared.  Even outside the EMU, the scope  for national 
exchange  rate  policy  would  be reduced for  Finland (Hetemäki  et ai. 1997).  Hence,  other 
means  should be found to  adjust  to  future price  disturbances caused by  shocks  in demand. 
To find  means  for the adjustment  is  beyond  the scope  of this  study,  but  an  obvious  way  to 
adjust  to demand fluctuations is  the adjustment  of  production  costs.  However,  the relatively  
long-term  price  contracts  that exist  in the input  markets  of Finnish industry  make rapid  
adjustment  difficult. 
Differentiation of Finnish forest industry products  via customer  orientation, established 
customer relationships,  product  development  and by  value-added products  may be equally  
important. For  example,  taking  into account the environmental issues  that have been  found 
to affect consumer behavior in Finnish export  markets  (Valtonen  et ai. 1995 and 1997)  
could form an important  advantage  in international competition.  To find efficient  ways  to  
adjust  in competition  for  market share is  most  important  in the Finnish sawnwood industry,  
where the exchange  rate  PT was  higher  than  in the paper industry.  
50 
Even though  the results  of the present study  increase information on  export  demand and 
price  formation in Finnish forest  product  markets,  various research  questions  remain to  be 
explored. The results  of the study  raise issues  concerning  the assumption  of perfect  
competition  in the modeling  of  forest  products  trade flows.  Thus, it would  be important  to  
explicitly  test  market competition  also  in other  important  export  countries (e.g.,  France,  the 
Netherlands). The scope of  this study  should also be extended to additional forest  products  
(e.g.,  plywood,  printing  and  writing papers, paperboards,  etc.)  and to  a  more disaggregated  
product  groups in order  to have a  more complete  picture  about price  formation. It  would 
also be necessary  to  study  the effects of other  currencies (e.g.,  US dollar and Swedish 
krona)  that  are  important  in invoicing  forest  products.  Finally,  a short term  analysis  
capturing  the dynamics of  export  demand and  price  formation in export  markets  could 
provide interesting  new information compared to  the long-term  elasticities presented  here. 
Studies on short-term linkages  between export markets and the domestic roundwood 
market  could also create  useful information for decision makers in the Finnish forest sector. 
51 
REFERENCES  
Andersen, S. 1993. Det europeiske  Avispapiermarkedenet  -  en struktur  ock  
utviklingsanalyse  (The  European  newsprint markets  -  a  structure  and  development  
analysis).  Doctoral thesis,  Agricultural  University  of Norway.  
Alavapati,  J.R.R.,  W.L. Adamowicz, and M.K. Luckert.  1997.  A  cointegration  analysis  of 
Canadian wood pulp  prices.  American Journal of Agricultural  Economics  79:975-896. 
Alston,  J.M., C.A.  Carter,  R. Green,  R.,  and D.  Pick.  1990. Whither Armington  trade 
models? American Journal of Agricultural  Economics  72(2):455-467.  
Ardeni,  G.P.  1989. Does the law of one price  really hold for  commodity  prices?  American 
Journal of Agricultural Economics 71:661-669. 
Armington,  P.S. 1969. A theory  of  demand for  products  distinguished  by  place  of 
production.  International Monetary  Fund Staff Papers  6:159-176. 
Athukolara, P.  1991.  Exchange  rate  pass-through:  The case  of  Korean exports  of 
manufacturers. Economic letters 35(1):79-84. 
Athukolara,  P.  and  J. Menon. 1994. Pricing  to  market behavior and  exchange  rate  pass  
through  in Japanese  exports.  Economic Journal 104  (423)  :27 1 —Bl. 
Athukolara,  P.  and J.  Menon. 1995. Exchange  rates  and  strategic  pricing:  The case  of 
Swedish machinery  exports.  Oxford Bulletin of  Economics  and Statistics  57(4):533-  
545. 
Babula,  R.A. 1987.  An Armington  model of  US  cotton  exports.  The Journal  of Agricultural  
Economics Research 39(4):  12-22. 
Banerjee,  A.,  J. Dolado, J.  Galbraith,  and  D.  Hendry.  1993. Co-integration,  error  
correction,  and  the econometric  analysis  of  non-stationary  data. Oxford  University  
Press.  329  p.  ISBN 0-19-828700-3. 
Blatner,  K.A. 1989. An approach  to  the estimation of  import price  elasticities by  supplier. 
Forest Science 35:30-41. 
The Board of Customs. Foreign  Trade 1980-1994. Helsinki. 
Booth, D.L.,  V. Kanetkar,  I. Vertinsky,  and D. Whistler. 1991. An empirical  model of 
capacity  expansion  and pricing  in an oligopoly  with barometric price  leadership:  a 
case  study  of  newsprint  industry  in North  America. Journal of Industrial Economics 
39(3):255-276.  
52 
Brooks,  D.J.,  A. Baudin,  and  P.  Schwarzbauer. 1995.  Modelling  Forest  Products  Demand,  
Supply  and  Trade (draft). UN-ECE  Timber and Forest  Discussion  paper. ETTS V 
Working  Paper  -  Draft,  ECE/TIMDP,  Geneva. 
Brännlund, R., L.  Lundberg,  and A. Wiberg.  1982. Determinants of Swedish market shares 
for pulp,  paper and sawnwood in Europe.  Swedish University  of  Agricultural  
Sciences,  Department  of  Forest  Economics,  Report  36. 28  p.  Umeä. 
Brännlund,  R.,  and K-G.  Löfgren.  1995. Cyclical  dumping  and  correlated business  cycles  in 
imperfect  markets:  empirical applications  to  the Canadian pulp  and paper industry.  
Applied  Economics 27:1081-1091. 
Buongiorno,  J., J.P. Chavas,  and  J.  Uusivuori. 1988. Exchange  rates,  Canadian lumber 
imports,  and United States prices: A time-series analysis.  Canadian Journal of Forest 
Research 18:1587-94. 
Buongiorno,  J., J-J.  Chou, and R.N. Stone. 1979. A monthly  model of  the United States 
demand for  softwood lumber imports.  Forest  Science  25:641-655. 
Buongiorno,  J., and  J.K.  Gilless.  1984.  A  model of international trade  of forest  products,  
with an  application  to  newsprint.  Journal of World Forest  Resource  Management  
1:65-80. 
Buongiorno,  J.,  and J. Uusivuori. 1992. The law of one price  in the trade of forest 
products:  Co-integration  tests  for  U.S.  exports  of  pulp and paper. Forest  Science 
38(3):  539-553. 
Byström,  S.,  and  O. Eriksson.  1991. A  test  application.  P:B2-92.  In: Modelling  the Swedish 
Forest  Sector. Lönner G. (ed.). The  Swedish University  of  Agricultural  Sciences,  
SIMS Report  20. 94 p. 
Canadian Forest  Service 1996. The State of  Canada's Forests 1995-1996. Ottawa. 110 p. 
Cassel,  G.  1921. The world's  money  problems.  E.P.  Dutton and  Co.,  New  York.  
Chou, J-J., and J. Buongiorno.  1983. United States demand for hardwood plywood  imports  
by  country  of origin.  Forest  Science  29:225-237. 
Chou,  J.-J., and  J.  Buongiorno.  1984. Demand functions for  United States  forest  products  
exports  to  the European  Economic Community.  Wood and Fiber Science  16(2):  158- 
168. 
Christensen,  L.R.,  D.W.  Jorgenson,  and L.J.  Lau.  1973.  Transcendental  logarithmic  
production frontiers. Review of Economics and Statistics  55:28-45. 
53 
CSO, Overseas  Trade  Statistics  of  the United Kingdom,  1960-1994,  London. 
Dagenais,  M. 1976. The determination of  newsprint  prices.  Canadian Journal of Economics 
9:442-461. 
Davis,  G.C.,  and  N.C.  Kruse.  1993. Consistent  estimation of  Armington  demand models. 
American Journal of  Agricultural Economics  75(3):719-723.  
Dickey,  D.A.,  and  W.A.  Fuller. 1979. Distribution of  estimates for autoregressive  time 
series  with unit root.  Journal of American Statistical Association 74:427-431. 
Dornbusch,  R.  1987. Exchange  rates  and  prices.  American Economic Review 77(  1):92- 106. 
Doornik,  J.A., and  D.F.  Hendry.  1994.  PCFIML 8.0. Interactive econometric modelling  of 
dynamic  systems.  Institute of  Economics and Statistics,  University  of Oxford. 
International Thomson Publishing.  London. ISBN 1-850-32113-2. 
Duffy, P.  A.,  M.K. Wohlgenant,  and J.W. Richardson. 1990. The  elasticity  of  export  demand 
for U.S. cotton.  American Journal of Agricultural  Economics 72(2):468-474. 
Dykstra,  D.P.,  and M. Kallio. 1987. Introduction to the lIASA  Forest  Sector Model. P. 139- 
167 in The  Global Forest  Sector:  an  analytical  perspective,  Kallio,  M.,  D.P.  Dykstra,  
and C.S.  Binkley  (eds.).  John Wiley  and Sons,  Chichester. 
Edison,  H.J.,  J.E.  Gagnon,  and  W.R.  Melick. 1997. Understanding  the empirical  literature 
on purchasing  power parity:  the post-Bretton  Woods era. Journal of  International 
Money  and Finance 16(1): 1-17. 
Engle,  R.F.,  and C.W.J. Granger.  1987.  Co-integration  and  error correction: representation,  
estimation and  testing.  Econometrica 55:251-176. 
European  Commission,  Intra- and  extra-  EU  trade, CD-ROM, 1988-1994. Luxemburg.  
FAO  1991. Yearbook of forest  products,  1980-1991. Rome. 
FAO 1994. Yearbook of forest products, 1983-1994. Rome. 
Feenstra,  R.C.,  J.E. Gagnon,  and M.M. Knetter. 1996. Market share and exchange  rate  
pass-through  in world automobile trade. Journal of International Economics  40:187- 
207. 
Froot,  K.A.,  and P.D. Klemperer.  1989. Exchange  rate  pass-through  when market share 
matters.  The American  Economic Review 79(4):637-654.  
Fuss,  M.A. 1977. The demand for energy in Canadian manufacturing.  An  example  of  the 
estimation of  production  structures  with many inputs.  Journal of  Econometrics 5:89- 
116. 
54 
Giovannini,  A.  1988. Exchange  rates  and  traded goods  prices.  Journal of  International 
Economics 24(1/2):45-68.  
Goodwin,  B.K.  1992. Multivariate cointegration  tests and the law of  one price in 
international wheat markets.  Review  of Agricultural  Economics 14(1):  117-124. 
Goodwin,  8.K.,  and  T.J.  Grennes. 1994. Real interest rate  equalization  and  the integration  
of international financial markets.  Journal of International Money  and Finance 
13:107-124. 
Granger,  C.W.J.  1981. Some properties  of  time series  data and  their use  in econometric 
model  specifications.  Journal of  Economterics 16:121-130. 
Granger  C.W.J,  and P. Newbold. 1974. Spurious  regressions  in econometrics. Journal of 
Econometrics  2:(1 11-120). 
Halonen,  M. 1990. Vertical oligopolistic  markets  -  The Export  Markets  for  Forest  
Products. Helsinki School of Economics. Series  B:101. 
Helpman,  E.,  and P.R.  Krugman.  1985. Market Structure and  Foreign  Trade.  Increasing  
Returns,  Imperfect  Competition, and the International Economy.  The MIT Press,  
USA. 271 p. ISBN 0-262-08150-4. 
Henning,  J., and L.  Martin. 1989. An economic evaluation of  expanded  3-M wheat exports.  
Canadian Journal of  Agricultural  Economics 37(3):445-465.  
Hetemäki, L. 1990. Factor substitution in the Finnish pulp  and paper  industry.  Summary: 
Panosten substituutio  Suomen massa-ja  paperiteollisuudessa.  Acta  Forestalia  Fennica  
211. 87  p. ISBN  951-40-1093-0. 
Hetemäki,  L.1997. Metsäsektori 2010. Metsälehti kustannus/Metsäntutkimuslaitos.  
Helsinki.  153 p. 
Hetemäki, L.,  R.  Hänninen, J. Kuuluvainen, P. Ollonqvist,  P. Ruuska,  R.  Seppälä,  A. 
Toppinen,  and  J.  Uusivuori.  1997. EMU ja  Suomen metsäsektori.  Valtioneuvoston 
kanslian julkaisusarja 17.  26 p. 
Hetemäki, M.,  E-L Kaski.  1992. KESSU  IV .  An econometric  model of the Finnish  
economy. Ministry of Finance. 135  p.  ISBN 951-47-5859-5. 
Honkatukia,  J. 1994. Imperfect  competition  and export  demand. Labour Institute  for 
Economic Research. Discussion  papers 118. 67 p. 
55 
Honkatukia,  J. 1995. Intraindustry  trade,  competitiveness,  and  price  rigidities  in the Finnish  
manufacturing  industries. Government Institute of  Economic Research  (VATT).  
Discussion  papers 90. 43  p.  
Hooper,  P., and  C.L.  Mann. 1989. Exchange  rate  pass-through  in the 1980 s:  The case  
of  U.S.  imports  of  manufactures. Brookings  Papers  on  Economic Activity  
1989(l):297-337.  
Hseu,  J-S., and J. Buongiorno.  1993. Price  elasticities  of substitution  between species  in 
the demand of U.S.  softwood lumber imports  from Canada.  Canadian Journal of 
Forest Research 23:591-597. 
Hung,  W.,  Y.  Kim,  and  K.  Ohno. 1993. Pricing  exports:  a  cross-country  study.  Journal of 
International Money  and Finance 12:3-28. 
Hänninen,  R.  1986. Suomen sahatavaran  vientikysyntä  Länsi-Euroopassa  vuosina 1962- 
1983.  Summary:  Demand  for  Finnish sawnwood exports  in western  Europe,  1962- 
1983.  Folia  Forestalia 657.  25  p.  
Hänninen,  R.  1994. Price elasticities of  substitution between Finland and  the other supplier  
countries  in  the demand of the United Kingdom  sawnwood imports.  P.  204-217 in 
Proc. of the Biennial meeting  of  the Scandinavian Forest  Economics,  Helles,  F.,  and 
M. Linddal (eds.).  Scandinavian Forest  Economics 35. Denmark. 
Hänninen,  R.H. 1998 a.  The law of  one  price  in United Kingdom  soft  sawnwood imports  -  a  
cointegration  approach.  Forest  Science 44(1): 17-23. 
Hänninen,  R.H.  1998b.  Exchange  rate  changes  and  the Finnish sawnwood demand and 
price  in the UK  market. Silva Fennica 32(l):61-73.  
Hänninen,  R.,  A.  Toppinen,  and  P.  Ruuska.  1997. Testing  arbitrage in newsprint  imports  to  
United Kingdom  and  Germany.  Canadian Journal of Forest  Research  27:1946-1952. 
Hänninen,  R.,  and  A.  Toppinen  1998. Long-run  price  effects  of exchange  rate changes  in 
Finnish pulp  and paper  exports.  Forthcoming  in Applied  Economics. 
IMF,  International Financial Statistics,  1978:4-1995:4. Washington  
Ito,  S.,  D.T. Chen  D.T. and W. Peterson. 1990. Modeling  international trade flows and 
market  shares  for agricultural commodities: a  modified Armington  procedure  for 
rice.  Agricultural  Economics 4:315-333. 
Isard,  P. 1977. How far can we  push  the "law of  one price"? The American Economic 
Review 67:942-948. 
56  
Johansen,  S. 1988. Statistical analysis  of  cointegration  vectors.  Journal of  Economic 
Dynamics  and Control. 12:  231-254. 
Johansen,  S. 1995. Likelihood-based inference in cointegrated  vector  autoregressive  
models. Oxford University  Press  Inc.,  New York.  267  p.  
Johansen, S.,  and K. Juselius. 1990. Maximum likelihood estimation and inference on  
cointegration  -  with applications  to  the demand for  money. Oxford  Bulletin  of  
Economics and Statistics  52: 169-210. 
Johansen, S.,  and  K.  Juselius. 1992. Structural tests  in a  multivariate cointegration  analysis  
of the PPP and the UIP for UK. Journal of Econometrics 53: 211-244. 
Johansen,  S.,  and Juselius,  K. 1994. Identification of  the long-run  and the short-run 
structure. An application  to the  ISLM model. Journal of  Econometrics 63(7):7-36. 
Jung,  C.,  and  K.  Doroodian. 1994.  The law of  one  price for  U.S.  softwood lumber: a  
multivariate cointegration  test. Forest  Science 40:595-600. 
Knetter,  M.M. 1989.  Price  discrimination by  U.S.  and  German exporters.  The American 
Economic Review,  79(1): 198-210. 
Knetter,  M.M. 1993. International comparisons  of pricing  to  market behavior. The 
American Economic Review,  83(3):473-486.  
Kongsted,  H.C. 1996. Modeling  price and  quantity  relations  for  Danish manufacturing  
exports.  In: "Dynamic  models of  foreign  trade under fluctuating  exchange  rates:  
Theoretical  and  Empirical  Applications.  Ph.D.  Thesis  in Economics.  University  of 
Copenhagen.  Rod Serie  (39)101-169.  
Krugman,  P. 1987. Pricing  to  market when the exchange  rate  changes.  P. 49-70 iti Real 
financial linkages  among open economies,  Arndt,  S.W. ,  J.A. Frenkel,  and  J.D. 
Richardson (eds.)  Cambridge:  MA Mit Press.  
Krugman,  P.  R.,  and  M. Obstfeld  1997. International Economics.  Theory  and  Policy.  
Addison-Wesley,  Massachusetts.  766 p. 
Laaksonen,  S.,  A.  Toppinen,  R.  Hänninen,  and J.  Kuuluvainen. 1997. Cointegration  in  
Finnish paper exports  to  the United Kingdom.  Journal of Forest  Economics  3(2):  
171-185. 
Laaksonen,  S.  1998.  Price  adjustment  for  forest  products  under fixed and  floating exchage  
rate  regimes.  College  of  Natural Resources,  University  of  California,  Berkeley.  
Unpublished  dissertation. 128 p. 
57 
Martinello, F.  1985. Factor substitution,  technical change  and  returns  to  scale  in Canadian 
forest industries. Canadian Journal of Forest Research 15:1116-1124. 
Menon, J. 1993. Exchange  rate  pass-through  elasticities  for the Monash  model: A 
disaggregate  analysis  of Australian manufactured imports.  Center of  Policy  Studies 
Monash University.  Working  Paper  No.  OP  -76 July.  49  p.  
Menon,  J. 1995. Exchange  rate  pass-through.  Journal of  Economic Surveys  9(2):  197-231. 
Miikkulainen,  P.  1989.  Suomen palvelujen  ulkomaankauppa.  Suomen Pankki  (Bank  of 
Finland Publications)  D:7l, 120 p. 
Naug, 8.,  and  Nymoen,  R.  1996. Pricing  to  market in a  small open  economy.  Scandinavian 
Journal of  Economics 98(3):329-350.  
OECD,  Main Economic Indicators. Historical Statistics  1962-1991, 1993, 1994, Paris. 
Officer,  L.H. 1986. The law of one price  cannot  be rejected:  Two tests based on the 
tradable/nontradable price ratio. Journal of  Macroeconomics  8(2):  159-182. 
Okun,  A. 1975.  Inflation: Its  mechanics  and welfare  costs.  Brookings  Papers  on  Economic 
Activity  2:351-401. 
Pick, D.H.,  and  T.A. Park.  1991. The competitive  structure  of  U.S. agricultural  exports.  
American Journal of  Agricultural  Economics 73(  1): 133-141.  
Pindyck,  R.S.  and D.L. Rubinfeld. 1988. Econometric models and economic forecasts. 
Second edition. McGraw-Hill Inc.  630  p. ISBN 0-07-050096-7. 
Puttock,  G.D., and D.M. Prescott. 1992. Factor substitution and economies of  scale in the 
southern Ontario hardwood sawmilling  industry.  Canadian Journal of  Forest Research 
22:1139-1146. 
Richardson,  J.D. 1978. Some empirical  evidence on  commodity  arbitrage  and the  law of  one 
price.  Journal of International Economics 8(2):34 1-51.  
Ripatti,  A. 1990. Raakapuun  kysyntä  ja tarjonta  Suomessa.  Empiirinen  tutkimus pitkän  
aikavälin riippuvuuksista.  M.Sc. thesis in Economics,  University  of  Helsinki. 93 p. 
Rogoff,  K.  1996.  The purchasing  power parity  puzzle.  Journal  of  Economic Literature 
34:647-668. 
Ronnila,  M.  1995.  Medium-term scenarios  for  the Finnish pulp  and  paper industry.  
Licentiate thesis in Management  Science,  Helsinki School of  Economics,  
Department  of  Economics. 104 p. 
58 
Sarker  R.  1996. Canadian softwood lumber exports  to  the United States:  A  cointegrated  
and error-corrected  system.  Journal of  Forest  Economics 2(3):205-231.  
Sarris,  A. 1983. European  community  enlargement  and  world trade in fruits  and  vegetables.  
American Journal of  Agricultural  Economics  65:235-246. 
Statistical Yearbook  of Forestry  1997. Yrjö Sevola (ed.)  The  Finnish Forest  Research 
Institute. Jyväskylä  1997. 348 p.  ISSN 0359-968 X 
Statistisches  Bundesamt,  Aussenhandel Nach Waren und  Ländern,  1980-94,  Wiesbaden. 
Statistics  Finland, Bulletin of  Statistics  (SVT).  1978-1995. Helsinki. 
Sukselainen,  T. 1986. Hinnanmuodostus Suomen teollisuudessa vuosina 1969-1981. 
Summary:  Price formation in the Finnish manufacturing  industry  in 1969-1981. 
Bank of Finland 8:42. 399 p. 
The  BOF4 Quartely  Model of  the Finnish Economy.  1990. Bank  of  Finland,  series  D:3. 327. 
ISBN 951-686-248-9. 
Tervo,  M.,  J.  Mäkelä,  and R.  Hänninen. 1988. Dynaaminen  kysyntämalli  Ison-Britannian 
maittaiselle sahatavaran tuonnille. Abstract:  A dynamic  demand model for British 
imports  of  sawnwood from different countries. Metsäntutkimuslaitoksen tiedonantoja  
(Finnish  Forest  Research Institute,  Research Papers)  313. 35 p. 
Toppinen,  A. 1998. Incorporating  cointegration  relations in a  short-run model of  the Finnish 
sawlog  market. Canadian Journal of  Forest  Research 28:291-298. 
Toppinen,  A.,  and  J. Kuuluvainen. 1997. Structural changes  in sawlog  and  pulpwood  
markets in Finland. Scandinavian Journal of Forest  Research 12(4):382-389. 
Uusivuori,  J,,  and  J.  Buongiorno.  1990. Short- and  long-run  effects  of  changes  in  exchange  
rates  on  United States  imports  of  paper from Finland and  Sweden. Scandinavian 
Journal of  Forest  research  5(2):285-296.  
Uusivuori,  J., and  J. Buongiorno.  1991. Pass-through  of  exchange  rates  on  prices  of  forest  
product  exports  from the United States  to  Europe  and  Japan.  Forest  Science  
37(3):93 1-948. 
Valtonen, K.,  H. Juslin,  and P. Laine. 1995. Metsäteollisuuden vihreät markkinat. 
Metsäntutkimuslaitoksen tiedonantoja  578.  Helsinki .ISBN 951-40-1482-0. 
Valtonen,  K.,  H.  Juslin, and H. Meriluoto, H. 1997. Metsäteollisuuden vihreät kuluttajat.  
Metsäntutkimuslaitoksen tiedonantoja  658. 158 p. Helsinki.  
59 
Vesala,  J. 1992. Incomplete  exchange  rate  pass-through  and hysteresis  in trade. A  survey  of 
recent  theories and  empirical  study  of export  pricing  of Finnish paper manufactures. 
Bank  of Finland  Discussion  Papers  29/92. 76  p.  
Volk, R.  1983. A model for  Finnish  exports  of  printing  and writing paper to the United 
Kingdom  and  Germany. Pellervo Economic Research  Institute,  PTT. Reports  and 
Discussion  Papers  29.  109  p. 
Wiberg,  A. 1987. Svensk  massa-  och  pappersindustris  marknadsställning.  Swedish  
University  of  Agricultural  Sciences,  Department  of  Forest Economics,  report  71. 
Umeä. ISBN9 I-576-3171-9. 19p. 
Yang, J. 1997. Exchange  rate  pass-through  in U.S. manufacturing  industries. The Review 
of Economics  and Statistics  79(1):95-104.  
60 
APPENDIX  1. Augmented  Dickey-Fuller  (ADF)  unit root  tests  for  the data of  the  study  of  
Hänninen (1998b)  with four  lags  (k  =4)  including  constant  and seasonals. 
Notes: * denotes the rejection  of  the null  hypothesis  of  nonstationarity.  The critical t -value  
is -2.90 at the 5%  level. 
Symbols:  X/=  quantity  of  Finland's sawnwood  exports  to  the UK,  1000 m\ ps-  Finnish  unit  
price of  exports  in the UK  (FIM/m
3
), er  =  exchange  rate  (FIM/GBP),  c/=  Finnish unit cost  
(index),  p„=  competitor's  price (GBP/m
3
)  and  x„  = the competitor's  export  quantity. 
Variables Test for  levels  Test for  first differences 
Finnish quantity,  x/  -2.39 -3.02* 
Finnish  price of  exports,  p/  -1.58 -3.68* 
Exchange  rate, er -1.83 -3.77* 
Finnish  unit cost, cj -2.68 -3.69* 
Competitors'  price,  p„ -0.09 -2.80 
Competitors'  quantity,  x„ -1.42 -2.66 
I  

204 
PRICE ELASTICITIES  OF SUBSTITUTION 
between  Finland  and  the  other  suppliers  in  the demand  
of the United  Kingdom  sawnwood  imports  
Riitta Hänninen 
The Finnish Forest Research Institute 
The Department  of  Forest  Resources 
Unioninkatu 40 A, SF-00170  Helsinki,  Finland 
SUMMARY 
The purpose of  the paper is  to  estimate substitution and own-price elasticities between 
Finland and its  main competitors  in the United Kingdom  soft  sawnwood  market. The 
model consists  of a cost minimizing  aggregate demand for sawnwood imports  to the 
United Kingdom  and of share  equations  for each  supplier  country.  The resulting  
estimates are  long term elasticities.  The  results  showed, that  the demand for  Finnish and 
Canadian sawnwood was  elastic  with  respect  to  the price,  while demand for Swedish 
sawnwood was  inelastic and the own-price  elasticity  for Russian sawnwood was  near  zero. 
The Morishima elasticities of  substitution referred  to substitutability  between supplier  
countries' sawnwood in the market.  The magnitude  of  the elasticities indicated that it  has  
been easier to  substitute competitors'  sawnwood for  Finnish sawnwood than  vice versa.  
Keywords:  Sawnwood,  imports,  two-level  demand,  Morishima elasticity  of  substitution, 
translog  function 
205 
1. AIM AND APPROACH 
The sawnwood industry  accounts  for more than half the value of timber and wood 
product  exports  of Finland. The main part  of  its production  is  exported,  the traditional 
customer  country being  the United Kingdom  that accounts  for 20 per cent  of  Finland's  
sawnwood exports.  So, the United Kingdom  is still an important  market  area  for  
Finland's sawnwood, even though  its share of total sawnwood exports  has declined 
considerably  during the last 30 years. Also Finland's market share and  export  quantities 
to the United Kingdom  has declined. There are probably  several reasons  for  this  
declining  development. It is  assumed here,  that the reason  is  Finland's  weakened price 
competitiveness  in the market. 
The price  competitiveness  can be measured by substitution elasticity.  It tells,  how easily  
a certain commodity  can be replaced  by other  commodities.  Here, it enables us to 
ascertain,  how the  mix of  sawnwood imported  from different supplier  countries might 
respond  to  a  rise  in one  supplier's  price.  Then,  the magnitude  of  the elasticity  shows  how  
much e.g. a rise  in Finland's export  price  with respect  to its  competitor's  price  reduces  
its export  quantity, ceteris  paribus,  in a given export market. The price  elasticity  
estimates of  sawnwood import  demand can be applied  in marketing planning.  If 
information is  available on the  absolute values of  the price  elasticities,  it is  possible,  for 
example,  to  select  a suitable price  policy  for different customer countries so as  to 
control export earnings.  
The principal  aim of the paper is to  estimate substitution elasticities for sawnwood 
imports  from the main supplier  countries,  Finland,  Sweden,  Canada and Russia  to the 
United Kingdom.  Also own-price  elasticities for imports  from these countries are 
estimated. Because  price  substitution can  be  measured only  if  sawnwood  of  each  supplier  
country has  a distinct price,  it is assumed that  there are price differences between 
suppliers  in the United Kingdom  market. In other words,  sawnwood  from different 
countries of  origin  are  assumed  to be imperfect  substitutes. 
Research on  substitution elasticities of  forest products  in foreign  trade have been carried 
out especially  in the United States. For example,  Chou & Buongiorno  (1982) studied 
substitution between imported  and domestic plywood  and Chou  & Buongiorno  (1983)  
studied substitution in  plywood imports  to the U.S. between certain Asian countries. 
Chou  & Buongiorno  (1984)  estimated cross  price  elasticities  between forest products  
from the U.S.  and from the rest  of  the world in the EEC imports.  Further, Hseu & 
Buongiorno  (1993)  analysed  substitution between species  in the demand of  U.S.  softwood 
lumber imports  from Canada. From Finnish studies we mention e.g. Hänninen (1986)  
who  estimated substitution elasticities between sawnwood from Finland and the rest  of 
the world in West European  countries  and Tervo  et ai. (1988),  who used the same 
approach  in studying  substitution in the United Kingdom  sawnwood imports.  
In the  earlier studies,  substitution in foreign  trade is  most often measured by  a cross  
price elasticity  that has been estimated from a model of a constant  elasticity  of  
substitution. The present paper is based on  the research by Hänninen (1993),  that 
analysed  substitution by using  a more general  functional form that doesn't include this  
assumption.  The  data of  the study  consist  of  annual time series from the years  1961-1990. 
206 
The properties  of  the time-series are  analysed  in order  to  make  more  reliable conclusions 
about the elasticity  estimates.  This is  not  usually  done in the previous  studies either. 
Substitution between Finland's and its  competitors'  sawnwood is measured by using  the 
Morishima elasticities of  substitution. Blackorby  &  Russel  (1989,  p.  883)  have shown, that 
Morishima elasticity  is a better measure than, for example,  Allen partial  elasticity  of 
substitution when substitution effects are  analysed  in many inputs'  case.  
2. THEORETICAL MODEL  FOR  SAWNWOOD  DEMAND 
In order to estimate  substitution and own-price  elasticities,  a  model for sawnwood import 
demand to the United Kingdom  is formulated. The  model is based on the two-level 
function approach  presented by  Fuss  (1977).  The two-level functions are  originally  
discussed in connection with production  theory  and applied  in the studies of  production  
structures and factor demand. To foreign trade this approach  was applied  first by  
Armington  (1969),  who used a two-level utility function. The two  stage model has  the 
advantage  that it saves  degrees  of  freedom in the estimation,  i.e. it is  a  way  to decrease 
the number of  explanatory  variables  in the estimable equations.  
The two-level demand model of  sawnwood implies  a  two  stage optimization  process  for 
a representative  firm in the end use  sector of sawnwood. It is  assumed, that a 
representative  producer  decides first, how much sawnwood and other aggregate inputs  
are needed to produce  a certain output quantity. After this decision,  the producer  
chooses between sawnwood  from different countries of  origin.  In the present paper, only  
the latter stage of  the model is  specified  and estimated. 
Let us  assume  that a  production  function of  a  representative  firm using  sawnwood as  an 
input  can be presented  as  follows: 
The  production function (1)  is  assumed to be well behaved,  i.e. it is assumed that the 
regularity  conditions presented  by Diewert (1971,  pp. 484-485)  hold. Applying  the 
assumption  of  weak homothetic separability,  the production function is written in the 
form: 
Where  M = M(Mt ,...M N) is  the aggregate production  function for sawnwood using  
sawnwood from different countries of origin as arguments. The production  function 
exhibits  constant  returns  to scale with respect  to Mj,...Mn,  and O. So, it is  possible  to 
(1)  Q = f(M,O) 
(2)  Q = O) 
where: 
Q = gross output 
M = aggregate quantity  of  sawnwood input 
0 = all other aggregate inputs,  e.g. substitutes of  sawnwood, capital, labour etc. 
207 
regard M as the output of a production process  that  uses sawnwood from different 
countries to "produce"  the aggregate quantity  of  sawnwood, M,  needed in the production  
of Q. 
Because the production  function is assumed to be weakly  separable,  the marginal rate  
of  substitution between  sawnwood from different countries, and M
jlt
 is  independent  
of the quantities of the other inputs  O demanded. It is  further assumed that the 
sawnwood aggregate (subfunction  M) is homothetic in its  components. This means that 
the sawnwood mix from different supplying  countries is  independent  of  the  total quantity 
of sawnwood demanded. 
If factor prices  and output levels are  exogenously  given  and cost-minimizing  behaviour 
is valid, the Shephard  duality theorem states that there exists  a cost  function dual to 
every production  function. Thus, the production  technology  implied  by  (2)  can  be 
represented  by the following  cost function,  which is also homothetic and weakly  
separable:  
Under constant  returns  to scale of  the production  function and linear homogeneity  of  the 
respective  cost  function,  the price  index of  sawnwood has the property:  
where M and M„..,MN  (j=1,..,N) are related to the sawnwood  function  M = 
as  in (2).  
This  two-stage optimization  process  should lead to  the same  result  as  straightforward  cost  
minimization,  given  prices  P M1 ,...PMN, and PO .  For  this,  the price index PM  must  depend 
only  on the prices  of  the sawnwood inputs  (e.g.  Solow, 1955-56,  p. 105): 
Since  P M  is  the price  per  unit of  sawnwood,  it is  also  the cost  per unit  to  the optimizing  
agent. It is  in fact  the minimized cost  of  total  imported sawnwood. 
The optimal  sawnwood mix  can  be derived from the above cost  function (5).  This  cost  
function is assumed to be homogenous  of degree one and it is also assumed to  be  
translogarithmic  (Christensen  et  ai. 1973).  So,  it is  presented  in the following  translog  
form: 
(3) C = h(P M(P MI ,...PMN), Po> Q), 
(4)  Pm M = (P ml M, +  PM2M2 +,..,+PmnMn) = E P„M> 
(5)  PM~  PM(PMI>">PMN) un't C. 
where: 
C = total production  cost 
Pm  = an aggregator function,  i.e. price  index  of  sawnwood 
from different countries of  origin  (j  =  l,..,N) 
Po = prices  of  the other inputs  
Q = gross output. 
208 
Assuming that the cost function (6)  is  twice differentiate and using Shephard's  duality  
theorem,  the derivative of  cost with respect  to price equals  the optimum  quantity of a  
sawnwood from a certain  supplier  country.  Cost  minimizing  behaviour implies  (Fuss  1977, 
p. 9  and e.g. Greene 1990, pp. 527-528)  that the demand functions for the suppliers, in 
terms of  shares in the cost  of  the sawnwood aggregate, take the form:  
Where Sj = PjXj/C  = the cost  share of  the ith  factor and i= sawnwood imports  from 
Finland,  Sweden, Canada,  Russia  and the rest of  the world. The respective  cost share 
equations  can be presented  as:  
The above demand system  is  assumed  to  satisfy  the adding-up  criterion 2Sj=l  and to 
impose  the properties  of  the  neoclassical production  theory:  
The estimation the equation  system  produces  parameters  for  the calculation of  own-price  
and  substitution  elasticities for  sawnwood imports from different supplier  countries. The 
robustness  of  these estimates depends  on the robustness  of the estimation results.  Thus, 
(6) C =Pw =B 0 + BF lnPF  + Bs lnPs + BclnP c + BsulnPsu + BR lnP R  
+  1/2 [BpF(lnP F)
2
 +  BssClnPs)
2
 + B
cc(lnPc)
2
 + B
susu(lnPsu)
2
 
+
2
] n p)(^nPs)  +  B FC (lnPF)(l nPc)"'"^FSu(l nPF)(l nPsu)  
+  BFR(InP F)(lnPR)+ Bsc(lnPs)(lnPc)+ Bssu(lnPs)(lnPsu)+ BSR(InPs)(lnPR ) 
+ öcsu(lnP  
c
)(lnP  su) + +  öSUR(InPsu)(lnPR ), 
(7) (ainC/ainPi)(P/C)  =  PjXi/C  = Si  = B;  +26^^, 
(8) SF  =BF  + BpplnPp  + BpjlnPg  + BFCInPc + + Bra lnPR 
S
s = B s + BSF InPF  +  BsslnPs + B sclnPc + +  BSRInPR 
S
c = Bc + BCFInPF  + +  Bcc lnPc + + BCRInPR  
S
su = Csu+  Csuf^PF+  ftsus'nPs BSUR^PR  
S
R  ~Br  4-  BRF InPF  + BRSInPs 4* BRC InPc + B lnPSLi  + BKRInPK 
(9)  Bf +  Bs +Bc + Ben H*  Br 1 
+ Bps + BpC + Breu  +  Bp»  
= 0 
B
SF  + Bcj  + Bsc + Bccti  + Bsr  =  0 
B
cf +  Bcs  + Bcc +  Bcsu  +  BCR = 0 
®SUC = 0 
BRP  +  BRS  + Brc  + Brsu  + BRR  
= 0 
and  to  satisfy  the Slutsky  symmetry  condition B;j  = ,  i+j.  
>i  = cost  shares of  the supplier  countries,  i=F,S,C,SU,R, 
'i  = sawnwood prices  of  the supplier  countries,i=F,S,C,SU,R,  
(F  = Finland,  S = Sweden, C = Canada, SU = Russia,  R  -  the rest  of  the world) 
mere: 
'
M
= unit cost  of  aggregate sawnwood imports  
',= sawnwood prices  from the supplier  countries,i=F,S,C,SU,R,  
(F = Finland,  S = Sweden, C  =  Canada,  SU = Russia,  R = the rest  of  the world) 
209 
the theoretical restrictions imposed  by the underlying  theory  must be tested. In other  
words,  the necessary  and sufficient condition for strict  quasi-concavity  of  the cost  function 
in input  prices  is  that the matrix of  the computed  elasticities (Hessian  matrix) must be 
negative  semidefinite at each observation point.  (Berndt  1991,  p.  476 and 493).  Moreover, 
the  underlying  cost  function must  be monotonically  increasing in input  prices.  This  means  
that the fitted cost shares must all be positive.  The cost  share equation  systems  must also 
satisfy  the symmetry  restrictions,  which are  tested by  using  the  Wald test (e.g.  Berndt 
1991, p. 465).  
3. THE OWN-PRICE AND SUBSTITUTION ELASTICITIES OF DEMAND  
For  the translog  cost  function,  the price  elasticities can be  presented  in the following  way.  
The own price  elasticity  of  input  demand can be written: 
and the  cross-price  elasticity  of  input demand can be presented  in the form: 
(see  e.g. Puttock & Prescott  1992,  p. 1141). In these formulae: 
The  formulae 12 and 13 are the Allen partial  elasticities. 
The Morishima elasticity  of  substitution (MES)  is  used to  measure the substitution 
effects.  It can be defined according  to e.g. Chambers (1988,  p. 96) as:  
This elasticity of substitution is  asymmetric  and the classification of  inputs  i and j as  
Morishima substitutes or  complements  depends  critically  on  which one  of  the input  prices  
changes.  According  to  14,  inputs  i  and j are  Morishima substitutes  (i.e.  i  is  substituted for 
j) if  and only  if  an increase in pj  causes  the input  ratio x/x,  to  increase  when p,  is  fixed. 
If  the estimate of  the Morishima elasticity  of  substitution is  positive,  the two inputs  in 
question can  be interpreted as  Morishima substitutes  or inputs  substitutable for each 
other.  If,  on the  contrary,  the elasticity  is negative,  the inputs  are Morishima 
complements.  Because the elasticity  estimates for the translog function normally  differ 
(10)  6H =„& 
(11)  Cjj  =ffjjSj. 
(12)  a,  = (Bjj  +  SiS,)/S,Sj  and 
(13)  a, =  (Bjj  +  S,(S,  -1))/S
2
i(  
(14)  MESjj  =  e, -ea 
where: 
Bij.6,  = estimated coefficients of  the cost  share  equations  
S, S,  = the fitted cost  shares for the supplier  countries i and j 
210 
at every  observation,  they  are  usually  computed  at the mean data values (e.g.  Greene 
1990, p.  528).  
4. DATA 
The estimation of  the model requires  time series data on quantities, prices  and cost 
shares  of  sawnwood imports  to the United  Kingdom  by  supplier  countries. The quantity  
of  aggregate sawnwood input  (M) demanded in the end use  sector  is  described by  the 
total  imports  of  coniferous sawnwood to the United Kingdom.  Changes  in the quality or  
species  composition  that may have occurred during the years 1961-1990 could not  be 
taken into account  because of a lack of information. 
Sawnwood demanded from different countries  of  origin  are  described by the imports  of 
coniferous sawnwood from Finland (F),  Sweden (S),  Canada (C), Russia  (SU)  and the 
rest  of  the countries  (R). The quantities (cbm)  and  unit prices  (f/cbm) of imported  
sawnwood are  obtained from the Overseas  Trade Statistics  of  the United Kingdom.  The 
unit prices  are  based on the CIF (cost,  incurance and freight) values. The series  were 
also checked and compared  with other sources  (FAO:  Yearbook of Forest Products,  
Finntimber: Statistical Yearbook, TTF: U.K. Yearbook, and Finland's official statistics  
of  foreign  trade)  and corrected  where necessary.  The import  quantities  from the rest  of 
the countries (R) are calculated as the residual MR =  M-(MF +MS+MC+MSU). The 
respective  import  price  PR is  calculated as  a  weighted  (weighted  by  cost  shares)  average 
of  the countries included in the  residual group. 
5. ESTIMATION METHOD AND TIME SERIES PROPERTIES 
The study  estimates the model by  using only  the cost share equations  (8). In order to 
specify a stochastic  framework for estimation,  a random disturbance term is  added to 
each cost  share equation.  The restriction of linear homogeneity  is  satisfied by  dividing 
the price  variables of  the cost  share equations  by  the prices  of the residual countries 
(P„). Before  the estimation,  the time series  properties  of  the cost  share  variables (Sj)  and 
transformed price  variables (P/P R ) were analysed.  
From the viewpoint of  the statistical validity  of  the estimation results,  it is  important to 
ensure  that the time series are  stationary  i.e. 1(0) processes.  A time series is  said to be 
(weakly)  stationary  if  its  first two moments, i.e.  mean and  variance,  are  constant.  Many 
economic  time series do not  satisfy  the stationarity  properties,  they  are  often 1(1) 
processes.  
The properties  of a  stationary  series 1(0) and a nonstationary  series 1(1) are quite 
different. Thus,  the degree  of  integration  of  the dependent  and independent  variables 
must match in the equations.  If both are 1(0) processes,  unbiased parameter estimates 
211 
can be obtained and the interpretation  of  the t-values is  standard. However, if both are 
1(1) processes  and the error  term is  1(0), unbiased estimates can be obtained, but the t  
values of the parameter estimates  do not  have standard interpretations.  In this case,  the 
dependent  (y,)  and independent  variables (z,)  are said to be  integrated.  If  y,  is  1(0) and 
some of the dependent  variables  are  1(1), results  are  not  reliable,  unless the subset of 
variables z,  cointegrates  to 1(0).  
In order to make conclusions concerning  the robustness  of the estimation results,  the 
properties  of the time series were examined by using autocorrelation functions,  
autoregressive  processes  of  the series and normality  and  stationarity  tests. Summing up 
the test results,  the stationarity  of  the variables required  by  the estimation method is  not  
fulfilled in the model, the variables seem  to be 1(1) series (Hänninen,  1993). 
The demand system  (8)  is  first estimated equation  by  equation  using  the ordinary  least  
squares method (OLS).  Then the symmetry  restrictions  are  imposed,  and the systems  are 
estimated by  using  Zellner's seemingly  unrelated estimator (ZEF)  (Pindyck  &  Rubinfeld, 
1988,  pp.  331-334).  Because  the cost  shares sum up to unity  at each observation over  all 
equations,  only  four of  the five share  equations  are  linearly  independent.  The common 
procedure  to  handle singular  systems  is  to  drop an arbitrary  equation  and then estimate 
the remaining  share equations  by using a method that  is invariant to  the excluded  
equation,  (e.g.  Berndt, 1991,  pp.  472-473).  However, no information is  lost,  because the 
parameters of  the excluded equation  can be calculated by using  the remaining equations  
and the parameter restrictions (9)  (see  e.g. Berndt, 1991, pp.  472-473).  
6. ESTIMATION RESULTS 
The Zellner iterative system estimation method with  symmetry  restrictions produced  the 
following  results  for the cost share equations:  
(15)  
Cost share of Finland: 
S
F  = 0.21 -  0.171nPF + o.lolnPs +  0.021nPc -  0.021nPsu 
(34.18)* (-3.10)* (1.42) (0.41) (-0.58)  
DW=l.67,  R
2
=.30 
Cost share of Sweden: 
S
s  =  0.24 +o.lolnPF + 0.021nPs +  0.151nPc -  0.211nPsu 
(20.46)*  (1.42) (0.15) (2.20)* (-2.86)* 
DW=O.36,  R
2
=.05 
Cost  share of  Canada: 
S
c  =O.lB + 0.021nPF + 0.151nPs -  0.151nPc + 0.031nPsu 
(20.71)*  (0.41) (2.20)* (-2.78)* (0.81)  
DW=O.B5, R
J
=.3O 
212 
Cost share of Russia: 
The  t values are  presented  in parentheses  under the coefficients and the statistically  
significant  coefficients (at the 5  per cent  level of  significance)  are  marked  with *.  
However,  the interpretation  of  the statistical significance  of  the coefficients is  problematic  
because the variables are  1(1) series.  The estimated coefficients are  unbiased, if all the 
variables really  are 1(1) series and if  the error  terms are 1(0) processes.  Here, the 
residuals may be 1(0) processes,  because they  seem to have tendency  to return  to  the 
mean and the mean seems  to be near  zero  (Hänninen,  1993).  However, in order to  draw 
proper conclusions about the stationarity of  the residuals they  should be  tested. Because 
the system  estimation does not maximize the individual equation  R
2
,
 it is not  an 
appropriate  measure of goodness-of-fit  in an equation  system context. Instead,  some 
other measure  is  needed to evaluate the results  (see  e.g. Berndt, 1991, p. 486).  
The coefficients  for the omitted residual countries,  S R ,  are calculated by  utilizing the 
estimation results  of  the above presented  equations  and the restrictions (9).  By  using  the 
above symbols,  the result is  as  follows: 
The  fitted values of  all the estimated cost shares are positive,  which indicates that the 
monotonicity  condition is  satisfied. The computed  Wald test value (7.16)  indicates,  that 
the  symmetry  restrictions are not  rejected in the model. 
6. ESTIMATES FOR  SUBSTITUTION AND OWN-PRICE ELASTICITIES 
The long-run  own-price and substitution elasticities of sawnwood demand from the 
supplier  countries were  computed  by  using  the estimation results  15-16  and  the formulae 
10-14. The own-price elasticities are presented  in Table 1. 
S
su = 0.21 -  0.021  nP F -  0.211 nPs + 0.031 nPc + 0.161  nPsu 
(28.70)* (-0.58) (-2.86)* (0.81) (2.44)*  
DW=O.59, R
2
=.34 
(16) Cost  share of  the residual countries: 
S
R 
=  0.16 +o.oBlnPF  -0.051nPs -0.051nPc +0.031nPsu -0.01PR. 
where: 
lnP
F  =  ln(Pp/P R) = relative price  from Finland 
lnP
s 
: =  
n il n 
Sweden 
lnP
c = ln(Pc/PR) = 
n II ii 
Canada 
lnP
Su=
= 
tl n n Russia 
213 
Table 1: Long  run  own-price  elasticities for  sawnwood from different  supplier  countries,  
1961-1990, 1961-1975 and 1976-1990. 
*) The  underlying parameter estimate (from equations 15) is statistically  significant  at  5 % level  of 
significance.  
The resulting  own-price  elasticities of  sawnwood demand from Finland,  Sweden, Canada 
and the residual countries are of the  (right) negative  sign  (Table 1).  For Russia the 
elasticity  is negative  only  in some years of  the period  (Hänninen,  1993) and its mean 
values are  positive.  Before discussing  more about the estimated elasticities,  it must be 
reminded that the interpretation of the statistical significance  of the elasticities is 
problematic,  because the variables of the  cost share equations  were not  stationary.  
Nevertheless,  it is  assumed  here, that the computed  own-price  and substitution elasticities 
are  statistically  significant,  if the underlying  parameter estimates  p i;  and p ;j  (in  the 
estimated cost share equations  15) are  statistically  significant at the 5  per cent  level of 
significance.  
According  to  the estimation results,  demand from Finland and Canada appear to be 
elastic.  This means, for example,  that a 1 percent  rise  in the price  of  Finnish sawnwood 
would decrease the demand for it by  1.60  percent.  Because  of  price-  elastic demand, 
Finland and Canada could affect their export  earnings  by pricing  policy in the United 
Kingdom.  They  could increase  their export  quantities  and total export revenues by  
lowering  the price.  Price rises are, however,  problematic.  If prices  are increased, the 
market shares  are  lost  to the competitors. When increasing  costs cause  pressure  to the 
prices, one  possibility  to  minimize the impact  is to  ensure  that price  changes  are  in line 
with those of  the competitors.  Because Finland's and the other suppliers'  price  
movements  seem to have been rather uniform on average, it seems, that  Finland has  not, 
however,  been able to much benefit from pricing  policy.  Therefore,  the  development  of  
production costs as compared  to  the competitors'  costs and exchange  rates have an 
important  role in Finland's sawnwood  exports.  
For the Soviet Union the own-price  elasticity  estimate is  almost  0  and of  the wrong sign  
and it is  not  statistically  significant.  Possible explanations  for the zero price-elasticity  of  
Russian sawnwood are  different motives for sawnwood exports  and different way  of  
trading. Sawnwood  demand from Sweden and from the  residual countries seem to  be  
price-inelastic in the United Kingdom  market. The low price  elasticity  of  Sweden's 
sawnwood indicates that it could have more possibilities  for  free pricing than its  
Supplier  Time periods 
country  1961-1990 1961-1975 1976-1990 
«a 
eii  
Finland -1.62* -1.54 -1.71 
Sweden -0.69 -0.68 -0.69 
Canada -1.58* -1.69 -1.46 
Russia 0.06 0.00  0.13  
Others -0.91 -0.89 -0.89 
214 
competitor,  Finland. In  price-inelastic  market it is  possible  to  raise  the product  price  if 
production  costs increase,  without losing  market share significantly,  ceteris paribus.  
Because  Finland seem not  to be able to use  effective pricing  policy to  increase its  export  
quantities and revenues  in the  United Kingdom  market, it could try  to push  the price  
elasticity  of sawnwood more near to zero. Price-inelastic demand could give  more 
possibilities  in free pricing  when the  production  costs  are  increasing.  The price  elasticity  
of the product  can be decreased by marketing activities,  for example,  by different 
promotional  strategies  and branding  that make the product  more  unique in the eyes  of 
the customers.  
The Morishima elasticities of substitution (MES), obtained here, are not  necessarily  
statistically  significant  (Table  2).  However, some inferences,  based on their computed  
values, are  made below. According  to the estimates,  substitutability  is  the dominating  
feature. Previous studies also indicate substitutability  between Finland and the  other  
suppliers'  sawnwood in the United Kingdom  market (e.g.  Hänninen 1986; Tervo et ai. 
1988). For Russia  the MES estimate could not  be computed,  because its  own-price  
elasticity  was  of the  wrong sign  (Table  1). The substitution elasticities calculated for 
different periods  (Table  2) indicate that  their values have not  changed  much during  the 
period  studied. 
Table 2: Long  run  Morishima elasticities of  substitution (MES)  for  sawnwood from 
different supplier  countries,  1961-1990, 1961-1975 and  1976-1990. 
MES  between Time periods  
supplier  
countries 1961-1990 1961-1975 1976-1990 
MES
,  MES,  M E  Si] 
Finland /  Sweden (FS)  1.40  1.36 1.44 
/  Canada (FC) 1.85 1.94 1.75 
/ others (FR) 1.47  1.44 1.50 
Sweden / Finland (SF)  2.22  2.15 2.29 
/ Canada (CS) 2.38 2.47 2.29 
/ others (SR) 0.85 0.86 0.85 
Canada/ Finland (CF) 1.91 1.85 1.97 
/ Sweden (CS) 1.70 1.79 1.60 
/ others (CR) 0.82 0.79 0.84 
Others  /  Finland (RF)  2.30 2.26 2.34 
"
 / Sweden (RS)  0.61 0.60 0.63 
/ Canada (RC) 1.48 1.56 1.41  
215 
It can be  seen from Table 2 that the Morishima elasticities of substitution are not  
symmetric.  For example,  a 1 per cent  rise in the price  of Finnish  sawnwood  would 
increase the relative quantity  demanded from Sweden (Mj/Mp)  by  about 2 per cent. On 
the other hand,  a  1 per cent  rise  in the price  of  Swedish  sawnwood would increase the 
relative quantity  demanded from Finland (MF/MS) by only  about 1.4 per  cent.  This means  
that a  rise  in Sweden's price  causes  a  smaller increase in demand from Finland than the 
increase in demand from Sweden caused by a rise  in Finland's price.  From this,  we can 
conlude,  that it has  been easier  to substitute Swedish sawnwood for Finnish sawnwood 
than to substitute Finnish sawnwood for Swedish sawnwood. When comparing  the 
substitution elasticities between Finland and its  other competitors,  Canada and the 
residual  countries,  the same kind  of  conclusion can  be drawn. In general,  when prices  
have increased, it has  been easier to substitute sawnwood from competing  countries for 
sawnwood from Finland,  than vice versa:  
Sweden/Finland = MES-SF= 2.22 > MES-FS= 1.40 
Canada/Finland = MES-CF= 1.91 > MES-FC=  1.85 
Others/Finland = MES-RF= 2.30 > MES-FR= 1.47. 
One should, however, notice  that the interpretation  of  the statistical significancy  of MES 
estimates is  problematic  and that the differences between the elasticities are  rather  
small. 
7. CONCLUSION 
The two-level demand system and the translogarithmic function were quite satisfactory  
approaches  in the calculation of the  substitution elasticities between Finland and its  
competitor  countries in the U.K. sawnwood import  demand. The resulting  estimates  for 
the Morishima elasticities of  substitution indicate that  sawnwood from Finland,  Sweden 
and  Canada  are substitutes in the markets of  the United Kingdom.  Compared  to the 
previous  research  the present  study  produced  the following  additional information on the 
substitution. First,  the estimated substitution elasticities vary pairwise  between supplier  
countries,  because constant elasticity  of substitution was  not assumed. Second,  the 
magnitudes  of  the Morishima elasticities  indicate that  it is  easier to substitute sawnwood 
from competing countries for Finnish sawnwood, than  to substitute sawnwood from 
Finland for competitors'  sawnwood when prices  are  increasing.  
Even though  the approach  of  the study  was  quite  satisfactory  in the estimation of  the 
elasticities,  the interpretation of  the statistical significancy of the  parameters was  
problematic  because of  the non-stationarity  of  the variables. So, in order to get more 
robust estimates of the  parameters,  the estimation method should be developed further. 
The time series properties  of the variables should be taken into  account in the estimation 
and also possible  dynamic  features of  sawnwood import  demand of the United  Kingdom  
could be included into the model. 
216 
LITERATURE CITED 
Armington, P.S. 1969: A Theory  of demand for products  distinguished  by place  of 
production.  IMF Staff  Papers,  16(1): pp.  159-176.  
Berndt,  E.  1991: The practice  of  econometricsxlassic  and contemporary. Addison-Wesley  
Publishing  Company,  Inc.,  The United States,  702 pp. 
Blackorby,  C. & R.R. Russel  1989: Will the real elasticity of substitution please  stand  
up?  -  (A  comparison  of the Allen/Uzawa and Morishima elasticities).  American 
Economic Review 79(4):  pp.  882-888.  
Chambers, R.G. 1988: Applied production  analysis  -  A dual approach.  Cambridge  
University  Press,  New  York, 331 pp. 
Chou,  J.-J. &  J.Buongiorno  1982: United States demand for hardwood plywood  imports  -  
a distributed lag  model. Agricultural  Systems  8(3): pp.  225-239.  
Chou,  J.-J. &  J.Buongiorno  1983: United States demand for hardwood plywood  imports  
by  country  of  origin.  Forest  Science 29(2):  pp. 225-237.  
Chou, J.J.  &  J. Buongiorno  1984: Demand functions for U.  S. forest products  exports  to  
the European  Eonomic Community. Wood and Fiber Science 16(2): pp.  158-168. 
Christensen,  L.R., D.W.Jorgenson  & J.J.Lau 1973: Transcendental logarithmic  
production  frontiers. Review of  Economics and Statistics,  55:  pp.2B-45.  
Diewert,  W.E. 1971: An application  of  the Shephard duality  theorem -  A generalized 
Leontief production  function. Journal of  Political Economy,  79: pp.  481-507.  
Fuss,  M.A. 1977: The  demand for energy in Canadian manufacturing.  An example  of  the 
estimation of  production  structures  with many  inputs.  Journal of  Econometrics,  5: 
pp.B9-16.  
Greene,  W.H. 1990: Econometric analysis.  MacMillan Publishing  Company.  New York. 
783  pp. 
Hseu, J-S & J.Buongiorno  1993: Price  elasticities  of  substitution between species  in the 
demand of  U.S.  softwood lumber imports from Canada. Canadian Journal of  Forest 
Research,  23: pp.  591-597.  
Hänninen,  R. 1986: Suomen sahatavaran vientikysyntä  Länsi-Euroopassa  vuosina  1962- 
1983. (English  summary: Demand for Finnish sawnwood exports  in western  Europe,  
1962-1983). Folia Forestalia,  657,  25 pp. 
Hänninen,  R.  1993: The  effect  of  relative  price  variations on the sawnwood imports  to  
the United Kingdom  from Finland. Licentiate thesis in forest products  marketing,  
University  of  Helsinki,  108 pp. 
Pindyck,  R.S. & D.L.Rubinfeld 1988: Econometric models and economic  forecasts. 
Second edition, McGraw-Hill Inc.  630 pp. 
Puttock,  G.D.  & D.M.Prescott 1992: Factor  substitution and  economies of  scale in the 
southern Ontario hardwood sawmilling  industry.  Canadian Journal of  Forest 
Research,  22: pp.  1139-1146.  
Solow,  R.M. 1955-56: The production  function and the theory  of capital.  The Review of  
Economic Studies,  23:  pp.lol-108.  
Tervo, M.,  J.Mäkelä & R.Hänninen 1988: Dynaaminen  kysyntämalli  Ison-Britannian 
maittaiselle sahatavaran tuonnille (English  summary: A  dynamic  demand model for 
British imports of  sawnwood  from different countries).  Finnish Forest  Research 
Institute,  Research  Papers,  313,  35 pp.  
Varian,  H.L. 1984: Microeconomic analysis.  Second edition,  W.W. Norton &  Company.  
New York, 348 pp.  
217 
Statistical Sources: 
CSO,  Overseas  Trade Statistics  of  the United Kingdom,  1960-1990. 
FAO, Yearbook of  Forest  Products,  1960-1990. Rome. 
Finntimber, Statistical Yearbook. Finnish  Timber Exporters'  Association. Helsinki.  
SVT, Foreing  Trade, Finland,  Al,  from different years. Helsinki.  
TTF, U.K. Yearbook of  Timber Statistics,  1976 and  1983-85. Charles K.  Norman (ed.). 
Published by  the Timber Trade  Federation,  London. 
II  

17 
The Law of One Price in 
United Kingdom  Soft  Sawnwood 
Imports—A Cointegration  Approach 
Riitta H. Hänninen 
ABSTRACT. The law  of one price  states that  prices  of homogenous commodities, defined in  a common 
currency,  are equal throughout the  world. It  implies cointegration of prices.  In  the  present study,  the  law  
was  tested  for  imports of soft sawnwood  to the  United  Kingdom  from  Finland, Sweden, Canada, and  
Russia  using the concept of  cointegration. The  data  are  quarterly and  cover  the  period 1978-1992.  The  
study  used  the  multivariate  cointegration method of  Johansen  instead  of  the  bivariate  method  commonly  
used  in earlier  studies.  The  existence  of the  law  was  tested  simultaneously  for  all  four  import prices and  
separately for  pairs of prices.  The  results  do not support the "law"  but  instead  indicate  the existence  of 
differences over  the  long run  between  different  suppliers' sawnwood  prices.  This  suggests that  imperfect 
competition models  should  be  used  in explaining and  forecasting  UK  sawnwood  imports. The  results  also  
raise  an important issue concerning testing for  competition in  forest  products markets, which  is  not  
usually done  in  connection  with the  modeling of markets. For.  Sci.  44(l):17-23. 
Additional  Key  Words: Forest  products, international  trade, Johansen's  method, unit  root  econometrics.  
In  
this
 
study
 
the
 
law
 
of
 
one
 
price
 
is
 
tested
 
for
 
United
 Kingdom (UK) sawnwood  imports from its main sup  
plier countries, Sweden, Finland, Canada,  and  Russia.  
The  law  of  one price  is a conventional  assumption in  studies  
of  commodity trade. It states that  each  good has  a  single  price,  
defined  in  a common currency  unit, throughout the world 
(Isard 1977). A number  of studies  that  modeled  forest 
product  markets  and  trade  have  assumed  the  law  was  correct,  
but  several  other  studies  have  allowed  for  price  differences  in 
the  market.  The  validity of  the  assumption has  not  usually 
been  tested. However, reliable  information  about  the  exist  
ence of price differences  in  the market  is  important  when  
choosing a suitable  modeling strategy. 
Allowance  is  made  for  price  differences  between  coun  
tries  of destination  or countries  of  origin,  for  example, in 
trade  models  based on the  theory  of Armington (1969). The 
same is  true  of  studies  modeling trade in  plywood (e.g.,  Chou  
and  Buongiorno 1983), pulp  and  paper  (e.g., Blatner  1989), 
and sawnwood  (e.g., Castillo  and Laarman 1984, Mohd  
Shahwahid  1991, Hseu  and  Buongiorno 1993, Hänninen  
1994). The  existence of  price  differences  has  been  justified,  
for  example, by  Buongiorno et  al.  (1979, p.  643), who  state  
that  lumber  quality (even for the same  species and  grade), 
business  practices,  habits, and  attachment  to a particular  
supplier  country  or specific  supplier may  differ  depending on  
the country of origin. 
The  other  approach, which  assumes the  law  of one price,  
is used, for example, in  the Timber  Assessment  Market  
Model  (Adams and Haynes 1980) and  in  the Global  Trade  
Model  (Dykstra and Kallio  1987). Brännlund et  ai. (1982) 
used  it  in  modeling Sweden's  trade  in different  forest  prod  
ucts,  as  did  Buongiorno and  Gilles  (1984) in modeling 
newsprint  trade  and  Boyd and  Krutilla  (1987) in  modeling  
lumber trade. 
There  are several  studies  testing the law  of one price for 
different  commodities, and  the results  do  not often  support 
the  law  (c.f., Isard 1977, Ardeni  1989, Knetter  1993). How  
ever,  there  are  only  a few  studies  testing the law  for  forest 
products. For  example, Buongiorno and  Uusivuori  (1992), 
tested  price  differences  in  US  pulp and  paper  exports  to 
western  European countries  and Japan. They used the  
cointegration approach and  the  Dickey-Fuller type bivariate  
method, as  did, e.g., Ardeni  (1989), and  found  evidence  for  
the  law  of one price. Because  simultaneity cannot be  taken  
into  account in  a bivariate  testing method, the  earlier  results  
may  have  suffered  from simultaneity problems. Nor does  the 
Riitta  H. Hänninen,  Researcher  of  Forest  Economics,  Forest  Research  Institute, Helsinki Research  Centre, Unioninkatu 40 A, 00170 Helsinki, Finland.  
Phone: +358 9  85705746;  Fax: +358 9 85705717;  E-mail: riitta.hanninen@metla.fi. 
Acknowledgments:  The author wishes to  thank Jari Kuuluvainen,  Lauri Hetemäki, Jussi Uusivuori,  and Antti  Ripatti  for  their valuable comments to  
improve  the manuscript  and Metsämiesten Säätiö Foundation for financial support. 
Manuscript  received  August 31,1995. Accepted November  1,1996, Copyright  © 1998 by  the Society  of American Foresters  
Reprinted  from  Forest  Science, Vol.  44, No.  1,  February  1998. Not  for  further  reproduction.  
18 Forest  Science 44(1) 1998 
multivariate  cointegration method  of Engle and  Granger 
(1987) account  for  simultaneity. 
By  contrast,  in  the  test  procedure presented by  Johansen  
(1988) and Johansen  and  Juselius  (1990, 1992), the  price  
equations can be  estimated simultaneously. Jung and  
Doroodian  (1994)  used Johansen's  procedure in  testing 
the law  for  four  regional softwood  lumber  markets  in  the  
United States and found  support for  the  law.  However,  
they  did not  resolve  the  question: for  which  two  of  the  four  
prices  does  the  law  hold. They  found  one cointegration 
relationship, which  means that  the  law  of  one price  cannot 
be  valid  for  all  four  prices  simultaneously.  In  the  present  
study,  Johansen's procedure was  also used, but  the  Jung 
and Doroodian  analysis  was  extended  by  conducting addi  
tional  tests.  These  tests  help to  determine  for  which  spe  
cific  prices  the  law  could  possibly  hold.  Also, the unit  root  
tests  for price  series  are carried out within  the  same 
framework. 
The  present  study  tested  the  law  of  one price  in  the  UK  
sawnwood  markets for  the period 1978 to 1992. The law  
implies that  prices of  sawnwood  from  different  countries  
should  be  equal in the  UK  market. If this  is  true,  the  market  
can be  characterized  as a perfectly  competitive  market, in  
which  sawnwoods  from the supplier countries  were per  
fect substitutes  for each other. Unlike  earlier studies,  a 
complete set  of  tests  was  carried  out,  and  the  multivariate  
method with  simultaneous maximum likelihood  estima  
tion was used  to allow for  more reliable  statistical infer  
ence. Contrary to the traditional  assumption and  to the 
results  of Buongiorno and  Uusivuori  (1992), the present  
results  do not support  the  law.  
Testing the Law of  One Price 
In  the present study,  the  law  of  one price  for  UKsawn  wood  
imports was tested  using  the  concept  of cointegration for  
nonstationary time  series.  The  study  used  the multivariate  
testing method  of Johansen  (Johansen 1988, Johansen  and  
Juselius  1990, 1992) instead of the  bivariate  tests  usually 
applied in  earlier  studies.  It also  expanded the  test  for  soft  
wood  lumber  prices  (Jung and  Doroodian  1994) by running 
additional  tests  in  Johansen's  framework: the stationarity of 
the price  variables  was  examined  and  the law  of  one price  was 
tested for pairs of prices.  
The  testing procedure consisted of two  stages.  First,  it  
was tested  whether  the law  holds simultaneously for  Finn  
ish,  Swedish, Canadian, and  Russian  prices.  For  this pur  
pose,  the  cointegration rank  of the  price  data  was  deter  
mined.  The rank  defines  how  many cointegration vectors, 
r,  can be  found  in  the data. Second, the  law  was tested  for 
pairs  of prices under  r.  
As an example of the  testing of the law,  let  Pit  and Pj, be  
sawnwood  prices  of  two  supplier  countries,  expressed  in  the  
UK  currency.  Assuming no transportation costs, the  law  of 
one price  implies  that  in equilibrium Pit =  Pjt.  Then, the  law  
can be stated as 
Using  lower  case letters  to  denote  logarithms and  includ  
ing an error  term,  Equation (I)  can be  written as 
The  law  of one price  implies that  the  constant  term,  a,  
should  not  be  significantly different  from  zero,  b  should  not  
be  significantly  different  from  one,  and  Uy,  should  be  distrib  
uted  identically  and  independently. Because  price  levels  are 
nonstationary,  the  ordinary  least  squares  method  would  not  
produce reliable  inferences.  Therefore, the  cointegration 
method was used.  
Assuming the  law  holds. Equation (2) can also  be 
written  as the  difference  between  the  two  prices,  Ujj,  =  pit 
~Pjr  If ,is  1(0), the  prices  are  cointegrated and the  law  
holds.  Due  to  this proportionality  of prices,  their  coeffi  
cients, P,|  and  P;1 ,  
in  the  estimated cointegration vector 
(Pi)  should  be  equal but  of  opposite sign:  P (!  =  — py, .  This  
property  is  used  in the following. 
In the  case of four  price variables, which  in  this  study  are 
prices  of sawnwood  from Finland, Sweden,  Canada,  and  
Russia, the  law  of one price  can be  written = p 2l =Py  = 
pit  in  equilibrium. This  can also  be  presented with three  
relationships that  form the first null  hypothesis: 
Hypothesis (3) was tested  by  examining whether  there  are 
three  cointegration vectors  in  the  cointegration space.  If this  
is  true,  the law holds  for  all  four  prices simultaneously. 
The  second  stage  of  the  testing procedure was  the  pairwise  
test  of  the  law  under  r. Because  only  one cointegration vector  
was found, the  law  cannot hold  for  four  prices  simulta  
neously.  Instead, it  can hold  for  two  of  the  prices.  The  second  
null  hypothesis is then  
This  hypothesis  was tested by  investigating whether  the  
restriction  P( j = — Py|  is  valid  for  any  two  prices  in  the  
cointegration vector  Pj  (restriction  A 4  in  Appendix 1). If the  
hypothesis  (4)  is  rejected,  the  law  of  one price  is  also  rejected 
for that  price pair. 
Data 
The data  for  the  study  consisted  of  prices  of  soft  sawnwood  
imported to  the  UK  market  from  Finland, Sweden, Canada, 
and  Russia. The data  were  quarterly, seasonally unadjusted, 
and  covered the  period from 1978  to 1992.  The  four  supplier  
countries  together have  accounted  for  about  80%  of UK  
sawnwood  imports  during the  period  studied, and  their  indi  
vidual import shares  have  been  quite similar  (Table 1).  This  
means that  suppliers  have  not  differed  much  in terms  of 
market  power.  
The import  prices  of  sawnwood  from  Finland, Sweden, 
Canada, and  Russia  to  the  UK were  described  by  the average  
import  unit  values  (£/m
3
)  based  on CIF (including cost,  Pu 
= (1) 
pu = a + bp  j, + u,j, (2) 
»o-  Pi,  = P  2,' Pi, = Pi, and Pi,  =  Pi, (3) 
Ho- Pi, = Pj, ('  =  1 4 . '* 7) (4) 
Forest  Science 44(I)  1998 19 
Table 1.  Suppliers'  percentage  shares  of  sawnwood imports to 
the United Kingdom.  
insurance, and  freight)  figures.  Data  on quantities and  values  
of sawnwood  (SITC 248.2-3)  imports were  taken  from  the  
Overseas  Trade  Statistics  of  the  United  Kingdom (CSO)  for  
the period  1978-1990  and  from  the  intra-and  extra-EU  trade  
statistics  (European Commission) for  the period  1990-1992.  
The  exchange rate,  £/ECU, needed for the  observations over 
the period 1990-1992  was  obtained  from  International  Mon  
etary Fund  (IMF) statistics. The  unit  price series  were de  
flated  by  the UK producer  price  index  (1985 =  100) obtained  
from the  Main  Economic  Indicators  (OECD).  All  the  analysis  
of the  study  was  carried  out  using real  price  series  in  loga  
rithms. 
Graphs  of  the prices  of sawnwood  show  sharp fluctuations 
during the period studied, 1978-1992  (Appendix 2). Fluctua  
tions in  the  suppliers'  prices  have  widened  since the  19605, 
which is presumably due to  the two main events  in the 
international  financial  markets: the breakdown  of  the Bretton 
Woods system  in  1971-1973  and  the  oil  crises  in  the  middle  
of the 19705.  After the  breakdown  of 1971-1973,  the interna  
tional  currency system changed; many countries  let  their 
currency  float, while  Finland  and  Sweden each  used  a cur  
rency index  to  define  the average  exchange rate.  
Average  import unit  values  were used  to describe  prices 
because  more  disaggregated data  on the quality composition 
of  each  supplier's sawnwood  were not available  for the  
period studied.  This  means  that  sawnwood  imported from  the  
four  supplier countries  to  the  UK  is  assumed to  be  homog  
enous enough to  allow  for  reliable  testing  of  the  law  of one 
price  using  unit  values.  For  Finnish, Swedish, and  Russian  
sawnwood, the  assumption is  well  justified  as producers  in  
these countries have much the same  raw  material and end-use 
sectors.  Canadian  sawnwood  competes with  the  other  suppli  
ers'  sawnwood  mainly  in  the  market  for  structural  sawnwood.  
Moreover, tree species  of  Canadian  sawnwood  differ  from 
the  competitors'  species  (NUTEK 1992). This  means  that  
sawnwood  from Canada  may  not  be  exactly  comparable to 
the other suppliers'  sawnwood, which must  be  taken  into  
account  in  drawing conclusions.  
Method and Empirical Results 
Cointegration and  Stationarity of  Price Variables  
The  cointegration method  presupposes  that  the  series  to  be  
tested  are nonstationary unit  root  processes.  The  existence  of 
a  unit  root  was first tested using the  Augmented Dickey-  
Fuller  (ADF) test  (Dickey  and Fuller  1979). After the 
cointegration estimation, the variables  were  also  tested  by  
Johansen's  method, and  the  results  were  compared. The  ADF 
test  results  presented  in  Table  2  are  based  on  the  equation 
including constant,  trend, seasonals,  and  three  lags. Accord  
ing to the results,  the  levels  of the  price  series  seem to  be  
nonstationary, while  the  first  differences  are  stationary. Thus, 
it  is  concluded  that  all  the levels  of  the prices  are nonstationary 
/(l) processes.  The  inferences  regarding nonstationarity are 
invariant  to  the  different  numbers  of  lags  (1  to  5)  included  in  
the  test  equation. 
Hypothesis  testing was initiated  by determining the  
cointegration rank,  r,  which defines  the  number of 
cointegration vectors in  the  price  data.  For  this purpose,  
Johansen's  method  was applied. Johansen's  method  uses a 
statistical  model  that  is  a  p-dimensional VAR(k)  process.  It 
can be  reparameterized in  error correction  form as 
where  Ax,  is  a 1(0) vector,  (X is  a vector  of constant  terms,  
D,  is  a seasonal  dummy, and  k  is  the  lag  length (k  =  1  ~..,A0 
The  constant  term can be  restricted  to the  cointegration 
space  in  the estimation  if  there  is  no linear  trend  in  the  data.  
r, rt_! and  n 
= -/ + n, +,...,+ are  coefficient  
matrices,  n  is  the matrix  of  long-run coefficients, and  it  
can  be  decomposed into  a matrix  of loadings, a, and  a 
matrix  of cointegrating vectors,  (3,  i.e., fl = ap\ The  
Table 2. The Augmented  Dickey-Fuller  (ADF)  and Johansen  unit root test results  for  prices  of  sawnwood with three  
lags. 
At, = r,A*M + +rt .,Ax,.t+l  + n*,.*  +  
+  <DD, + £,, f= l T 
Average percentage  shares  
Supplier of total  import quantity 
country 1978-1984 1985-1992 1978-1992 
(  /o) 
Finland  17 12 15 
Sweden 25 23 23 
Canada 22 31 27 
Russia 17 14 15 
Other 19 20 20 
I 
100 100 100 
Price  ADF' test statistic Decision 
Johansen
b
 
LR test  statistic Decision 
Levels  
Finland,  pf -3.14 KD 22.25* Id) 
Sweden, ps  -2.35 1(0 23.01* KO 
Canada,  pc -3.03 Id) 27.39* KD  
Russia,  pr  -3.08 KD 24.66* 1(1) 
First-difTerenced  
hp/ -4.04* 1(0) —  — 
Aps -4.47* 1(0) — 
—  
Ape -4.07* 1(0) 
— — 
Apr -3.78* 1(0) —  — 
3
 The null hypothesis  is  nonstationarity.  Critical  t-value -3.49 at 5% level. 
b The null  hypothesis  is  stationarity.  Critical x2 value 7.82 at 5%  level.. 
20 Forest  Science 44(1) 1998 
Table 3. Diagnostic  tests for residuals of price  equations  ob  
tained from  cointegration  estimation of Model (5). 
cointegration vectors  define  the stationary  linear  combi  
nations  (long-run relations) of  the  variables  in  xr  e, 
is  a 
vector  of error  terms assumed  to be  NID(O,C2), i.e., the  
error  terms  are normally and  independently distributed  
with  expectation zero and variance  matrix O. 
Different  values  found  for  r  in  the  cointegration estima  
tion  have  different  implications:  (a) If r  =  0,  the  price  series  
arenotcointegrated andOLS  in  differences  should be  used  in  
testing (Banerjee et  al.  1993, p.  256); b)  if  r  =  4,  the series  are 
individually stationary  and  OLS in  levels  can be  used  in  
testing; (c) if  0  < r  <  4  the  price  series  are  cointegrated. 
The  first  hypothesis (3) was  tested  by  determining the  
number  of cointegration vectors, r,  in  the  data.  The  Model  
(5)  was  fitted  to  the  data for  the  period 1978  through 1992.  
The  dimension  of the  estimable  model  is  four  (p = 4). It  
consists  of equations for the prices of sawnwood  from  
Finland, Sweden, Canada, and Russia.  Because  the restric  
tion  of  the  constant  term to  the  cointegration space  did  not  
affect the test  inferences, the test results  are presented 
without this restriction.  
The  diagnostic  tests  for  the  cointegration estimation  of 
Model  (5) are satisfactory (Table 3).  Autocorrelation  of the 
residuals  was examined  using  the Lagrange-Multiplier (LM) 
test, which  is  valid  for  systems  with  lagged dependent vari  
ables.  The  null  hypothesis  (no  autocorrelation) was  accepted 
with  three  lags (k  = 3)  at the 5%  level; hence  three  lags were 
used  in the  model.  Normality  of  the residuals  was  tested  with  
the  Jarque-Bera test  (Jarque and  Bera  1980) at  the  5%  level.  
The  null  hypothesis  (normality) is  rejected for  Canada.  How  
ever,  the  test  value in  that case is very  close to the  critical 
value, which  indicates  that  nonnormality is  not  a  serious  
problem. The  standard  errors of  the  equations for  Sweden  and  
Canada are high, indicating that the model  does  not fit the 
price  data  very  well.  The  high standard  errors  may  be  due  to  
sharp price  fluctuations  (Appendix 2),  which  are not  neces  
sarily easy  to model.  
The  results  of  the  cointegration estimation  of  Model  (5) 
indicate  that r  = 1. Johansen's  method  (Johansen 1988) 
formulates  two likelihood  ratio  tests  for  the  cointegration 
rank,  r
,
 the trace  test  and  the  maximal  eigenvalue test.  Since  
Table  4. Test  results  for  number of cointegration  vectors (r)  in 
price  data for sawnwood. 
both  tests  gave  similar  results,  only  the trace test  results are 
presented here.  According to  the  X trace likelihood  ratio  test 
(Appendix 1), the  hypothesis of  no cointegration  (r  =  0) is 
rejected, and  one cointegration vector  is accepted  at  the  5%  
level  (Table 4).  This  result implies that Hypothesis (3)  of  the  
simultaneous  law  of one price can be  rejected. 
The  normalized  eigenvectors  (P,)  and  their  weights  (a,) 
obtained  from Model  (5) are shown  in Table 5. Of these  
eigenvectors,  the  first  relation, P|,  which  is  presented in the  
first  column  of  Table  5,  is  most  highly correlated  with  the  
stationary part  of the process  Ax,  when  corrected for  the  
lagged values  of  the  differences.  Thus, fl s  is  the  cointegration 
vector  found  in  the present price  data. 
The  normalized  coefficients, a ( ,  in  the  first column  of  the  
weights in  Table  5, are the  weights with  which the  error  
correction  term  enters  the  four  equations of the system. The 
a's  are adjustment coefficients  that  describe  the  average 
speed of adjustment toward the estimated  equilibrium state. 
A low  coefficient  indicates  slow  adjustment, while  a high 
coefficient  indicates  rapid adjustment. 
After determining the cointegration rank,  r,  the unit  
root  test  was  applied for each  price  variable, under  r( r  was 
found  to  be  equal to 1). The  null  hypothesis was  that  of 
stationarity:  H0:  p it  
= 1(0),  («  = 1,...,4). The  test  was  done  
by  restricting  the  coefficient, P,],  of  the  testable  price,  p ir  
to  unity  and  the coefficients  of the  other  three  prices  to 
zero  in  the  cointegration vector  P| (restriction A  3  in 
Appendix 1). According  to  Johansen's  test, the  null  hy  
pothesis  of  stationarity is  rejected for  all  variables  (Table 
2). Thus, the  results  of the ADF and  Johansen's  test  are in  
line  with  each  other  and  indicate  the  nonstationarity of  the  
variables.  The  basic  difference  in these  tests  is  in  respect  
to the null  hypothesis. The ADF test  assumes 
nonstationarity as the  null, while  Johansen's  test  assumes 
stationarity. 
Hypothesis  (3) is  rejected in  the present study,  because 
only  one cointegration relation  was  found  in  the  data. This 
means that  the  law  of one price  cannot  be  valid  for  all  four  
prices  simultaneously. The  result  r=  1 indicates,  however,  
Table 5. Results  from  cointegration  estimation  of  Model (5).  Normalized eigenvectors  (β1,,..., 
β 4
) for  Finnish, Swedish,  
Canadian,  and Russian sawnwood prices  and respective  weights  (α1,,...,α4). 
Tests/  Finland Sweden Canada Russia 
standard errors A  pf  A ps  Ape Apr 
Autocorrelation* 7.58  6.04 6.03 8.97 
Normality
6
 4.82 3.96 6.21* 0.42 
Standard errors  0.07  0.12 0.10 0.07 
a
 Critical x2 value  9.49 at 5% level. 
b
 Critical value  5.99 at 5% level. 
Null 
hypothesis 
Eigenvalues 
\ 
LR-test statistic Critical values 
for  trace  test X '  
* trace 
H», r  = 0  0.43 53.52* 47.20 
H„:r< 1 0.25 20.43 29.90 
H
0 :  r  
< 2 O.I  5 9.10 1 5.40 
//„: r < 3 0.03 1.63 3.80 
3 Critical  value at 5% level (Doornik  and Hendry  1994}. 
Price  variables  Eigenvectors Weights 
by  country  p, P» P, P. a. a-2 a, 
Finland 1.00 1.17 1.31 -2.08 -0.55 -0.07 0.04 0.04 
Sweden -0.77 1.00 -0.60 0.08 0.76 -0.13  0.18  0.05 
Canada -0.37 0.96 1.00 0.22 -0.23 -0.14 0.02 -0.06 
Russia 0.58  -0.58 -1.04 1.00 -0.44 -0.00 0.15 -0.01  
21 Forest  Science 44(I) 1998 
that  a stationary  relationship between  any  two  price  vari  
ables  is  possible.  This  is why testing of the  law  was 
continued  by  examining  pairs  of  prices.  
Testing the  Law  of  One  Price for Pairs  of  Prices  
Next,  the  cointegration vector,  (3|  (Table 5),  estimated  
from Model  (5)  was  used  to  test  whether  the  law  of one price 
holds  in  pairwise  comparisons of prices  (Hypothesis  4).  The  
cointegration vector  can be  interpreted in  terms  of  an error  
correction  mechanism measuring the  changes in  prices  of 
sawnwood  from  Finland.  The  long-term equilibrium rela  
tion, p|,  between  price  variables  can be  presented as 
where  pf,  ps,  pc,  and pr  are  prices  of sawnwood  from 
Finland, Sweden, Canada, and  Russia.  In  Equation (6), the 
coefficient of  the  Canadian  price  is  small  compared to  the 
coefficients of  the  Finnish  and  Swedish  prices.  Moreover, 
the  coefficient of  Russian  price  is  negative. One  reason for  
the  small coefficient  of the Canadian  price may  be  in the 
quality composition of  Canadian  sawnwood.  As was men  
tioned earlier, sawnwood  from  Canada  may not  be  quite 
comparable  to  the  other  suppliers'  sawnwood  in  the  UK  
market.  The  negative coefficient of the Russian  price  is 
more difficult  to explain. It may  be  due  to the  export 
pricing policy,  which  has  been  different from  that  of the  
other countries.  
The  validity of  the  law  of  one price  was  tested  under  r  =  1 
for  two  price  variables  at  a time  (Hypothesis  4) by assuming  
that they  are proportional. According to  Equation (6), mag  
nitudes  of the  price  coefficients  differ  so widely from  each  
other  that  proportionality  between  any  two  prices  is  unlikely. 
Proportionality  would  imply that  the  coefficients  in  the  
estimated  cointegration vector,  (3,  (Table 5),  would  be  about  
the  same magnitude and  of  opposite sign. To  test  the  law  by  
price  pairs,  Johansen's  restriction  (A 4),  presented in  Appen  
dix  1,  was  applied. The  Hypothesis  (4)  was  tested  by  restrict  
ing two  price  coefficients  of  the cointegration vector  at a time  
to  P,|  =  -  (3 |.  The  test  results,  presented in  Table  6,  indicate  
rejection  of  the  null  hypothesis  for  all  pairs of  prices.  Conse  
quently,  the  law  of  one  price  is  rejected for  the  UK  sawnwood  
imports for the four  supplier countries.  
Table 6. Test  results  for  law of  one  price by  price  pairs  obtained 
from  Model (5)  with restictions  on β1 under  r 
=
 1. 
Summary  and Conclusion 
In  this  paper  the law  of one price  was  tested for  the  
United  Kingdom's soft  sawnwood  imports  from  Finland,  
Sweden, Canada, and Russia. Johansen's  multivariate  
cointegration approach was  used  in  the  testing. The  test  
was  applied using quarterly import unit  values  deflated  by  
producer prices for the  period 1978 through 1992. 
The  results  do  not  support  the  law  of  one price  for  UK  
sawnwood  imports.  Thus, the  price  differences  among  
supplier  pairs  did  not  tend  to  disappear over the  long run. 
This  result  differs from the  traditional  assumption that  
commodity  prices  are perfectly  arbitraged, with  instanta  
neous exchange-rate pass-through, at  least  in  the  long  run. 
It also  differs  from  the  results  of  Buongiorno and  Uusivuori  
(1992), who  tested  the  law  for  US  paper  product exports  
across destinations.  
The  result  do  agree,  for  example, with  Isard  (1977),  
who  compared prices  of  different commodities, including  
paper  products,  across exporters  (US  and  Germany);  Ardeni  
(1989),  who  tested  the  law for  certain  exporters  of agricul  
tural  and  metal  products; and Knetter  (1993), who found  
price  differences  across exporters  for  certain  agricultural  
products.  
The  reasons for  the  price  differences  between  suppliers  
may  be  connected  with, for  example, exchange-rate pass  
through and  institutional  factors that  influence  the  prices  
of  each  country. Moreover, the  costs  of  arbitrage may have  
been  high  in  sawnwood  markets, in  which  established  
customer  relationships  play  an important role.  However,  
the  price differences  between  Canada  and  the  other  coun  
tries  may  also  be  a result  of the  difference  in  quality  
composition  between  the  Canadian  and  other  sawnwoods.  
Canadian  sawnwood  may  not  be  quite comparable with  the  
other  suppliers' sawnwood.  
The  results  of the  study  suggest  that  models  distin  
guishing products  by  country of origin  and allowing for  
price  differences  between exporters  can be  applied in  
modeling UK sawnwood  imports.  They also  raise  an im  
portant issue  concerning the  assumptions to be  made  
regarding  competition in  sawnwood  markets. Certain  fea  
tures  of  western  European sawnwood  trade  support the  use  
of  imperfect  competitive  models.  The market  can  be  char  
acterized  as having four  dominant  exporters:  Finland, 
Sweden,  Canada, and  Russia.  This  intuitive  feature  should,  
however, be tested empirically  if  it  is to  be  used  in  model  
ing  markets.  
Literature Cited 
Adams, D.M., and  R.W.  Haynes.  1980. The 1980 softwood timber assess  
ment market model: Structure, projections and policy  simulations. For.  
Sci. Monogr.  22: 1-65. 
Ardeni, G.P.  1989. Docs  the law of one  price  really  hold for commodity  
prices?  Am. J. Agric.  Econ.  71:661-669. 
Armington, P.S. 1969.  A theory  of demand for products  distinguished  by 
place of production.  Internal. Mon. Fund Staff Pap.  6:159—176. 
Banerjee, A., J. Dolado, J. Galbraith, and D. Hendry. 1993.  Cointegration,  
error correction and the economic analysis  of non-stationary  data.  Oxford  
University  Press  Inc. 329 p.  
pf  = + 0.77  ps  + 0.37  pc  -  0.58  pr (6) 
Testable  price  pair  ~u '
n
 
pf&ps P„ = -P„ 21.75* 
pf&.pc P„  = -P„  
29.90* 
pf  & pr  Pll = ~Pjl 29.92* 
ps  & pc  P2I = P3I 28.43* 
ps  &  pr  P„ = -P<, 27.67* 
pc&  pr  Pjl  ~ P4I  
24.46* 
Symbols:  pf, ps,  pc,  and pr are  prices  
of sawnwood from Finland, Sweden, 
Canada, and Russia and P (1 is  coefficient of  respective  price  in 
cointegration vector pr 
8 Critical  x
2
 value  7.82 at 5%  level. 
22 Forest  Science 44(1) 1998 
Blatner, K.A. 1989. An approach to the estimation of import  price  elastici  
ties by  supplier.  For.  Sci. 35:30-41. 
Boyd,  R.,  and K.  Krutilla.l9B7. The welfare impacts of  U.S. trade  restric  
tions  against  the  Canadian softwood lumber  industry:  A  spatial  equilib  
rium analysis.  Can. J.  Econ.  1987:17-35. 
Brännlund, R., L.  Lundberg,  and  A. Wiberg. 1982.  Determinants  of Swedish 
market shares for  pulp, paper  and sawnwood in Europe.  Swedish Univ.  of 
Agric.  Sci., Dep.  of For.  
Buongiorno, J., and J.K.  Gilles. 1984. A model of international trade of 
forest  products,  with an application  to newsprint.  J. World For.  Rcsour.  
Manage.  1:65-80. 
Buongiorno,  J.,  and  J.  Uusivuori.  1992.  The law of  one  price  in  the  trade  of 
forest  products:  Co-integration  tests  for  U.S. exports  of  pulp  and paper.  
For.  Sci.  38:539-553. 
Buongiorno, J., J-J.  Chou, and R.N. Stone. 1979.  A monthly model of the 
United States  demand for  softwood lumber  imports. For.  Sci. 25:641- 
655. 
Castillo, S.,  and  J.G. Laarman. 1984. A market-share  model to asses  price  
competitiveness  of  softwood lumber  exports  to Caribbean  markets. For.  
Sci.  30: 928-932. 
Chou, J-J.,  and J. Buongiorno. 1983. United States  demand for hardwood 
plywood  imports by  country of origin.  For.  Sci.  29:225-237. 
CSO.  Overseas  Trade  Statistics  of the  United Kingdom,  1978-1990,  London. 
Dickey,  D.A., and W.A. Fuller. 1979. Distribution of estimates for 
autoregressive  time series  with unit root. J.  Am. Statist. Assoc. 74:427- 
431. 
Doornik, J.A., and  D.F. Hendry.  1994. PcFiml 8.0.  Interactive  econometric 
modelling  of dynamic  systems.  Inst,  of Econ.  and Stat., Univ. of Oxford. 
International Thomson Publishing,  London.  436 p.  
Dykstra,  D.P.,andM. Kallio. 1987. Introduction to the lIASA Forest  Sector  
Model. P.  139-167 m The  global  forest  sector: An  analytical  perspective,  
Kallio, M., et ai. (eds.). Wiley,  Chichester.  
Engle, R.F.,  and  C.W.J. Granger. 1987. Co-integration  and  error correction: 
Representation,  estimation and testing.  Econometrica 55:251-276. 
European Comission.  Intra- and Extra-EU trade, CD-ROM,  1988-1992, 
Luxemburg.  ISSN 1017-6594.  
Hanninen, R. 1994.  Price elasticities of substitution  between Finland  and the 
othcrsupplicr  countries in the demand for  the United Kingdom  sawnwood 
imports.  P. 204-217 in Proc.  of  the  biennial  meeting  of Scandinavian 
forest  economics,  Helles, F.,  and M. Linddal (eds.).  Scand.  For.  Econ. 35, 
Denmark.  
Hseu,J-S.,  and  J. Buongiorno. 1993. Price  elasticities of  substitution between 
species  in the  demand of  U.S. softwood lumber  imports  from  Canada. 
Can.  J. For.  Res.  23:591-597. 
IMF.  International Financial Statistics, 1993:4.  Washington,  DC. 
Isard,  P. 1977.  How far can we push  the "law of one price"? Am. Econ.  Rev. 
67:942-948. 
Jarque, C.M., and A.K.  Bera. 1980. Efficient tests for normality,  
homoskedasticity, and serial  independence  of regression  residuals.  Econ.  
Letters  6:255-59. 
Johansen, S. 1988. Statistical analysis of cointegration  vectors. J. Econ. 
Dynam.  Control. 12:231-254. 
Johansen, S.,  and K. Juselius. 1990. Maximum likelihood estimation and 
inference  on  cointegration —with applications  to the demand  for  money. 
Oxford  Bull. Econ.  Statist. 52:169-210. 
Johansen, S., and  K. Juselius. 1992. Structural tests in a multivariate 
cointegration  analysis  of the  PPP and the UIP for UK. J. Econ.  53:211- 
244. 
Jung, C, and  K. Doroodian. 1994. The law of one price  for U.S. softwood 
lumber: A multivariate cointegration  test. For.  Sci.  40:595-600. 
Knetter, M.M. 1993. International  comparisons  of pricing-to-market  behav  
ior.  Am. Econ.  Rev.  83:473-486.  
NUTEK.Närings-ochteknikutvecklingsverket.  1992.Sägverksindustrininför  
sekelskiftct.  Del 11,  R 1992:40. Stockholm.  370 p.  
Mohd Shahwahid, H.O.  1991. Further  assessment of  the  price competitive  
ness  of Malaysian  Lauan lumber imports  in the United States.  For.  Sci.  
37:849-859. 
OECD. Main Economic Indicators,  Historical Statistics, 1962-1991 and 
1993:10, Paris.  
Appendix  1. Johansen's Testing Method 
In  the  following, Johansen's  testing  method  is  briefly  
summarized  (for details, see  Johansen  1988, Johansen and  
Juselius  1990, 1992). The  maximum  likelihood  estimators 
of the  a  and  (3  parameters for  Model (5) are obtained  by  
solving an eigenvalue problem (Johansen 1988). To  deter  
mine  the  cointegration rank  (i.e., number  of  cointegration 
vectors),  Johansen  formulates  two  likelihood  ratio  tests:  
the  trace  test and  the  maximum eigenvalue test.  The  null  
hypothesis can be  presented as 
Because  p = 4 in the model, the determination  of 
cointegration rank  involves  four  different  stages in  testing 
H
O.  The  null  hypothesis applied for  the  trace  test  is  that  the 
number of  cointegrating vectors  is  less  than  or equal to  r, 
where  ris  0,  1, 2,  or  3.  In  each  stage  the  null  hypothesis  is 
tested  against the  general alternative:  r =  0 against  r  > 1, 
r  > 1 against r  >2,  etc.  The  hypothesis (Al) is  tested  using 
the likelihood ratio test of the form 
where  T  is  the  number  of observations  and the  Xfs  are the  
smallest  squared canonical  correlations  (eigenvalues). The  
testing  strategy is  a  multivariate  analogue of the  Dickey-  
Fuller  (1979)  test.  
When  the number of  cointegration vectors,  r, is deter  
mined  by  the  data, it  is  possible to  test different  hypotheses 
by  restricting  the  cointegration vectors,  p,  or  their  weights, 
a. Testing is  done  by estimating Model  (5)  with  restric  
tions  and  comparing the  results  to  the  Model  (5) without  
restrictions.  The  test  statistics  are asymptotically  y} dis  
tributed.  
Under  r, the Johansen  method  (Johansen and  Juselius  
1990, 1992) formulates  the  following restriction on p,  
which  can be  used  to test  the stationarity  of a variable:  
In  (A  3), p  is  the  dimension  of the  model (here p  =  4), r 
is  the  number  of  cointegration vectors,  of  which  rj  vectors  
H
0  :  rank(fl)  
< r (Al) 
-2  In  (Q)  =  -r£ln(l-X,.) (A  2)  
i=r+l  
P  =  («,¥) 
(A  3) 
where  H(p x r,), *|/(  p  x r 2)  and /■ =/j  +r 2 
Forest  Science 44(1)  1998 23 
are known  a priori. Matrix  H  is a design matrix, where  the 
restriction  on the (3  matrix  is  formulated  by  the  researcher.  
The  test  can be  done  by  estimating  Model  (5)  with  restric  
tions, so that  r ,  cointegration vectors  are restricted, and  
the  remaining r2 cointegration vectors  (included in  matrix  
y) are unrestricted.  
Because  the number of cointegration relations  was found  
to  be  one,  r2 -  
0 and  r l  =1 
in  the present  study.  Stationarity 
is  tested  by restricting  the  coefficient  of  the testable  variable  
(P,l) to unity  and the  other  coefficients to zero in  the  
cointegration vector  p,  and  by  examining whether  the  result  
ing linear  combination  is stationary.  These restrictions  are 
defined  by design matrices  H,  which  are of  the  form  (1,0,0,0), 
(0,1,0,0), etc.  Each  variable  is  tested  separately  by  restricting  
the  coefficients  respectively  in  the  //matrix.  The  null  hypoth  
esis  for  the Johansen  test  is  stationarity. 
The  hypothesis  presented in  Equation (4), i.e., the  law  of 
one price  by  pairs of  prices,  can be  tested  using  the  following 
restriction  of Johansen:  
In  (A 4)  the  restrictions  are defined  by  <p, which  is  an s 
x  r  matrix, where  r(p  -  s)  is  the  number  of  restrictions  and 
r  is  the number  of  cointegration vectors.  The testable  null  
hypothesis  (4) between  two price  variables  can be  ex  
pressed  as  P,|  =  — Py, .  With  four  variables, under  r=  1, the 
design matrices, H,  are of  the  form  (1, -1,0,0),  (1,0,-1,0), 
etc.  If the  restriction  is  rejected, the  law of one price  
between  the  two  prices  is  also  rejected.  
Appendix  2. Real  Unit Prices of  Soft Sawnwood Imports  to the United Kingdom  from Finland,  
Sweden, Canada, and Russia.  
B = Hw  or n  = aip'  W H v v
(A  4) 
where H(px. s), cp(i x r) and  r< s < p  

III 

1946 
Testing  arbitrage  in  newsprint  imports  to  
United  Kingdom and  Germany 
Riitta Hänninen, Anne Toppinen, and  Pertti  Ruuska 
Abstract:  The arbitrage was tested  for  United  Kingdom and  German newsprint imports from Finland,  Sweden, and Canada 
using the hypothesis  of  "the law of  one price"  (LOP).  The  LOP  states  that  prices  of  homogenous commodities, defined  in  the 
common currency,  are equal throughout the world. Concept of cointegration and Johansen's  multivariate  method were used  in 
testing the  LOP  in the  long  run.  The empirical data  were  quarterly covering the period 1980-1994.  The hypothesis that  the 
LOP  holds  for  all three  supplier countries simultaneously was  not supported by the multivariate test  results  in either importing 
country.  Cointegration tests rejected the simultaneous  LOP  between the three  import prices in both  markets. The LOP  was 
also tested  pairwise, and  it was  found to  hold between Swedish and Canadian prices  in the  U.K. market. 
Rlsuml :  On  a  verifiö l'arbitrage en utilisant l'hypothese de la « loi d'un seul prix  » (LOP) pour  les  importations de papier 
journal du Royaume Uni  et de l'Allemagne en provenance  de Finlande,  de Suede et du Canada. La  LOP  stipule que les  prix  
des biens  homog£nes de consommation  courante, definis par  une devise commune,  sont egaux  ä travers le  monde. Le concept  
de la co-intdgration et la methode multivariee de Johansen ont etd  employes  pour  verifier la LOP  ä long  termc. Des  donnees 
empiriques trimestrielles couvrant  la periode 1980-1994 ont ete  utilisees. L'hypothese selon  laquelle la LOP  est  valide  pour  
les  trois  pays  fournisseurs  simultanement, n'a pas  ete confirmee par  les  resultats  de la methode multivariee pour  aucun des 
pays  importateurs. Les tests  de co-integration ont rejet6 la simultaneity de la LOP  entre les  trois  prix  d'importations pour  les  
deux marches.  La  LOP  a aussi  ete  verifiee  par  paire, et on a  constate  qu'elle etait validee entre  les  prix  suedois et canadien sur 
le marche  du Royaume Uni. 
[Traduit par  la Redaction] 
Introduction  
The  arbitrage  implies  that  the  prices  of  a particular  commodity  
in  various  international  markets  are equalized when  expressed  
in  the  same currency  and  consequently,  the  "law  of  one price"  
(LOP)  holds.  The  LOP  assumption  implies  perfect  competition 
in the  market, and  it  is a commonly  used  assumption  in  mod  
elling commodity  trade.  The  present study examines  for  arbi  
trage in  the  U.K.  and  German  newsprint imports from Finland,  
Sweden, and Canada  by  using the  multivariate  cointegration 
approach. The  focus  of  this  study  is  on the  long-run  relation  
ship between  the  prices of the  three  supplier countries.  
A  number  of  studies that  model  forest  products  markets  and  
trade  have  assumed  perfect  competition, but there  have  also  
been  several  studies  that  have  assumed  imperfect competition, 
allowing for price differences across markets.  The  trade  mod  
els  assuming  imperfect  competition and  allowing  for  price  dif  
ferences between products of different countries are 
commonly based  on the  Armington (1969) model.  The  basic  
assumption of the  Armington (1969) model  is  that a product 
originating from  a  supplier  country  is  not a perfect substitute  
for  the  same product originating from  another  supplier coun  
try. For  example,  price  differences  are assumed in  studies that  
model  trade  in  plywood  (e.g., Chou  and  Buongiorno 1983), 
pulp and  paper  (e.g., Blatner  1989  and  Laaksonen  et al. 1997), 
Received  January 2, 1997.  Accepted  September 2, 1997. 
R.  Hänninen, 1 A. Toppinen, and P. Ruuska.  Finnish  Forest  
Research  Institute,  Helsinki Research Centre,  Unioninkatu 40 A,  
00170 Helsinki,  Finland. 
1 Author  to  whom  all correspondence should be addressed,  
e-mail: riitta.hanninen@metla.fi  
Can. J. For.  Res.  27: 1946- 1952 (1997) 
and  sawnwood  (e.g., Hseu and  Buongiorno 1993; Hänninen  
1994). The  other  approach, which  assumes perfect competition 
and  the  LOP,  is  used,  for  example,  in  the  Timber  Assessment 
Market  Model  (Adams and  Haynes 1980) and in  the Global  
Trade  Model  (Dykstra  and  Kallio  1987). Brännlund  et ai.  
(1982) applied the  assumption  in  modelling Sweden's  trade  in  
newsprint and  the other  forest products. 
For  the newsprint market, both  perfect and  imperfect com  
petition have  been  assumed, and  it  is  not clear  which  one is  
preferable in  modeling this  market. Previous  models  of the 
newsprint industry in  North America (e.g.,  Dagenais 1976; 
Booth et  al. 1991) were based on  oligopolistic market structure  
with  dominant  firm price leadership, whereas  Buongiorno and 
Gilless  (1984) modelled  newsprint trade, and  Andersen  (1993) 
studied  the  structure  and development of the European news  
print market  assuming competitive markets.  
If empirical  study  assumes LOP  when  it  actually  docs  not  
hold, inferences  are invalid  and  the forecasts  are unreliable.  
Under  perfect competition, relative  costs determine  export 
quantities, whereas under  imperfect competition relative  
prices  are the  key  determinants  of  exports.  Policy  recommen  
dations  based on competitive market model  will  not have  the 
expected  effects in the  presence  of  imperfect competition, and  
the  costs  of  incorrect  policies may  be  large. Thus,  in  modelling 
and forecasting  the  demand for  and  trade  of  forest  products, it  
would be useful to have direct tests  for  the existence  of the LOP. 
The  present study  tests the LOP  in  the  U.K.  and  German 
newsprint  markets  between  the  main  supplier countries, Sweden, 
Finland, and Canada  by  using Johansen's  multivariate  cointe  
gration analysis  (Johansen 1995).  Earlier, Johansen's  method  
in  testing  the  LOP  in  forest  products'  markets  has  been  used  
by,  for example, Hänninen  ( 1 998),  who  tested  the  LOP  for  the  
U.K. sawnwood  imports from the  four main  supplier  countries.  
d 1997 NRC Canada 
Hänninen et ai.  1947 
The  LOP  between the  suppliers  was  rejected in  the  simultane  
ous  and  pairwise  tests.  On  the  other  hand, Jung and  Doroodian  
(1994),  who  studied  the  LOP  in  connection  with  four  regional 
softwood  lumber  markets  in  the  United States  accepted  the  
LOP.  They found  one cointegration vector, but  did  not  deter  
mine  for  which  one of  the  four prices  does  the  LOP  hold.  One  
cointegration vector in  the  case of  four  prices  means that  the  
law cannot  hold  for  all  four  prices  simultaneously (e.g.,  Good  
win  and  Grennes  1994). The  main  finding of  the  present  study  
is that  simultaneous  LOP  was  rejected in  the  U.K. and  German  
markets.  In  pairwise  analysis,  the  only  case where  the  LOP  
was  accepted  was  between  the  prices  of  Sweden  and  Canada  
in the U.K. market.  
The  paper  is organized as follows.  The  outline for  the test  
ing procedure for  the  LOP is described  in  the  next  section, 
following by  the  data  section  and  empirical results.  Summary 
and  conclusions  are presented in  the final section.  
The  LOP testing procedure 
At the  most  aggregate level,  long-run relations  of  price  indices  
of total  trade  for different  countries are examined  using the 
theory of  purchasing  power  parity  (PPP).  The  LOP,  being gen  
erally concerned  with  the  commodity level, can  be  seen as a 
disaggregated counterpart of  the  PPP.  Because  of  the  impor  
tance  of  the  LOP  assumption it  has  been widely tested  at the  
aggregate level  of economy  (e.g., Isard  1977; Ardeni  1989; 
Knetter 1993) using  the  concept  of  cointegration and  the  Engle 
and Granger (1987) method  with  two variables.  The results  
from the  previous studies generally do  not  support the  LOP  
(see,  for  example, the  survey  of  Officer  1986). For  forest  prod  
ucts  there  are  only a few  studies  that examine  the  validity of 
the LOP. 
Buongiorno and  Uusivuori  (1992) tested price  differences 
in  U.S.  pulp and  paper exports  to different  countries  using 
bivariate  cointegration analysis  and  found  evidence  that  the  
LOP holds.  However,  when  pairs of prices are  considered  
separately, any linkages that may operate  through a third  mar  
ket  price  arc ignored. Also,  the approaches using  the  
Dickey-Fuller  type  methods actually  test the  null  hypothesis 
that  the  LOP does  not hold  and  thus  perform  the  test  indirectly. 
Our  testing procedure  differs  from most  of  the  earlier  studies  
with  respect  to estimation  method.  We  employ the  multivariate  
cointegration method  as developed by  Johansen  (1988)  and  
Johansen  and Juselius  (1990). In using the multivariate  
method, the  price equations can be  estimated  simultaneously. 
In  Johansen's  procedure, it  is possible  to  test  directly the  null  
hypothesis,  that  the  LOP  holds  simultaneously  in a multivari  
ate context. 
Traditionally, tests  of  the LOP  have  used  the following bi  
variate  procedure that we  consider  here  as  an example.  Let  pu 
and  pj,  be  newsprint  prices  of two  supplier countries, expressed  
in  the  currency  of  the  destination  country. Assuming no trans  
portation  costs,  the  LOP  means that  in equilibrium p„ =pjr We  
write  the  following regression equation, with prices  in  loga  
rithmic  form: 
If  the  LOP  holds  in  [l],  the  restrictions  a = 0  and  P  = 1 should  
be  accepted. The test  is,  however, conditioned  on the assump  
tion  that prices  are  not  simultaneously determined  and  that  
both  price series  are stationary, i.e., 1(0)  processes.  If one of  the  
two  prices  is  nonstationary (integrated)  and  the  other  is not,  the  
LOP  between  them  is  automatically rejected. If both prices 
have  the  same  order  of  integration and  if  they form a  cointe  
grated system, the  theory of  cointegration (Engle  and  Granger 
1987) can be  applied to examine  the  law.  Specifically in  [l],  if  
prices  are nonstationary (i.e., /(1)), and  if  u,  is  stationary  
(i.e.,  /(0)),  then  pjt and  p,,  arc  cointegrated.  In  this  case,  the  
LOP  is  said  to  hold  in  the  long run and  the  cointegration test  
can  be  transformed  into a stationarity  test  of the  difference  
between  the  two prices  (Baffes  1991). Consequently, the  law  
can  be  tested  by  taking the  difference:  
and  testing whether z,= 1(0).  Because  the ordinary  least  
squares  estimate  of P  from [l] is biased  in  small samples 
(Stock  1987), [2] was preferred over [1 ] in  many of the pre  
vious studies that test for cointegration. For example,  
Buongiorno and  Uusivuori  (1992) used  the difference  of two  
price  variables  to  test  cointegration and  examine  the  strong  
LOP.
2 
In the  case of three  price  variables, which  in  this  study  were 
the unit  prices  of newsprint  from  Finland, Sweden, and  Canada, 
the  LOP  holds  if  the  suppliers'  prices  in the  long-run equilib  
rium  are equal in a destination  country. This  situation  can be  
described  by  p u= p 2,  = Pj,  in  equilibrium. It  can also  be  pre  
sented in  the  form  of  a  null  hypothesis  such  as 
If the prices are nonstationary, /(l) processes,  the hypothe  
sis  [3] can be  tested  by  taking the  differencesP\,~P2t =z \t  and 
Ph  ~ Pii = z2i  afid  testing whether  z u  and  z2,  are  7(0).  
Under  the assumption that  the  LOP holds in  the  market, the 
prices  are jointly  determined.  This  means that  simultaneity 
should be  taken  into  account in  testing. The cointegration 
method  of Engle and Granger (1987) may suffer from the si  
multaneous  equation problem. Moreover, in  the  case of  three  
variables, the  number  of possible cointegration vectors  may  be 
one or two. By  using the  method  of  Engle and  Granger it  is  
possible to  obtain  only  one vector.  In  addition, testing for  pa  
rameter  restrictions  is not possible. These problems are re  
solved  in  the  present  study  by  using the  multivariate  maximum  
likelihood  procedure of Johansen  (1988) and  Johansen  and  
Juselius  (1990, 1992). The  method  is  briefly summarized in  
the  following paragraphs. A comprehensive description  of  the  
method  can  be  found in  Johansen (1995)  and  Banerjee et  al.  
(1993). 
The  rank  test  in  the  Johansen procedure defines  how  many 
cointegration vectors,  r, can  be  found in  the  data.  The  cointe  
gration rank  is  determined  by  estimating a  
VAR(A) model  system (e.g., Johansen 1995, p. 11): 
where  x,  is  a vector  of  empirical variables, which  arc  the  news  
print  import prices  from  Finland, Sweden, and  Canada  into  the  
United  Kingdom and  Germany.  Thus,  the  dimension  (p )  of  the 
2 Buongiorno and Uusivuori  (1992)  also tested  a  weaker form of 
the  LOP.  This  was  done by  testing the integration of u, in cq. 1  
without  restricting a and P, respectively,  to zero and one in 
ordinary least  squares  regression.  
© 1997 NRC Canada 
[l] Pi,  =°- + PPjl  +U l 
[2] z,=  Pi,~  Pj, 
[3] Ha:  P\,=p2, and  pl,=pit  
[4] x,=A\Xj_i  + ...  +A +  jx + ODt  
+  £|, t—  \  j..T 
1948  Can.  J. For. Res. Vol. 27, 1997 
Tabic 1. Percentage shares  of Finland,  Sweden,  and Canada  from 
imports of  newsprint to the United Kingdom and Germany, 
1980-1994. 
VAR  model  is  three.  Symbol n  is  a  vector  of  constant  terms,  D,  is  
a vector  of  seasonal  dummies, kis  the  lag length (k=  1,..., N)  
and  e, is  a vector  of  error  terms  assumed  to  be  N1D(0, CI).  
Equation 4 is reparameterized in  error-correction  form  
(e.g., Johansen  1995, p.  89)  such  as 
where Ax,  is  an 1(0) vector.  In  [s]  the  constant  term,  |x,  can be  
restricted  to  cointcgration space  in  the  estimation  if  it  is  as  
sumed  that  there  is  no linear  trend  in  data,  1*),..., Tt_, and n  = 
-I  +  n| + ...  +  n t are coefficient  matrices.  
T describes  short  
term  dynamics of  the  process  and  n  is  the  matrix  of  long-run 
coefficients.  The  coefficient  matrix  n  can be  decomposed into  
a matrix  of  loading  vectors,  a, and  a  matrix of  cointegrating 
vectors,  P,  such  that n  = otP'.  The  cointegration  vectors  define  
the  stationary linear  combinations  (long-run relations) of  the  vari  
ables  in  xr Further,  Johansen (1988)  formulates  a likelihood  
ratio-based  trace  test  for  testing cointegration  rank,  r,  where  
the  null  hypothesis is H o \  r=  r 0 against  ro<r<,p. 
When  the  number  of  cointegration vectors,  r,  has  been  de  
termined, it  is  possible  to  test  hypotheses  under  r  by  imposing 
linear  restrictions  on the  matrix  of  cointegration vectors,  P,  and  
loadings, a (Johansen and  Juselius 1990, 1992). The tests for 
these  linear  restrictions  are asymptotically  %
2
 distributed.  Ac  
cording to Goodwin  and  Grenncs  (1994), simultaneous  LOP 
between  prices requires n -  1 cointegration vectors,  where  n 
is  the  number  of  price  variables.  In  our  case (ii  = 3 in  each  
market) this  would mean that  r=  2. Under cointcgration a 
one-to-one relationship between  prices  is  required for  the  si  
multaneous  LOP  to  hold  (Baffes 1991). This  restriction  is jus  
tified  since  wc  use CIF (cost,  insurance,  and  freight)  prices  in  
each  destination  country, and  they arc  neutral  to  the  effect of 
freight costs  unlike  free  on board  prices.  Subsequently, hy  
pothesis [3],  i.e., the  simultaneous  LOP,  can be  tested  with  the  
following restriction  on the  P  matrix  assuming that  two  coin  
tegration vectors  exist in  the data: 
where  H  is  a design  matrix  giving the  restriction  on  the  P  
matrix  and  (p  is  a (2  x  2)  matrix  of  the  parameters reduced  by  
means of H.  If restriction [6]  is  rejected, the  law  is  also  rejected 
for  all  prices  simultaneously,  in which  case the  second  testable  
hypothesis is  that  the LOP holds  between  any  two prices.  
In  the  market where  the  LOP  holds, some prices may  still  
be  driving forces  behind  the  other  prices.  Finally, we test  
whether  any  of the three  newsprint import prices in  the United  
Kingdom or Germany  has  such  a  property. This  can  be  tested 
by  the  weak  cxogcncity  test  (e.g., Silvapullc and  Jayasuriya 
1994). The  weak  exogeneity  can  be  tested  by  imposing restric  
tions  on  the  loading  vectors  of  the  variables  to  equal  zero in  
the  cointcgration space.  If this  restriction  is accepted  for  a 
variable,  it  is weakly exogenous  and  consequently does  not 
adjust  to deviations  from any  equilibrium relation, as defined  
by  the cointcgration relation  (Johansen  1992). 
Data 
The  data  used  in  this  paper  consist  of  unit  prices  of  newsprint 
imported from Finland, Sweden, and Canada  to the United  
Kingdom  and  Germany.  The  data  arc  quarterly, seasonally  un  
adjusted,  and cover the  period from  1980  to  1994.
3
 The  import 
prices  of  newsprint  from  Finland,  Sweden, and  Canada  to  the  
importing countries  were described  by  the  average  import  unit  
values  (£/ton  for  United  Kingdom  and  DEM/ton  for  Germany)  
based  on CIF figures. Data  on quantities and  values  of  news  
print (SITC 641.1) imports to  the  United  Kingdom were taken  
from the  Overseas  Trade  Statistics  of  the  United Kingdom  
(CSO).  The  newsprint import quantities and  values  to Ger  
many  (SITC 4801 0010, including item 4801  650  for year 
1980,  item  480 1 790  for years  1981-1987, and  item  4801  0090  
for 1988 onwards) were obtained  from Aussenhandel  nach  
Waren und  Ländern  (Spezialhandel). The  unit  price series  
were  deflated  by  producer  price  indices  (1985 = 100) for  the  
United  Kingdom and  Germany, obtained  from the Main Eco  
nomic  Indicators  (OECD).  
Average import  unit  values  were used  to  describe  prices  
because  more  disaggregated data  on the  quality composition 
of  each  supplier's  newsprint  were not  available.  This  means 
that  newsprint  imported from  the  three suppliers  was  assumed  
to be  homogenous enough to allow  reliable  testing of  the LOP 
using unit  values. Relatively  speaking,  newsprint is  the  most  
homogenous grade among  the  different  paper  products  and  is  
used  for  similar  end  uses,  mainly for  printing newspapers  and  
less  demanding advertising publications. For  the  three  supplier 
countries  at  hand, the  homogeneity assumption  was  also  justi  
fied, because  these  countries  have  similar  fibre  raw  material  
basis,  i.e., coniferous  wood.4 
Empirical  results  
The  three  supplier  countries  have  accounted  for  about  85% of 
newsprint  imports  to  the  United  Kingdom  and  68%  to  Ger  
many  during the  period studied.  Their  average  import  shares  
are shown  in  Table  I. It  can  be  seen that the combined  import 
shares of these  three  countries  have  slightly fallen  during the 
period, both  in  the  United  Kingdom and  Germany. However, 
loss  of  import share  has  concerned only Finland  and Canada.  
Sweden  has  managed to maintain  its  share  in  Germany and  has  
3
 In principle it  would be  possible to obtain monthly price  data. 
However,  considering  newsprint markets, quarterly frequency 
seems more  suitable, since  newsprint trade in Europe is  often 
based  on long-term contracts (see  Andersen 1993). 
4
 Possible  consequences  due to differences  in quality composition 
between  different exporters  or the  effect  of  different ports  of  
entry  and exit,  data errors,  etc.  cannot be  taken into  account, but  
they  need to  be borne  in mind when interpreting  empirical  results.  
© 1997 NRC  Canada 
[s] At,  =r  i  Ax,_,  + ...  +  rt_,AtM+l  +  Ilt,_t  +n  +  00,  +e„  
I=l r 
1 f 
[6] P  =  //<p= -1 ocp  
0 -1  
Period 
United Kingdom Germany  
Finland Sweden Canada Finland Sweden Canada 
1980-1987 35 20 33 24 36 II 
1988-1994 20 32 29 . 19 36 10 
1980-1994 28 26 31 22 36 10 
Hänninen et ai. 1949  
Fig. 1. Real unit values of newsprint imports (in logs) from 
Finland,  Sweden, and Canada to the  United Kingdom and 
Germany, 1980 (first quarter)  -  1994 (fourth quarter). 
strongly increased  it in  the United  Kingdom. In Germany, 
Sweden  has  been  clearly the  most  significant exporter  during 
the  whole  period. Canada's  average  share  has  been  only  about  
10% of total German  imports.  Germany is far  less  dependent 
on imported newsprint  than  the  United  Kingdom. The  ratio  of 
domestic  newsprint production to imported newsprint has  been  
around  90%  in  Germany and  about 40%  in  the  United  Kingdom. 
Graphs of  newsprint  prices  are presented (real price  series  
in  logarithms) in  Fig.  1.  After  stable  price  development during 
the 1980 s,  there  has  been  a  sharp fall  in  all  newsprint prices  in  
both  importing countries  because  of  economic  recession  in  the  
early 19905. In relative  terms,  newsprint imported from Sweden 
has  had the highest, and the newsprint imported from Canada, 
the  lowest  average  unit  price  in  both  the  United  Kingdom and  
Germany during the  period studied.  
Visual  inspection of  the  graphs of the  newsprint series  
(Fig. 1) suggests nonstationary prices. The test  results  from  
Augmented Dickey-Fuller (ADF)  unit  root  test  (Dickey  and 
Fuller  1979) with  one lag, presented in  Table 2,  indicate  that 
Table  2. Augmented  Dickey-Fuller (ADF) 
unit root  test  results  for  prices  of newsprint 
with  one lag.  
all  the  levels of prices are nonstationary /(1) processes,  while  
the  first  differences  of  prices  arc  clearly  stationary. ADF-tcst 
statistics  were estimated from a  model  that includes  a constant 
and  a trend  variable.  We performed the  test  with  1 to  5 lags, 
and the  results  were found  to be  invariant  to the  use  of different 
number  of lags. 
Hypothesis  [3]  was tested  by determining the cointegration 
rank,  i.e., the  number  of cointegration vectors,  in  both  destina  
tion  countries.  To  define  the  proper  lag length for  the cointe  
gration estimation, model  system [s]  with alternative  number  
of  lags  was fitted  to  the price data.  Diagnostic  tests  for the  
residuals  of  each  equation and  the  corresponding vector test  
statistics  supported the  model  with  four  lags  (k  =4)as  a suffi  
cient  description of  the  data (Table 3).  However,  null  hypothe  
sis  of normality was rejected for  the  residuals  of all  price  
equations for  the  United  Kingdom  and  also  for  Finland's  price  
equation in  Germany at  the 1% level.  According to Gonzalo  
(1994),  non-normality of  residuals  does  not bias  the  results  for  
Johansen's  cointegration tests.  Moreover, non-normality was 
mainly caused  by  outliers  in  the original data in  U.K. customs  
statistics.  
The  cointegration test  results  from  model  [s]  with four  lags 
indicated  one cointegration vector  in  the U.K.  data, while  no  
cointegration was found  in  the German  data.  The  test  results  
are presented with  an unrestricted  constant,  allowing for  linear  
trend in  the  data  (sec  Fig.  1). According (o  the  Johanscn  rank  
test, the  hypothesis  of no  cointegration (;•  = 0)  could  be  re  
jected and one cointegration vector  was accepted  at the  5%  
level  in  the United  Kingdom (Table 4). In the case  of the  Ger  
man price  data,  all  the  test  statistics were  well  below  the  critical  
values  and  ;•  = 0 could  be  rejected only at  50%  level  (cf.  Jo  
hansen and  Juselius  1990; critical  value  20.19). 
The  three  normalized  eigenvectors  ((3)  with weights (a)  ob  
tained  form the  cointegration estimation  of  model  [s]  for both  
destination  countries  arc shown in Table 5. From the three 
vectors  of  the  United  Kingdom the first  relation, p h is  inter  
preted as the  cointegration vector,  because  the  first  relation  is  
always most  highly correlated  with  the stationary part  of  the 
process  Ax,  when  corrected  for the  lagged values  of the differ  
ences. The normalized  a-coefficients  are presented on the 
right-hand side  of  the  long-term P-coefficients  in  Table  5.  The  
a's can be  interpreted as adjustment coefficients  that  describe  
O 1997 NRC  Canada 
Price  by  
ADF  test  values  by  
destination country  
supplier country  U.K. Germany 
Levels 
Sweden,  ps  -0.46 -0.48 
Finland, pf  -1.32 -1.36 
Canada,  pc  -2.22 -1.77 
First  differences  
Sweden,  Aps  -7.21* -5.73* 
Finland, Apf -9.05* -5.50* 
Canada,  Ape -4.83* -7.94* 
Note: the  rejection of the null  
hypothesis  of nonstationarity. The critical /-value 
is  -3.49 at the 5% level. 
1950 Can. J. For. Res. Vol. 27, 1997 
Table 3.  Misspecification tests  for  residuals  from  Johansen's  cointegration estimation  of 
model  [s]  with  four  lags. 
Table 4. The result  of the  cointegration rank  test.  
the average speed of adjustment of individual  equations to  
ward  the estimated equilibrium state (sec  Johansen  and  
Juselius  1990). A low  coefficient  indicates  slow  adjustment, 
while  a high coefficient  indicates  rapid adjustment. For exam  
ple,  the  coefficients of  the  a,-vector  represent  the  weights with  
which  Sweden's  newsprint price  enters  the price  equations of 
Finland and Canada. 
The  results  of  the  rank  test  in  Table  4  imply  that hypothe  
sis [3]  of  simultaneous  holding of  the  LOP  can be  rejected for  
both  destination  countries,  because  r< 2. As there  was no 
cointegration in the German market,  the simultaneous  long-run 
LOP cannot hold.  In the U.K.  market, the simultaneous  LOP  
between three  prices,  i.e., hypothesis [3] from  eq. 6, was re  
jected. The  value  of  the  X
2
(2)  test statistic was 8.68, while  the  
5%  critical  level  is  5.99.  The  rejection  of  hypothesis  [3]  indi  
cates  that  the  LOP cannot be  valid for  all  three prices  simulta  
neously. However,  a stationary  relationship between  any two 
price variables  is  still  possible in  the  United  Kingdom. This  is  
Tabic  5. Normalized eigenvectors, β1. and their weights, (α1,  
obtained  from cointcgration estimation  of model  [s].  
why  the  law  was tested  in  the  next  phase also  for  the  pairs of 
prices  under  one cointegrating vector  indicated  by  the  trace  
test.  
For a  pairwise  testing of  the  LOP,  the  coefficients  of  the  
cointegration vector,  |3h  were restricted. The  long-term equi  
librium  relation, (3,, normalized  to  Swedish  price  can be  pre  
sented in vector form as  
where  1.00 is  the  coefficient  of Swedish  price,  -0.22  is  the  
coefficient  for  the  Finnish  price,  and  -0.96  is  the  coefficient  for 
the  Canadian  price.  The  coefficients  indicate  that  the  LOP may  
hold  between  Sweden  and  Canada,  but  not between  Finland  
and  the others. The  formal test  was performed by  restricting 
two price coefficients  at a  time to equal 1 and  -1 and  the  third  
to  equal  zero in  (3,  vector in  [7].  Test  results for  the  United  
Kingdom under  one cointegrating vector,  presented in  Table 6,  
© 1997 NRC Canada 
[7] p, = [l.OO -0.22  -0.96] 
Tests  for  residuals  and standard errors  
Autocorrelation0 Heteroskedasticity''  Normality® Standard  error  
Equation (FAR (4.36))  WW»,32)) (X
2
(2)) (c
e
) 
United Kingdom 
Apsk  0.80[0.53] 0.03 [0.99] 27.86[0.00]* 0.07 
Apfk 0.54[0.71] 1.30[0.29] 36.91 [0.00]* 0.08 
Apck  1.48[0.23] 0.10[0.98] 16.67[0.00]* 0.05 
System VFar (36,77)=0.76[0.82] Vx
2
(6)=89.49[0.00]* 
Germany 
Apsd  0.91 [0.47]  0.42[0.79] 0.95[0.62] 0.02 
Apfd 0.33[0.86] 0.12[0.98] 9.66[0.01]* 0.04  
Aped  2.05[0. 11] 1.27[0.30] 5.02[0.82] 0.07 
System VFAR(36,77)  = 1.80[0.02] Vx
2
(6)= 16.98[0.01]*  
Note: A denotes the first  differences of the variables,  psk,  pfk, and pek  denote the newsprint  prices of 
Sweden, Finland, and Canada  in the U.K. market,  and  psd, pfd, and ped are  the corresponding  prices in  
the German market.  Values  in square brackets  arc  marginal  significance  levels  and 
*
 indicates that the 
null hypothesis  is rejected  at the 1% level.  
"Autocorrelation of the residuals  was examined using  the F-form of the Lagrange-Multiplicr  (LM)  test, 
which is  valid  for  systems  with lagged  dependent  variables. 
''Heteroskedasticity  was  tested  using  the F-form  of  the LM test against  fourth-order  autoregrcssive  
conditional heteroskedasticity. 
c
Normality  of the  residuals  was  tested with the Doornik-Hanscn  test (Doomik  and Hendry  1994). For 
further detail and test references,  see  Doornik and Hendry  (1994).  
Null hypothesis Eigenvalues X trace  95% confidence 
H
0 :  r£i  (X,) statistics  (7X0) values (C(i)) 
United Kingdom 
r= 0 0.35 35.33* 29.70 
r<\ 0.18 11.40 15.40 
r<  2 0.00 0.02 3.80 
Germany  
r= 0 0.28 23.09 29.70 
r< 1 0.07 4.40 15.40 
r<  2 0.00 0.16 3.80 
Note: rejection  of the  hypothesis  r -  0. 
Prices  by  
Eigenvectors Weights 
countries p, P2 P3 a, a2 «3 
United Kingdom 
Sweden, psk  1.00 -0.51 -9.94 -0.23 0.39  0.00 
Finland,  pfk  -0.22 1.00 2.92 0.24 -0.35 0.00 
Canada,  pek  -0.96 -0.32 1.00 0.56 0.21  0.00 
Germany 
Sweden, psd  1.00 -0.21 -2.29 -0.03 -0.05 -0.01 
Finland, pfd -1.40 1.00 1.02 0.10 -0.05 -0.01 
Canada,  ped -0.20 -0.99 1.00 -0.03 0.07 -0.04 
Hänninen et ai. 1951 
Table 6. Test  results  for  the LOP  by  pairs  of  prices  
obtained from the  U.K. model by  restricting (3, under 
r  = 1. 
rejected the  law  for  the  price  pairs  for  Sweden  and  Finland  
(psk,  pfk)  and  Finland  and  Canada  (pfk,  pek).  Only  the  restric  
tion  [l,O,  -1 ] was not rejected, which  means that  for  the  prices 
of Sweden  and  Canada (psk,  pek), the  test  supported the LOP. 
The  results  for the  weak  exogeneity test  in  Table  7  reveal  
that  the Swedish  and Finnish  newsprint prices  are weakly ex  
ogenous  to  the  long-run parameters in  the  U.K. market.  Thus, 
Swedish  and  Finnish  prices  do  not  adjust to  deviations  from  
any  equilibrium state  defined by  the  cointcgration relation, but  
are rather  determined  outside  the  model  system.  For the  Cana  
dian  price, weak  exogeneity was rejected, suggesting that  Ca  
nadian  price  is affected by  the  Swedish  and  Finnish  prices  in 
the U.K. market. Furthermore, the existence  of  the LOP be  
tween Swedish  and  Canadian  prices in  United  Kingdom  can 
be  interpreted so that  Canadian  price  is  the  adjusting part  and  
the  Swedish  price is  the  determining part  of  the  relationship. 
The  result  of the weak exogeneity tests  could  also  reflect  the 
fact  that Scandinavian  producers  are dominating the  European 
newsprint market, while  Canadian  imports to the area are  the  
residual  of the market demand  (see also  Brännlund  and  
Löfgren (1995) on the results  of testing for  cyclical dumping 
in  the newsprint markets). 
Summary and  conclusions 
In this paper  the  long-run LOP  was tested  for  U.K. and German 
newsprint  imports from  Finland, Sweden, and  Canada  by  using 
the  theory of cointegration (Hngle and  Granger 1987). Johan  
sen's  (1995) multivariate  cointegration method  was applied in  
the  empirical  analysis.  The  test  results  were calculated  using 
quarterly  import unit  values  deflated  by  producer  prices  for  the  
period 1980-1994  (r=  60). 
The  cointegration tests  indicate  that  the  LOP  is not  valid  in  
either market.  The  results  supported one cointegration  vector  
in  the U.K. data, while  in  the  German  data  no cointegration was 
found. Also  the  results  of  the  pairwise  tests  in  the  U.K.  market 
indicated  rejection  of the  LOP  except  for  a  pair  of  Swedish  and  
Canadian  prices.  Differences in  cointcgration rank  test  results  
between  the  two  destination  countries  may  be  due  to  fact that  
Germany  is  less  dependent on the  imported  newsprint than  the  
United  Kingdom. Therefore  the price of  their  own newsprint 
production in  Germany, which  was  not  available,  may  affect 
the  price  structure  of  the  German  market.  However, the  exclu  
sion  of  German  price  can be  justified  by the  fact  that newsprint 
production in  Germany is  mainly based on  the  use of  recycled  
fibre  (Andersen 1993). Therefore  its  price  is not perfectly com  
Table 7. Test  results  for  weak  exogeneity  of 
supplier countries' newsprint  prices  in the 
United Kingdom. 
parable with the  prices  of the  mainly  fibre-based  newsprint 
imported from Sweden, Finland, and  Canada.  
The  results  of  this  study  differ  from  the  traditional  assump  
tion  of  trade  theory, i.e., that  commodity prices  are perfectly  
arbitraged (e.g., Buongiorno and  Gilless  1984). They also  dif  
fer  from the test  results  of Buongiorno and  Uusivuori  (1992),  
who  accepted  the  LOP  for  U.S.  paper  exports across destina  
tions.  The  results  do agree, for example, with  the  results  of 
Isard  (1977), who  compared export  prices  of different  products 
between  the United  States  and  Germany; Knetter  (1993), who  
found  price  differences  between  exporters  for  agricultural prod  
ucts;  and  Hänninen  (1998), who  tested  the  law  for  U.K. sawn  
wood  imports across the main  supplier  countries.  
Newsprint  is a relatively  homogenous product, so the  arbi  
trage  could  be  expected to hold.  However,  our  results  indicate  
long-run deviations  of  the  LOP  and  support imperfect compe  
tition  in  the  newsprint markets.  Thus,  the  models  allowing for 
price  differences  between  the  main  supplier  countries  can  be  
used  in  modeling the  U.K.  and  German  newsprint  markets.  The  
results  of  the  study  also  stress  the  importance of  formal  testing 
for  competition in  forest  products  markets, which  is  often  
missing in  the studies of forest  products  trade.  
Acknowledgements 
We thank  Heikki  Kauppi, Jari  Kuuluvainen, Juuso  Vataja, and 
the  referees of the  Journal  for helpful comments  on the earlier 
versions  of this  paper. 
References  
Adams,  D.M.,  and Hayncs,  R.W. 1980.  The 1980 softwood timber  
assessment  market  model:  structure,  projections and policy  simu  
lations. Monogr. 22. F0r.  Sci.  Suppl. 26(3): 1-65. 
Andersen,  S. 1993. Det curopciske Avispapicrmarkedenet—en struk  
tur och utviklingsanalyse  (The European newsprint  market— a 
structure and development analysis).  Doctoral  thesis,  Agricultural 
University of  Norway,  As. 
Ardeni,  G.P. 1989.  Docs  the  law  of one  price  really  hold  for  commod  
ity  prices?  Am. J.  Agr.  Econ.  71: 661-669. 
Armington, P.S. 1969. A theory of demand  for products  distinguished 
by  place  ofproduction. Int. Monet. Fund  Staff  Pap. 6(1): 159-176. 
Ausscnhandcl  nach Warcn und Ländcm. 1980-1994. Statistichcs  
Bundesamt. Organization for  Economic Co-operation and Devel  
opment,  Wiesbaden. 
Baffcs, J. 1991. Some further  evidence on the law of one price: the 
law of  one price  still holds.  Am. J.  Agric.  Econ.  73:  1264-1273. 
Bancrjcc, A., Dolado, J., Galbraith,  J., and Hendry,  D. 1993. Co  
integration, error-correction,  and the  econometric  analysis  of non  
stationary  data. Oxford University  Press,  Oxford. 
O 1997 NRC Canada 
Testable 
prices  
Hypothesis Test  
(//(,: Pjf=Pn) (LR  test  statistic") 
psk,  pfk  
psk,  pek  
pfk,  pek  
Pn=i*  P21—1» P31-0 16.68* 
Pi  
1=1»
 p2i=o>  p3i=-i  
Pi  1=0» p21=l» Pjl=-I 9.88* 
Note: psk,  pfk and pek  are the  newsprint  prices  of 
Sweden, Finland and Canada in the UK. 
"Critical  x
2
 value  is  5.99  at  the  5%  level, the 
rejection of the law of one price.  
Testable Test  (under r=l)  a=0 
price  (LR  test statistic") 
Sweden,  psk  1.32 
Finland,  pfk  1.30 
Canada,  pek 9.71* 
"Critical  x 2  value 3.84  at  the  5%  level.  
•denotes the rejection of weak cxogencity.  
1952  Can.  J. For. Res. Vol. 27,  1997 
Blatner,  K. A. 1989.  An approach to the estimation  of import price  
elasticities by  supplier. For.  Sci.  35(1): 30-41. 
Booth, D.L., Kanctkar,  V., Vcrtinsky,  1., and Whistler,  D. 1991. An 
empirical model  of capacity expansion and pricing in an oligopoly 
with barometric price  leadership: a case  study of newsprint indus  
try  in North  America. J.  Ind.  Econ.  39(3): 255-276. 
Brännlund,  R.,  and Löfgren,  K.G.  1995. Cyclical  dumping and corre  
lated  business  cycles in imperfect markets: empirical applications 
to  the Canadian pulp and paper industry. Appl. Econ.  
27:  1081-1091. 
Brännlund,  R.,  Lundberg, L.,  and Wiberg, A. 1982.  Determinants of 
Swedish market  shares  for  pulp, paper and sawnwood in Europe. 
Department  of Forest  Economics,  Swedish University  of  Agricul  
tural  Sciences,  Umeä. Rep. 36. 
Buongiorno, J., and Gilless,  J.K. 1984. A model  of  international trade 
of  forest  products,  with an  application to newsprint.  J.  World For.  
Res.  Manag.  1: 65-80. 
Buongiorno, J., and Uusivuori,  J.  1992. The law of one price  in the 
trade of forest  products: co-intcgration tests  for  U.S. exports  of 
pulp  and paper.  For.  Sci.  38(3): 539-553. 
Chou,  J-J., and  Buongiorno, J.  1983.  United States  demand for  hard  
wood plywood imports by country of origin. For. Sci.  
29(2): 225-237. 
CSO.  1980-1994. Overseas  trade  statistics  of  the  United Kingdom, 
CSO,  London. 
Dagcnais,  M. 1976.  The determination of newsprint prices.  Can. J.  
Econ. IX: 442-461. 
Dickey,  D.A., and Fuller,  W.A. 1979. Distribution of  estimates for  
autoregrcssivc  time scries  with unit root. J. Am. Stat.  Assoc.  
74: 427-431.  
Doornik,  J.A., and Hendry, D.F. 1994. PcFiml 8.0. Interactive  
econometric  modelling of dynamic systems. Institute of Econom  
ics and Statistics, University  of Oxford. International  Thomson  
Publishing, London. 
Dykstra,  D.P., and Kallio, M.  1987.  Introduction to the  II AS A Forest  
Sector  Model. In The global forest  sector:  an analytical perspec  
tive. Edited  by M. Kallio,  D.P. Dykstra,  and  C.S. Binkley. John 
Wiley & Sons,  Chichester,  pp.  139-167. 
Englc, R.F., and Granger, C.W.J.  1987. Co-integration and error  cor  
rection: representation, estimation and testing. Econometrica, 
55(2):  251-276. 
Gonzalo, J.  1994. Five  alternative methods of estimating long-run 
equilibrium relationships. J. Econom.  60: 203-233. 
Goodwin, 8.K., and  Grennes,  T.J.  1994. Real  interest  rate  equaliza  
tion and the  integration of international financial markets.  J.  Int. 
Money  and Fin. 13: 107-124. 
Hänninen,  R. 1994. Price  elasticities  of substitution between Finland 
and the other  supplier countries in the demand  for  the United 
Kingdom sawnwood imports.  In Proceedings of the Biennial 
meeting of  the  Scandinavian forest  economics,  22-25 Nov.  1994,  
Gillcleje, Denmark.  Edited by  Finn Helles and Michael Linddal.  
Department  of Economics  and Natural Resources,  Royal  Veteri  
nary  and Agricultural University,  Copenhagen, Denmark.  Scand.  
For. Econ. 35: 204-217. 
Hänninen,  R.  1998.  The law of one price  in the  United Kingdom soft  
sawnwood  imports—a cointcgration approach.  For.  Sci.  44(1). 
In press.  
Hseu,  J-S., and Buongiorno, J.  1993.  Price  elasticities of substitution 
between  species  in  the demand of U.S. softwood lumber imports 
from Canada. Can. J. For. Res.  23: 591-597. 
Isard, P. 1977. How far  can  wc push the "law of one price"? Am. 
Econ.  Rev.  67(5): 942-948. 
Johansen, S. 1988.  Statistical analysis of cointcgration vectors. J.  
Econ. Dynam.Control, 12: 231-254. 
Johansen, S. 1992. Cointcgration in partial systems and the  efficiency 
of single equation analysis.  J.  Econom.  52: 389-402. 
Johansen, S. 1995. Likelihood based inference  in cointcgrated vector 
autoregrcssivc  models. Oxford  University  Press,  Oxford.  
Johansen, S., and  Juselius,  K. 1990. Maximum likelihood estimation 
and inference  on  cointcgration
— with applications to the  demand 
for money.  Oxford  Bull. Econ.  Stat.  52:  169^-210. 
Johansen, S., and Juselius,  K. 1992.  Structural  tests  in a multivariate 
cointegration analysis  of the PPP  and the UIP for  UK. J.  Econom.  
53:211-244. 
Jung, C., and Doroodian,  K. 1994. The law of one price  for  U.S. 
softwood lumber: a multivariate cointcgration test. For. Sci.  
40(5): 595-600. 
Knetter,  M.M. 1993.  International  comparisons of pricing-to-inarkct  
behavior. Am. Econ.  Rev.  83(3):  473-486. 
Laaksonen,  S., Toppincn, A.,  Hänninen,  R., and Kuuluvainen,  J. 
1997.  Cointcgration in Finnish paper  exports to the United King  
dom. J.  For.  Econ. 3(2): 171-185. 
OECD. 1962-1991,  1993. Main economic  indicators,  historical sta  
tistics. Organization for  Economic Co-operation and Develop  
ment, Paris.  
Officer, L.H.  1986. The law of one price  cannot be  rejected: two tests  
based on the tradable/nontradablc price ratio. J. Macroccon.  
8: 159-182. 
Silvapullc, P., and Jayasuriya,  S. 1994. Testing for  Philippines rice  
market integration: a multiple cointcgration approach. J.  Agric. 
Econ. 45: 369-380.  
Stock,  J.H.  1987. Asymptotic  properties of least  squares  estimators  of 
cointcgration vectors. Econometrica, 55: 1035-1056. 
O 1997 NRC  Canada 

IV 

Silva Fennica  32(1)  research  articles  
61  
Exchange  Rate  Changes  and the 
Finnish  Sawn  wood  Demand and Price  
in the UK Market  
Riitta  H. Hänninen  
Hänninen, R.H. 1998.  Exchange rate  changes  and  the  Finnish  sawnwood  demand  and  
price in the  UK  market. Silva  Fennica  32(1):  61-73. 
This  paper  examines  the  long-run influence  of exchange rate  changes on the  Finnish  
sawnwood  price  in the  United Kingdom (UK) using quarterly data for  the  period 1978 
1994.  The degree of  influence was  measured by  a  pass-through coefficient (PT)  obtained 
from  a markup pricing  relation  of a system  model.  The  model, which included export  
demand and price equations, was  estimated  with the  cointegration method  of  Johansen. 
The  results  indicated  a large PT, which means that  exchange rate  changes are reflected  
almost proportionately  in  Finnish export  price  expressed  in  pounds  sterling.  Thus, the 
Finnish  price  of sawnwood  in  pounds has  lowered  as a result  of depreciation of the  
Finnish markka  (FIM).  This  has improved Finnish  competitiveness  and  market share in  
the  UK.  Appreciation of the  FIM has  had  the opposite effect. It seems  that Finnish  
exporters  have made use  of depreciations and  devaluations of the  FIM to maintain  and  
increase  their  market  shares  but  not  necessarily  their  markups. For  Finland,  which is in 
the  process  of joining  the  European economic and  monetary union  (EMU),  knowing the  
size of the  PT is  also  important in assessing  the economic impact  of membership. 
Keywords  exchange rate,  export  demand, Johansen's  cointegration  method, pass-through,  
sawnwood  price,  United Kingdom 
Author's  address  Forest  Research  Institute, Helsinki  Research Centre, Unioninkatu 40 A,  
00170  Helsinki, Finland  Phone +358  9  8570 5746  E-mail riitta.hanninen@metla.fi  
Received 22  May 1997 Accepted  22  December  1997 
1 Introduction  
This study  examines  the effects  of exchange  rate  
changes  on  Finland's  sawnwood exports  to the 
United Kingdom (UK),  which is  a  major  market 
for Finnish  sawnwood. In 1996 the UK account-  
Ed  for 18 percent of Finland's sawnwood  ex  
ports  (Facts and Figures  1997). Exchange  rate  
effects  are analysed  by a demand and supply 
model for Finnish sawnwood exports.  The focus 
of the  study is  on the exchange  rate  pass-through  
(PT),  i.e. the percentage change  in export  prices  
Silva  Fennica 32(1)  research  articles  
62  
associated with  a one  percent  change  in the ex  
change  rate.  
If the exchange  rate has a small effect  on for  
eign  currency  prices,  currency realignments  
should not  have much effect  on  exports.  Previous 
results are scanty  for forest  products,  but they  
indicate that  trade flows of  forest products  (meas  
ured in quantities)  have not  been very  sensitive to 
exchange  rate  variations (Buongiorno  et al.  1988).  
Also  Uusivuori and Buongiorno  (1990)  found no 
long-run  relationship  between exchange  rates  and 
quantities  of paper exports  from Finland and 
Sweden to the USA.  For Finland,  which is  in the 
process  of joining  the EMU, the magnitude  of  the 
exchange  rate  effect  is  important in assessing  the 
economic impact  of  membership.  In the EMU,  it 
will  no  longer  be  possible  to  improve  competitive  
ness  by  manipulating  the exchange  rate.  Estimates 
of the degree  of  pass-through  are  also informative 
as  to market  competition,  which is  important  in 
modeling  trade  flows. 
The  pass-through  concept can be defined as  
the extent  to  which a change  in a  country's  nom  
inal  exchange  rate  induces a price change  in 
terms  of  the foreign  currency. The magnitude  of  
the pass-through  also reflects  the extent  of mar  
ket  competition.  For  example,  in  perfect  compe  
tition, export  prices  in foreign  currency do not 
change  as  a  result of  a  devaluation (revaluation)  
of the exporter's  currency. In this case,  Finnish 
exporters'  prices  in  FIM would increase (de  
crease),  their markup  margins would increase 
(decrease)  and the PT would be zero. On the 
other  hand, in imperfect  competition  Finnish ex  
porters  could change  their foreign  currency pric  
es  when  the exchange  rate  changes.  If  exchange  
rate  fluctuations were  partly  reflected in foreign 
currency prices and partly  in FIM prices,  PT 
would be between zero  and unity. If PT  is  unity, 
Finnish  exporters lower  (raise)  their foreign  cur  
rency  export  prices  pro rata  to  a devaluation 
(revaluation) of the  FIM 
Pass-through  has not  been extensively  studied 
for  forest  products  trade, although  it has been 
analyzed  for trade in many other commodities 
and for aggregate imports  and exports (see  e.g. 
the survey  Menon 1995). PT is found to  vary 
widely  by industry  (e.g. Feenstra et al. 1996), 
which means that it  is  necessary  to study  pass  
through  at the commodity  level. Most of the few 
studies analyzing  the PT for forest products deal 
with the USA.  For example,  there are  very  few 
results  for Finnish exports.  In the study  of  Vesa  
la (1992) PT  for  export prices  of  Finnish paper 
manufactures was  found to  be between 0.66 and 
0.69 for western  Europe  and between  0.16 and 
0.30 for the USA.  According  to Vesala, the small  
er PT for the USA is  due to  the large  US domes  
tic market,  in which the Finnish share  is  much 
smaller than in Europe.  Other studies  have also 
found that prices  of several  commodities import  
ed to the USA  only  slightly  reflect changes  in 
exchange  rates  (e.g.  Yang  1997). 
Uusivuori and Buongiorno  (1991)  estimated 
PT  for  US  forest  products  exports to  Europe  and 
Japan.  Pass-through  was incomplete  in most  of 
the product  categories.  However,  in lumber ex  
ports  to Japan,  PT was  high:  from 0.79 to  1.04, 
depending  on  the species.  Also in the study  of  
Menon (1993  a),  who estimated PTs  for  Austral  
ian imports,  wood products  had a relatively  high 
PT (0.80),  while for  paper and board the PT was 
0.45. The earlier results indicate that PT is  high  
er for  wood industry  products  than for paper 
products.  However, exact  conclusions about the 
size  of the PT  for  forest  products  are  difficult to  
draw, because  studies  are  scarce  and they con  
cern  only  a  few products  and countries. 
Based on the earlier  models for Finnish sawn  
wood exports  to the UK (e.g.  Hänninen 1986, 
Tervo et al. 1988, Hänninen 1994) and studies 
that test  arbitrage  in UK  sawnwood imports  (Hän  
ninen 1998), it can  be assumed that the PT is 
relatively  large.  The assumption  is  also  support  
ed  by  the structure  of  UK  sawnwood imports,  
which is  dominated by  four  large  supplier  coun  
tries,  i.e. Finland, Sweden,  Canada and Russia.  
The analysis  of the present  study  is based on  a 
structural multivariate model formulated for Finn  
ish sawnwood exports  to the UK. The  model 
consists  of an export  demand equation  and a 
markup  price  equation.  In earlier studies simul  
taneous  multivariate models have not  been ap  
plied in analyzing  pass-through  in forest  prod  
ucts markets and they  have rarely  been applied  
in other commodities  markets (see  however 
Rockerbie 1992, Menon 1993 a,  1993b, Kong  
sted 1996).  Previously  used models have usually  
been bivariate, with a price variable being  re  
gressed  on an exchange  rate  (e.g.  Knetter 1989,  
Hänninen Exchange  Rate  Changes  and  the  Finnish  Sawnwood Demand and Price  in  the  UK  Market  
63 
1993, Pick  and Park 1991, Uusivuori and Buon  
giorno 1991). 
Multivariate approaches  based on the markup  
concept have  also been applied,  but these appli  
cations have  been single  equation  models (e.g.  
Dornbusch  1987, Hooper  and Mann 1989, Athu  
kolara 1991, Athukolara and Menon 1994, 1995, 
Hung  et al. 1993). The  present study  uses  the 
Johansen (Johansen 1988, Johansen and Juselius 
1990 and  1992)  multivariate cointegration  meth  
od for nonstationary  data  for estimation purpos  
es. This allows for the estimation of a model 
system,  unlike the traditional Engle  and Granger  
(1987)  procedure.  
The findings  indicate that PT is  large, which 
means that in the long-run  changes  in the ex  
change  rate  (FIM/pound  sterling) are  almost com  
pletely  reflected in Finnish  prices  as measured in 
pounds.  Thus, for  example, depreciations  and 
devaluations of the FIM have improved  Finnish 
competitiveness  and market share by  lowering 
the relative Finnish price in the UK market. 
2 Model  
In the present  study, a  theoretical model is  formed 
to describe Finnish sawnwood exports  to the 
UK.  The  model includes a demand equation  and 
a price equation.  An estimate for the exchange  
rate  pass-through  (PT)  is  obtained from the price  
equation.  The  same type of model construction 
has earlier been applied  by  Kongsted  (1996) in 
modeling  Danish manufacturing  exports.  
The model is  constructed on  the assumption  
that competition  between supplier  countries is 
imperfect  in UK sawnwood imports.  Thus,  the 
elasticity  of substitution model describing  de  
mand for goods  from different origins  is  used to 
describe the export  demand for Finnish sawn  
wood. The derivation of the export  demand as  
sumes  a two-stage optimization  of a  representa  
tive sawnwood importer  in the UK. First,  costs 
are minimized subject  to  the expenditure  on a 
good (e.g.  sawnwood)  and,  secondly,  this ex  
penditure  is  allocated optimally  between the prod  
ucts  from different countries of  origin. Thus, the 
export  demand for Finnish sawnwood can be 
represented  as 
where  Xf  and Pf are  the Finnish quantity  (1000  
m  3) and nominal unit  price  of  sawnwood exports  
(FIM/m3 ) to the UK market,  X
a
 and P
0
 are the 
respective  quantity  and  price (£/m
3
)  of  competi  
tors' sawnwood, ER is the nominal exchange  
rate  (FIM/£),  bf  is  a  constant  and  rj  is  the elastic  
ity of substitution (assumed  to be constant).  Af  
ter  logarithmic  transformation of  the variables in 
(1),  the export  demand relation becomes 
where lower-case letters denote logarithmic  val  
ues  of the corresponding  upper-case letters in 
equation  (1),  c is a constant  term and £ is a 
disturbance term. The symbol  rj  is  the price elas  
ticity of demand and equation  (2) is  homoge  
nous  of degree  zero  in the nominal variables. In 
this elasticity  of substitution model (2),  the change  
in the relative  quantities  demanded is  assumed 
to  be proportional  to the change  in the relative 
price  of  exports.  
An equation  for Finnish sawnwood  price is 
based on  a markup  model that has earlier been 
applied  in the estimation of pass-through  e.g. by  
Dornbusch (1987),  Hooper  and Mann (1989),  
Athukolara (1991), Athukolara and Menon 
(1995),  Hung  et al. (1993).  In deriving  the price 
equation,  it is  assumed that a  representative  Finn  
ish exporter firm produces  exclusively  for an 
imperfectly  competitive  UK market,  employs  
constant-returns-to-scale technology  and unit pro  
duction cost,  Cf.  The firm maximizes profit  by  
taking  the competitors'  price  and the supply  of 
competitors'  sawnwood as  given  and by  setting  
the price  in FIM,  Pf,  as a  constant  markup  over 
unit production  costs, Cf.  With Xf denoting  ex  
port  quantity,  the  exporter's  profit,  Vf, is 
defined 
as 
Profit maximization yields  
where rj  is  the price  elasticity  of demand. Ac  
cording  to  Hung  et al. (1993),  a more general  
Xf  = bfi  X„  (PfIP O ER)i (1)  
*/=  ~t] (Pf-Po  ~ er)  +  x 0  +  c  + £ (2)  
Vf={Pf-CfiXf (3)  
Pf- (r\-\) (4)  
Silvo  Fennica 32(1)  research  articles  
64 
case,  in  which competitors'  prices  determine the 
exporter's  price,  can  be presented  by  using  the 
concept of a variable markup.  Then,  a variable 
markup  can  be defined by  assuming  that the 
coefficient T)  depends  partly on price competi  
tiveness  in the export  market.  Competitiveness  
can  be described as  the relative  price  (P„  ER)/Pf,  
where (P„  ER) is  the competitors'  price in the 
terms  of exporter's  currency (ER  = FIM/£).  Thus, 
the price  elasticity  of  demand is 
From (4)  and (5),  the pricing  behavior of  a  profit 
maximizing  exporter can  be described as  a  vari  
able markup  over  the unit cost: 
The variable markup,  0, depends  on the relative 
price  and may be approximated  as  
where 0  (>  0)  is  the relative  price  elasticity  of  the 
markup.  The constant  markup  is obtained if 0  = 0 
and <p'  = 77/(77 1). The second equality  in (7)  
comes  from the  log-linear  approximation  of the 
nonlinear function </>. Substituting  (7) into (6) 
and taking a logarithmic  transformation,  a rela  
tion for the  price  of Finnish sawnwood in the 
UK market  is obtained: 
where 7=  1/(1  + 0),  0  < y< 1, <5  =  ln07(l  + 0)  is  
a constant, and u  is  a disturbance term that cap  
tures  all other factors.  The other  symbols  are  the 
same as  above. Lower-case  variables  denote logs  
of the corresponding  upper-case variables. The 
export  price,  pj,  is  homogenous  of  degree  zero in 
the exchange  rate  and competitors'  prices,  and 
the equality  restriction is  imposed  on the coeffi  
cients of  p 0  and er  in  the estimation. 
The degree  of exchange  rate  pass-through  (PT)  
can be derived as the absolute value of the ex  
change  rate  elasticity of export  price  measured 
in foreign  currency (e.g.  Kongsted  1996). From 
equation  (8)  we obtain 
If 7=  0,  then the exchange  rate  affects  the Finn  
ish price in FIM (equation  8)  and thus the chang  
es  in  er  (FIM/£)  are absorbed by the variable 
markup.  This means that Finnish exporters  do 
not  pass  through changes  in the exchange  rate  to 
their export  prices  in pounds,  i.e.  PT  = 0,  which 
indicates perfect  competition  and the existence 
of the law of  one price  in  the market. 
If  7= 1, Finnish  export  prices  in FIM,  pj,  are  
proportional  to  production  costs, c/,  and  PT  = 1. 
This indicates that  Finnish exporters  fully  pass  
through er  changes  to their export  prices  in 
pounds  and keep  the markup  constant.  This indi  
cates imperfect  competition  in the market. Be  
tween the above extremes (o<y<  1) we have 
incomplete  pass-through.  In this  case,  changes  
in er are reflected partly  in FIM-denominated 
and partly  in foreign  currency-denominated  ex  
port  prices.  
The model used  in the study  may naturally be 
affected also by  factors other than  those includ  
ed in the model. If production  costs include an 
important  imported  component,  it is  evident  that 
the degree  of pass-through  will be less  than one,  
even if 7= 1. In the Finnish  sawnwood industry,  
inputs  are for the most part  of domestic origin, 
which suggests  that this effect  is not  very  impor  
tant. 
3 Estimation  of  the Model  
3.1 The Data 
During  the period  studied, the FIM fluctuated 
widely  with respect  to the pound  sterling  (graph  
1C in Appendix  1). Moreover, the  exchange  rate  
regime  was revised several  times. The FIM was 
fixed with respect  to other currencies in the years  
1978-1992. However, the value of  the FIM was  
changed  by  specific  decisions of  the Finnish gov  
ernment, being  revalued in 1979, 1980 and 1989 
and devalued in 1978, 1982, 1991 and 1992. 
Revaluations can be seen as  declines and deval  
uations as  rises in the graph (1C). During  the 
fixed exchange  rate  period,  the value of  the  FIM 
was  measured by  a  currency index that included 
Tl  = T]({Po  ER)IPf) (5) 
Pf=<pCf (6) 
<P  =  <P{(P„  ER)/Pf)  =  f((P n  ER)/Pf) 0 (7) 
Pf=S+(l-y)(er  +  p  0)  +  ycf +u (8) 
PT =  -{dipf -  er)/der)  = % O < PT  < 1 (9) 
Hänninen Exchange  Rate Changes  and  the  Finnish  Sawnwood Demand and  Price  in the UK  Market 
65 
the currencies of  Finland's most important  trad  
ing  partners, until June 1991. After this,  the FIM  
was  unilaterally  linked to the ECU basket  until 
September  1992 and then allowed to float until 
October 1996. 
The development  of the exchange  rate  (graph  
1C) can  be divided into three subperiods. The 
period  of depreciation, 1978-1980, was  followed 
by  a long  period,  1981-1990, during which the 
FIM  appreciated  by about 22  per cent. In the 
years  1990-93 the FIM again  depreciated by 
about 26  per cent  with respect  to the pound  ster  
ling.  The  same subperiods  can also be distin  
guished  in the development  of the other  varia  
bles. A comparison  of the graphs  (18, IC, ID, 
IF) indicates that FIM  depreciations  have  low  
ered the relative  Finnish price  in pounds  and 
increased  export  quantities  and Finnish  market 
share  in the UK.  FIM appreciations  have had the 
opposite  effects. 
It seems that exchange  rate  changes  have  af  
fected Finnish competitiveness  and market share 
by  changing  the Finnish  price in terms of  pounds,  
while the markup  has remained relatively  con  
stant.  Comparison  of the markup  on  production  
costs  (pf  -  Cf)  and  the exchange  rate,  er,  indicates 
that exchange  rate  fluctuations have not  had much 
of an impact  on the markup.  The markup  series 
is  relatively  stationary up to 1990. After 1990, 
when FIM was  devalued two  times by  large  per  
centages, the markup  jumped  up to a  higher  lev  
el (graph  IE). It  is possible  that the markup  
absorbed a  larger  part  of the exchange  rate  change  
in the 1990 s  than in the 1980 s.  If this  is  true, PT 
has  decreased in the 19905. It  is  also likely  that 
the markup  increased partly  as  a result of a  de  
crease  in production costs. Finnish prices  of  saw  
logs  decreased by about 30  percent due to the  
recession in the Finnish  economy in the years  
1990-93. Exchange  rates  and prices  of products  
and inputs  seem to have  been much more turbu  
lent in the 1990 s  than earlier,  which makes it 
difficult to draw conclusions.  
The data used in this study  were seasonally  
unadjusted,  quarterly, and covered the period  
from 1978 to 1994. Finnish  and competitors'  
prices  of sawnwood were described by import 
unit values of sawnwood to the UK (£/m
3
)  based  
on CIF figures  (CIF includes  cost, insurance and 
freight). The Finnish price  in pounds  was  con  
verted into FIM, pf,  at the nominal exchange  
rate.  The competitors'  price,  p„, is  a (quantity  
share)  weighted  unit price of imports  (£/m
3
)  from 
Sweden, Canada and Russia.  Competitors'  im  
port  quantity, xa,  is  the sum of UK imports  from 
these three countries. Data  on  quantities  and val  
ues  of  sawnwood (SITC  248.2-3)  imports were 
taken from the Overseas  Trade Statistics of the 
United  Kingdom (CSO)  for the years 1978-1990 
and from the intra- and extra-EU trade statistics 
(European  Commission)  for the rest  of the ob  
servations. 
The exchange  rate  variable, er  (FIM/£), was 
the nominal quarterly average. Price observa  
tions for the period  1990:1-1994:4 were trans  
formed using  the £/ECU rate.  Both exchange  
rates  were  obtained from International Monetary 
Fund (IMF) statistics. Production costs  of Finn  
ish  sawnwood, c/,  were  described by  the produc  
tion price index (1990=  100, manufacture of 
wood and furniture), which was  obtained from 
the Bulletin of  Statistics  (Statistics  Finland).  The 
use  of the production  price  index as  a proxy  for 
the domestic  cost  of the  sawnwood industry  may 
cause uncertainty  in the testing  of PT,  but it is 
used because proper cost  variables are  not  avail  
able. Another possibility  would have been to 
describe the production costs by the price  of  
rawmaterial. Quarterly  time series for  prices  of  
sawlogs  are  however available only  from year 
1985, which would make the estimation period  
too  short. 
3.2 Estimation Method 
The cointegration  estimation of the model com  
prising  Equations  (2)  and (8)  involves  a  vector  
of six  time series. Provided that  the data is  cointe  
grated,  it may  be possible  to identify  two  cointe  
gration  vectors  the coefficients of which describe 
the long-run  equilibrium  relationships  implied 
by  equations  (2)  and (8). 
The present study  uses  Johansen's  cointegra  
tion method (Johansen  1988, 1995), which is  
suitable for the estimation of nonstationary  data. 
The main advantage  of  Johansen's  method com  
pared to  the earlier alternative multivariate meth  
ods  (e.g.  Engle  and Granger  1987) is  that  it  makes  
it possible  to  estimate and identify  multiple 
research  articles Silva  Fennica  32(1)  
66 
cointegration  vectors.  In more complex  cases  
than the two-variable case,  there may  exist  up to  
p -  1 (p  denotes the number of  variables) cointe  
gration  vectors.  The  earlier methods do not  solve  
this problem.  When a cointegration  relationship  
was  identified, it was  assumed to be unique.  This 
need not  be true in a multivariate case. Johansen's  
method provides  a log-likelihood  ratio test sta  
tistic for determining  the number of  cointegra  
tion vectors in the data. 
If there  exist  more than one cointegration  vec  
tor linking  the variables together,  the relation  
ship  estimated by  the Engle-Granger  method will  
be invalid. The relationship  may simply  repre  
sent  complex  linear combinations of all  the  
cointegration  vectors.  Johansen's method,  un  
like the Engle-Granger  procedure,  also accom  
modates  short-run dynamics  in the cointegration  
regression.  This helps  to reduce biases and im  
proves  efficiency  in using  the information con  
tent  of the data in the estimation. Johansen's 
method also enables testing of  several  economic 
hypotheses  by means  of linear restrictions  in the  
same cointegration  framework.  
The stationarity  of the variables was  not  tested  
separately  prior to the analysis,  as  would  be the 
case  e.g. if the Engle  and Granger  (1987)  meth  
od  had been used. The property  that unit vectors 
can  be cointegration  vectors  in Johansen's mod  
el means that one can include in the cointegra  
tion  analysis  /(l) or  /(0) variables provided  they  
are  economically  meaningful  (Johansen  1995, p. 
74). By including  a stationary  variable in  the 
vector  of  variables,  X,  one adds an extra cointe  
gration  vector, i.e. an extra dimension in cointe  
gration  space. 
The analysis  of the present study  starts by 
estimating  an  unrestricted statistical VAR model 
including  all  six variables. The method of Jo  
hansen  (1988)  and Johansen and Juselius (1990) 
uses  a statistical model that is a  
VAR process  of order k,  where p  is  the number 
of variables. The VAR  model can  be formulated 
as 
where Ax,  is  an 1(0) vector of  the six  first-differ  
enced variables. In the VAR model the constant  
term, fi, can be restricted to cointegration  space 
so as  to represent the absence of a  linear trend in 
the data. However, in the present  study  a linear 
time trend was  assumed to  exist  in the data and 
thus the constant  term was not  restricted. D,  is  a 
seasonal dummy and k  is the lag  length.  Intro  
ducing  a sufficient number of lags  usually  pro  
duces a  well-behaved error  term NID(O,Q). F
b...,
 
rv, and n = -  I + lli  + n2 +,...,+ 11* are  coeffi  
cient  matrices and e,  is  a vector  of error  terms 
assumed to be  normally  and independently  dis  
tributed with  expectation  zero  and variance ma  
trix £2. 
In (10) the level terms capture the long-run  
steady state relationships  and the first-difference 
terms show  the short-run dynamics.  II is the 
matrix of  long-run  coefficients of  the lagged  lev  
els.  It can be decomposed  into a  matrix of  load  
ings,  a,  and  a matrix of  cointegrating  vectors,  [3,  
i.e. Il = a(3'.  If  x,  is  1(1), i.e.  integrated  of  order 
one,  the components of  x, are  cointegrated.  The 
number of cointegration  vectors  that exist  among 
the variables is determined by  estimating  the 
rank of the matrix  11.  Johansen's method (Jo  
hansen 1988)  formulates two  likelihood ratio tests 
for the cointegration  rank,  r,  from which the 
trace  test  is  applied  here. The trace  test for test  
ing  that there are  at most r  cointegration  vectors 
in the set  of  p  variables is  defined as 
where T is  the number of observations and the 
Vs  are  the smallest squared  canonical correla  
tions (eigenvalues).  When the number of  cointe  
gration  relations,  r,  is  determined it is  possible  to 
test  hypotheses  on  the long-run  matrix,  II  =  a(J'.  
The present  study  tests restrictions  on the Py  
coefficients. In the Johansen framework these 
tests are conducted under the hypothesis  
where H is  a  (p  x  s) matrix, <p is  a(jxr) matrix 
and  r<  s < p. Restrictions,  s, are defined by  cp, 
and  r  is  the number of cointegration  vectors.  The 
exclusion  test (excluding  a  variable from a  cointe  
gration relation)  was  performed  in order to iden  
Ax,  = T|Ax,_!  +,..., + rt_,  Ax,^ + i  
+  FLc,_t  +  H  +  <t>D,  + t=1,...,T (10)  
Trace  (r)  =  -T  £/«(l-X.;) (11)  
i=r+ l  
P  = H(p or  n  = acp'H' (12)  
Hänninen Exchange  Rate  Changes  and the Finnish  Sawnwood Demand  and Price  in the UK  Market 
67 
tify the economic long-run relation as  being  that 
represented  in equations  (2)  and (8).  The testable 
null hypothesis  for  exclusion of a variable was 
Pij  =O. Homogeneity  between Finnish  export  
price,  pf,  and the other  nominal variables was  
tested by  restricting  the respective  coefficients 
accordingly.  Finally, the markup  relationship  was  
tested by  examining  if  the coefficient of  C/ could 
be restricted to unity. 
4  Results  
4.1 Cointegration  of  the  Empirical  
Variables 
The cointegration  estimation was  based on the 
VAR(4)  model given  by  equation  (10),  with six  
equations  (p = 6)  for the period  1978-1994. The 
lag length,  k,  of the VAR model was  determined 
by  the Schwarz  (SC) and Hannan-Quinn  (HQ)  
information criteria,  using  likelihood ratio tests. 
Starting  from k  = 5 (see  Doornik and Hendry  
1994, p.  287),  a  reduction of the VAR from k  = 5 
to k  = 4 was accepted.  Because the reduction 
from k  = 4 to  k  = 3 was rejected,  k = 4 was  used 
for further modeling.  
The diagnostic  tests on the residuals of the 
VAR(4)  model are presented  in Table 1, and 
they  support the model with k = 4. Autocorrela  
tion of the residuals was  examined using  the F  
form of the Lagrange  Multiplier  (LM)  test, which 
is  valid for systems  with  lagged  dependent  vari  
ables. The null hypothesis  of  no serial autocorre  
lation was  accepted  at the 5  percent  level. Heter  
oskedasticity  was  tested using  the F-form of the 
LM test against  4th order autoregressive  condi  
tional heteroskedasticity.  The  null hypothesis  of 
no heteroskedasticity  was  accepted  at the 5  per  
cent  level. Normality  of  the residuals was  tested 
by  means of the Doornik-Hansen test (Doornik  
and Hendry  1994) and the null hypothesis  of 
normality was accepted  for  all the equations.  
Also the corresponding  vector tests  for the equa  
tion system  accepted  normality  and indicated no  
autocorrelation. For further details and referenc  
es  concerning  these  tests, see  Doornik and Hen  
dry (1994).  
The results of the cointegration  estimation of 
the VAR(4) model indicate that r  = 2  (Table  2).  
According  to Johansen's  trace  test, the hypothe  
ses  of r  = 0 and r < 1 can be rejected.  Thus two  
cointegration  vectors  are accepted  at the 5  per  
cent  level. The eigenvectors  (Pj)  and  their weights  
(otj)  obtained from the cointegration  estimation 
of model (10)  are shown  in Table 3. Of  the six  
eigenvectors,  the first  two  relations ((3i  and p 2) 
are most highly  correlated with the stationary 
part  of the process  Axt corrected for the lagged  
values of  the differences. Thus,  Pi  and  P2  are  the 
two  cointegration  vectors  determined by  the mod  
el (Johansen  1995). They  are  normalized by  the 
coefficients of  Finnish export  quantity,  Xf,  and 
Finnish  export  price,  pf  
The normalized vectors of loadings,  otj, are 
Toble 1. Misspecification  tests for  the  residuals  of  the  VAR(4) model.  
Equation 
Autocorrelation 
Far  (4,32)  
Tests for  the residuals  and the standard errors 
Heteroskedasticity Normality Standard errors 
Farch  (4,28) X
2
N(2) <Je  
A(*/) 0.97[0.44] 0.02[  1.00] 0.05 [0.98] 0.16 
A(Pf)  1 .96[0. 12] 0.33[0.86] 2.00[0.37] 0.05 
A  (er)  0.44[0.78] 0.22[0.93] 0.99[0.61] 0.04 
Mcy)  1.14[0.36] 0.48[0.75] 4.01 [0.13] 0.01 
&(Po)  2. 14[0. 10] 0.14[0.97] 0.77[0.68] 0.07 
A(x„)  0.69[0.60] 0.34[0.85] 0.32[0.85] 0.12 
System:  VFar  (144,48) = 1.10[0.36]  VJC
2
n  (12)  = 9.98[0.61] 
Note: Values in square brackets  are  marginal significance  levels.  
Silva  Fennica 32(1)  research articles  
68 
Table 2.  Results  for  the  cointegration  rank  test,  
presented  under the (3j  vectors  in Table 3.  The 
(Xjj's  represent  the weights  with  which the error  
correction terms enter each equation  and they  
indicate the average speed  of  adjustment  toward 
the estimated equilibrium  state. A  low coeffi  
cient indicates slow adjustment,  while a  high 
coefficient indicates rapid  adjustment.  Table 3 
shows  that the loadings  of the two  cointegration  
vectors  (Pi and  p2) are rather  low.  
4.2 Long-run  Exchange  Rate Pass-through  
The two  estimated cointegration  vectors  are  test  
ed with restrictions  in  order to identify  the model 
system  comprising  export  demand and price re  
lations for Finnish  sawnwood. The unrestricted 
cointegration  vectors  from Table 3 are 
where, Xf  and pj  are  Finnish  quantity  and unit 
price (FIM/m
3>  of sawnwood exports  to  the UK 
and x„ and pa are the respective  quantity and 
price  (£/m 3)  of  competitors'  sawnwood.  er  is the 
exchange  rate  (FIM/£)  and Cf  is  the  unit cost  of  
Finnish  sawnwood output. 
The first  cointegration  vector,  P, (equation  13), 
was  identified, as  was  equation  (2),  by  excluding  
Finnish  production  cost  (9)  from the relation  and  
Table  3.  Normalized eigenvectors,  βj,  with  correspond  
ing weights, αj, obtained  from the  unrestricted 
cointegration estimation. 
assuming  that the coefficient of is (3ei  =-1 ■  The 
exclusion  of  c/was  accomplished  by  restricting  its 
long-run  coefficient, p4 |,  to zero.  The second 
cointegration  relation,  p2 (equation  14),  was  iden  
tified, as  was  equation  (8),  by  excluding  Finland's 
and the competitors'  quantity.  
When the above restrictions  and the homoge  
neity assumption  required  by economic theory  
were tested for the  demand and price  relations, 
the test  rejected  this structure  (structure  I in Ta  
ble 4).  The homogeneity  of the equations  in the 
nominal variables pj., (p„  +  er)  implies  that 
the coefficients of pa  and er  should be equal  and 
that  the coefficient of  cs  should equal  the  differ  
ence  between the coefficients of  ps  and  er in the 
cointegration  vectors. However, the coefficients 
of the unrestricted equation  (13)  in particular  
indicate that homogeneity  does not  necessarily  
hold. Thus testing  was continued by applying  
structure II (Table 4),  where the homogeneity  
restriction is  applied  only  in the price equation.  
The structural form (II)  is  accepted.  The type 
of demand equation  (1) applied  in the present 
P,:  1 .OOxf-  6.15pf-  4A9er 
+  8.90cy-0.89 />„-2.3(k0  and (13) 
P  2:  o.44xf+\.oopf+o.s7er-0.72cf  
-  1.26 p„  +  0.50  a„ (14)  
Null hypothesis  Eigenvalues X trace 
statistics 
95% critical 
values  
H0:rSi Xi T(i) C(i)  
r=  0 .56  122.9 * 94.20 
r<  1 .36  70.83 * 68.50 
r<2 .29  42.23 47.20 
r<  3 .22  20.69 29.70 
r<  4 .06  4.88 15.40 
r<  5 .02  1.18 3.80 
Note: 
*
 indicate rejection of  the  null hypotheses  implying that  the  rank  
is  2. 
Variables Eigenvectors 
p. P2 PJ  P4  Ps  P& 
x
f 
1.00 0.44 -0.03 0.83 -0.41 —0.81 
Pf -6.75 1.00 
-0.48 1.50 -0.82 —0.25 
e
r  -4.49 0.57 1.00 -2.35 0.97 2.29 
C1 
8.90 -0.72 0.26 1.00 -0.22 0.21  
Po -0.89 -1.26 0.38 -2.16 1.00 -0.32 
Xo -2.30 0.50 0.27 -1.15 0.06 1.00 
Variables Weights 
ai  «2 <*3 0<4 as Ob 
xf -0.05 -0.09 -0.29 0.26 0.24 0.11  
Pf 
0.05 -0.03 0.16 -0.17 -0.10 0.01  
Cr  -0.01 -0.02 -0.52 -0.06 -0.03 0.01 
C
f 
-o.oi  0.04 0.05 -0.02 -0.01 0.00 
Po  -0.04 -0.04 0.57 0.14 -0.11 0.03 
XQ 0.16 0.15 0.10  0.32 0.06 0.05 
Symbols:  x/=  Finnish quantity, p/= Finnish  price  (FIM), er= ex-  
change rate (FIM/£), c/= 
Finnish unit cost,  p<} =  competi- 
tors'  price  in pounds  sterling and  x„  =  competitors' quanti- 
ty-  
Hänninen Exchange  Rote  Changes  and the  Finnish  Sawnwood Demand and Price  in  the UK  Market  
69 
Table 4. Tests for  the  restrictions  on the  unrestricted  cointegration  vectors  β1  and  β2  under  r  = 2. 
Notes:  1) Symbols:  x/  = Finnish quantity, p/  =  Finnish price  (FIM), er = exchange rate (FIM/£), cf -  Finnish unit cost,  
Po =  competitors' price in pounds sterling and x 0 = competitors'  quantity. 
2) 
*
 indicates  rejection of  the  restricted  model structure. 
study  is  not  often tested for homogeneity  in ear  
lier applications.  The homogeneity  condition has 
usually  been  satisfied simply  by  expressing  the 
exporter's  and  competitors'  prices  in a common 
currency in the form PFIPO.  Because the price  
relation fulfills the restrictions  and produces  the 
PT estimate,  the estimation is  continued by  re  
stricting  it further. 
In the accepted  structure  (II),  the price  relation 
p 2 resembles the markup  pricing  relation with 
7= 1 (Table  4).  The coefficient of unit cost,  cf,  
is  
close to unity (-0.92),  while the coefficients  of 
the exchange  rate  (er) and competitors'  price  
(p0 ) are close to zero. Finally,  the markup  as  
sumption  was  tested by restricting  the price  rela  
tion accordingly.  The resulting  structure  (III) is 
also accepted  by  the test and the  final long-run 
equilibrium  relations  can be  presented  as  
where the symbols  are the same as  above. Be  
cause  the demand equation  failed the homogene  
ity  test,  interpretation  of its  coefficients is  prob  
lematic. However, the signs  of the own-price  
and exchange  rate  elasticities of Finnish  sawn  
wood export  demand are  consistent with the eco  
nomic theory.  Moreover, the magnitude  of the 
price elasticity  (-2.44) is between the earlier 
estimated results  of Hänninen 1994 (-1.71,  esti  
mated from annual data, 1976-90) and of  Tervo 
et ai. 1988 (-3.1,  estimated with Aimon polyno  
mials,  1-12 lags  from quarterly data, 1966-85). 
The relatively  large  elasticity of the exchange  
rate  (6.56)  implies  a large  effect  of exchange  
rate  on Finnish sawnwood exports  to the UK. 
The restricted price  relation (16)  representing  
markup  pricing,  with y= 1,  indicates that Finn  
ish sawnwood export  price  in FIM  is  proportion  
ate to production  cost  and the effect  of the ex  
change  rate  on the FIM price  is  very  small (zero).  
Thus, exchange  rate  pass-through  is large. A 
large  pass-through  coefficient indicates that ex  
change  rate  changes  are  reflected almost pro rata  
in the Finnish export  price  in pounds  sterling.  
5  Conclusions  
The study examined the long-run  exchange  rate  
pass-through  (PT)  for the Finnish  price of  sawn  
wood in the  UK market by  estimating  a demand 
and price  equation  system.  The data were  quar  
terly and covered the years 1978-1994. Jo  
hansen's cointegration  method, which is  suitable 
for analyzing  nonstationary data, was  used in  the 
p'i:  I.ooxf=  -2.44 pf+  6.56(er  +  p0 )  
+  1.00;t
o
and (15) 
P'2:  I.oop/= +  0.00(er  +  p0)  + I.ooc/ (16) 
Variables/ Restricted and normalized demand (fin) and price  (Bj2)  relations (i  = 1 6; j = 1,2) 
LR-tests  
(I) (ID (ill)  
Psj  = P3j  and  P31  =  -P21. P31 *-p2l. Mark-up: P42 = 1, 
P41 =0,  Pöl =  -1,  P12 = 0, P62 =  0, P42 =  —<p22 + P32) others  unchanged others  unchanged 
Pil Pi2 Pil P,2 Pil Pi2 
x
f 
1.00 0.00 1.00 0.00 1.00 0.00 
Pf 2.13 
1.00 1.57 1.00 2.44 1.00 
er  -2.13 -0.04 -6.16 -0.08 -6.56 0.00 
c
f 
0.00 -0.96 0.00 -0.92 0.00 -1.00 
Po  -2.13 -0.04 -6.16 -0.08 -6.56 0.00 
Xo -1.00 0.00 -1.00 0.00 -1.00 0.00 
LR x
2 (1/4  4)= 14.62*  X
2
(l/4  3)  =  6.84  X
2
(l/4  5) = 7.63  
Silva  Fennica 32(1)  research articles 
70 
estimation. The main advantage  of Johansen's 
method in  this  study  is that it allows for  the 
estimation and identification of a  system model, 
unlike the traditional Engle and Granger  (1987)  
procedure.  It also accommodates short-run dy  
namics in  the cointegration  regression, which 
helps  to reduce biases. Johansen's method also  
enables testing  of several  economic hypotheses  
by  means of  linear restrictions  in the same cointe  
gration framework. 
According  to  Johansen's rank test, two  cointe  
gration vectors  were  determined among the vari  
ables. These vectors  were restricted in order to  
identify them with the theoretical export  demand 
and price  relation. The restrictions  were  accept  
ed in both cointegration  vectors, but in the ex  
port demand relation (15)  the assumption  of  price 
homogeneity  did  not  hold. Thus, the coefficients 
of the demand equation  are  problematic  to  inter  
pret.  However, the price  and exchange  rate  elas  
ticities were  of the expected  signs  and the mag  
nitude of the  price  elasticity  was  consistent  with 
the earlier estimated  results (Hänninen  1994 and 
Tervo et ai. 1988).  The  elasticity  of  the exchange  
rate  estimated from the demand equation  was 
relatively large, implying  a  large  effect for the 
exchange  rate  on Finnish sawnwood exports  to  
the UK.  The result does not  agree with the earli  
er  studies,  which indicate  that the trade flows of 
forest  products  are not  very sensitive to the ex  
change  rate  variations (Buongiorno  et al. 1988 
and Uusivuori and Buongiorno  1990). The pre  
vious results  are however scanty  and they  con  
cern  exports  to the  US  market. 
The  responsiveness  of  trade flows to  exchange  
rate  changes  is  connected with  the magnitude  of 
PT.  For  example, a low  PT indicates that trade 
flows remain relatively  insensitive to exchange  
rate  changes  (Menon 1995). Thus, the above  
mentioned results of Buongiorno  et al. (1988)  
and Uusivuori and Buongiorno  (1990)  indicate 
low  PT  for exports  to  the  USA. Several studies 
have  estimated low PTs (e.g. Yang 1997) for 
U.S. imports  of many commodities. This indi  
cates  that trade flows directed to US market  have 
not  been  very  sensitive to exchange  rates.  
The pass-through  coefficient obtained from the 
restricted price  equation  (16)  was  close to unity 
(0.92) and it could be further  restricted to one. 
This means that exchange  rate  pass  through  in 
Finnish sawnwood price  expressed  in pounds  
sterling  has been large.  The result is  consistent 
with the relatively  large exchange  rate elasticity  
obtained from the demand equation.  However, 
the  lack  of  proper data on  sawnwood production  
costs may have compromised  the measurement  
of PT. 
Several  earlier findings  for  forest  industry  prod  
ucts  indicate relatively  large PTs  with the excep  
tion of  US imports.  For  example,  Uusivuori and 
Buongiorno  (1991)  estimated PT to be between 
0.79 and  unity  for US lumber  exports to  Japan.  
According  to Menon (1993  a),  PT was  0.80 for 
Australian imports of wood products  and 0.45 
for  paper products.  The  results  of  Vesala (1992)  
showed that PT for the export  prices  of Finnish  
paper manufacturers varied between 0.66 and  
0.69 in western  Europe  and between 0.16 and  
0.30 in the USA. 
A large  PT  for  Finnish sawnwood exports  
means that variations in exchange  rates  are  al  
most completely  passed  through  to  the foreign  
currency  prices  in  the long-run and the markup  
is  kept  unchanged.  For example,  depreciations  
and devaluations of the FIM have  lowered the 
relative Finnish price and improved  competi  
tiveness  and market share  in the  UK. Because 
prices  expressed  in foreign  currency change  as  a 
result of an exchange  rate  change,  a large PT 
also implies imperfect  competition. Also in this 
respect,  the results are consistent  with earlier 
results implying  imperfect  competition  and re  
jection  of the law of  one price for UK  sawnwood 
imports (Hänninen  1998). 
The resulting  large PT indicates that Finnish 
sawnwood exporters have made use of  devalua  
tions to  increase their market shares  and  export 
quantities but not  necessarily  their  profit mar  
gins. If EMU is  realized and Finland joins,  the 
possibility  of using  exchange  rate  policy  to im  
prove Finnish competitiveness  and market share 
will disappear.  Even outside the EMU, the scope 
for national exchange  rate  policy  would be re  
duced (Hetemäki  et al. 1997).  Hence, other means 
must be found to  adjust  to future disturbances 
caused by  demand and price decreases in the 
world market. 
One way  to improve Finnish competitiveness  
during recessions  could be  to adjust  production 
costs. However, a  rapid  adjustment  is  problem  
Hänninen Exchange  Rate  Changes  and  the Finnish  Sawnwood Demand and  Price  in the UK  Market  
71 
atic due to  the relatively  long-term price con  
tracts that exist  in the Finnish input  markets. 
Hence, it is  also necessary  to differentiate Finn  
ish  products,  for example  via customer  orienta  
tion,  established customer  relationships  and by 
product  planning.  Because PT seems to be high  
er  for wood industry  products  than for paper 
products,  according  to the earlier results,  the 
sawnwood industry  may have  more difficulty 
than the paper industry  in adjusting  to  the EMU 
environment. 
For further research,  the short-run dynamics  
of exchange  rate  changes  vis-ä-vis  Finnish  sawn  
wood prices would provide  a useful and interest  
ing  challenge.  Earlier studies of pass-through  
indicate that price  adjustment  does not  necessar  
ily  happen immediately (Dornbusch 1987).  The 
pass  through effects  could also  be examined with 
respect  to other forest products  and other curren  
cies,  for example  the Swedish krona,  which  has 
played  an important role in western European  
sawnwood trade, particularly in the 19905. 
Acknowledgements  
Helpful  comments  of  Joseph  Buongiorno,  Lauri 
Hetemäki, Sören Johansen, Hans-Christian Kong  
sted, Jari Kuuluvainen, Bo Jellesmark Thorsen, 
Anne Toppinen  and Jussi  Uusivuori in preparing  
this paper are gratefully acknowledged.  This 
study was  supported  financially  by  Metsämiesten 
Säätiö Foundation. 
References  
Athukolara, P. 1991.  Exchange rate  pass-through:  the 
case  of Korean exports of manufacturers. Eco  
nomic  Letters  35(1):  79-84. 
&  Menon, J. 1994.  Pricing to market behavior  and  
exchange rate  pass-  through in  Japanese exports.  
Economic  Journal  104(423): 271-281.  
&  Menon,  J. 1995. Exchange rates  and  strategic  
pricing:  the  case  of  Swedish machinery exports.  
Oxford  Bulletin  of  Economics  and  Statistics  57(4):  
533-545.  
Buongiorno, J., Chavas,  J.P. &  Uusivuori, J. 1988.  
Exchange rates, Canadian lumber imports, and 
United States prices:  A time-series analysis.  Ca  
nadian  Journal  of Forest  Research 18:1587-1594. 
CSO,  Overseas  Trade Statistics  of  the United King  
dom, 1978-1990. 
Doornik, J.A. &  Hendry, D.F. 1994. PcFiml 8.0. In  
teractive  econometric  modeling of  dynamic sys  
tems.  Institute  of Economics and  Statistics,  Uni  
versity  of  Oxford. International Thomson  Pub  
lishing,  London. 436  p.  
Dornbusch, R.  1987. Exchange rates  and  prices.  Amer  
ican  Economic  Review  77(1):  92-106. 
Engle, R.F.  &  Granger. C.W.J. 1987. Co-integration 
and  error  correction: representation,  estimation and  
testing.  Econometrica 55:  251-276.  
European  Commission, Intra- and  extra-  EU  trade, 
CD-ROM, 1988-1992. Luxemburg. ISSN 1017- 
6594. 
Facts  and  Figures  1997.  Finnish  Forest  Industries  Fed  
eration, Helsinki.  33  p. ISSN 1238-4143.  
Feenstra, R.C., Gagnon, J.E. &  Knetter, M.M. 1996. 
Market  share  and  exchange rate  pass-through in  
world  automobile trade.  Journal of International 
Economics  40: 187-207.  
Hetemäki, L., Hänninen, R.,  Kuuluvainen,  J.,  Ollon  
qvist,  P.,  Ruuska, P., Seppälä, R.,  Toppinen, A. & 
Uusivuori,  J. 1997.  EMU ja  Suomen metsäsek  
tori. Valtioneuvoston kanslian julkaisusarja  17. 
26 p.  + appendices.  
Hooper, P.& Mann, C.L. 1989. Exchange  rate  pass  
through  in  the  1980  s:  the  case  of  U.S.  imports  of 
manufactures.  Brookings Papers on Economic  
Activity  1989(1): 297-337. 
Hung, W.,  Kim, Y. &  Ohno, K.  1993.  Pricing  exports:  
a cross-country  study. Journal of International 
Money and  Finance  1993(12): 3-28.  
Hänninen, R.  1986. Suomen  sahatavaran  vientikysyntä  
Länsi-Euroopassa vuosina  1962-1983.  Summary:  
Demand  for  Finnish sawnwood exports  in  western  
Europe,  1962-1983. Folia  Forestalia  657. 25  p.  
1994. Price elasticities of substitution between 
Finland and  the other  supplier countries  in the 
demand  for the  United Kingdom  sawnwood  im  
ports. In: Helles, F.  &  Linddal, M. (eds.).  Proc.  of 
the  Biennial  meeting of the  Scandinavian  forest 
economics. Scandinavian Forest Economics 35:  
204-217.  
1998. The law  of one price in the  United  Kingdom 
sawn softwood  imports -  a cointegration  approach. 
To be  published in  Forest  Science 44(1), 1998. 
research  articles Silva  Fennica 32(1)  
72 
IMF, International Financial  Statistics,  1978:4-1995:4. 
Washington.  
Johansen, S. 1988. Statistical analysis  of  cointegra  
tion vectors.  Journal of  Economic Dynamics  and  
Control  12: 231-254. 
1995.  Likelihood-based  inference in  cointegrated 
vector  autoregressive models.  Oxford University  
Press  Inc.,  New  York.  267 p.  
&  Juselius, K. 1990. Maximum  likelihood  estima  
tion and  inference on cointegration  -  with appli  
cations to the demand for  money.  Oxford  Bulletin 
of Economics and  Statistics 52:  169-210. 
&  Juselius, K.  1992. Structural  tests  in  a  multivar  
iate  cointegration  analysis  of  the PPP  and  the UIP 
for  UK. Journal  of Econometrics  53: 211-244.  
Knetter, M.M. 1989. Price  discrimination by U.S. and 
German  exporters.  The American Economic Re  
view 79(1):  198-210. 
1993. International comparisons of  pricing  to  mar  
ket behavior. The American Economic Review 
83(3): 473-486.  
Kongsted, H.C. 1996. Modeling price and  quantity 
relations for Danish manufacturing  exports. In: 
Dynamic  models of  foreign trade  under  fluctuat  
ing  exchange rates:  theoretical and  empirical  ap  
plications. PhD.  thesis in economics. University  
of Copenhagen. Rod  Serie  (39).  p. 101-169.  
Menon,  J. 1993 a. Exchange rate  pass-through elastici  
ties  for the  Monash model: a  disaggregate  analy  
sis  of Australian manufactured  imports.  Center of 
Policy  Studies Monash University,  Working  Pa  
per  No.  OP-76  July.  49  p.  
1993b.  Exchange rate  pass-through for  Australian  
manufactured  imports:  estimates from the  Johansen 
maximum-likelihood  procedure. Center  of  Policy  
Studies Monash  University,  Preliminary  Working 
Paper No.  IP-60  July.  7  p.  
1995. Exchange rate  pass-through.  Journal  of Eco  
nomic  Surveys  9(2):  197-231. 
Pick,  D.H.  &  Park, T.A 1991.  The competitive  struc  
ture of U.S.  agricultural  exports.  American Jour  
nal  of Agricultural  Economics 73(1):  133-141. 
Rockerbie,  D.W. 1992.  Exchange  rates, pass-through, 
and Canadian export competitiveness:  an  analysis  
using vector  autoregressions.  Applied Economics  
24: 627-634. 
Statistics  Finland,  Bulletin of  Statistics (SVT),  1978 
1995. Helsinki. 
Tervo, M., Mäkelä  J.  &  Hänninen, R.  1988.  Dynaami  
nen  kysyntämalli  Ison-Britannian maitoiselle saha-  
tavaran  tuonnille. Abstract:  A dynamic demand  
model for  British imports  of sawnwood from dif  
ferent  countries.  Metsäntutkimuslaitoksen tiedon  
antoja  (Finnish  Forest  Research  Institute,  Research  
Papers)  313. 35  p.  
Uusi vuori, J. &  Buongiorno, J. 1990.  Short- and  long  
run effects  of  changes  in  exchange rates  on United 
States imports  of  paper from Finland and  Sweden. 
Scandinavian  Journal  of  Forest  Research  5(2):  285- 
296. 
&  Buongiorno, J. 1991.  Pass-through  of  exchange 
rates  on prices  of forest product  exports  from the 
United  States  to  Europe and  Japan. Forest  Science 
37(3):  931-948. 
Vesala, J. 1992. Incomplete exchange rate  pass-through 
and hysteresis  in trade. A survey  of recent theories 
and empirical  study  of  export  pricing  of  Finnish 
paper  manufactures. Bank  of  Finland Discussion 
Papers  29/92. 76 p.  
Yang, J. 1997. Exchange rate  pass-through in  U.S. 
manufacturing  industries. The Review  of  Econom  
ics  and  Statistics 79(1): 95-104. 
Total  of  36 references 
Hänninen Exchange  Rate  Changes  and  the Finnish  Sawnwood Demand and Price  in the UK  Market 
73 
Appendix  1. Quarterly  time series data in logarithms covering  the period 1978-1994. 

V 

LONG-RUN PRICE  EFFECTS  OF EXCHANGE  RATE  CHANGES  IN 
FINNISH  PULP  AND PAPER EXPORTS 
Riitta Hänninen* and Anne Toppinen  
Finnish Forest Research Institute 
Helsinki Research Center 
Unioninkatu 40 A, 00170 Helsinki,  Finland 
￿Corresponding  author 
tel:+3sB 9 8570 5746 
fax:  +358 9  8570 5717;  
e-mail: riitta.hanninen@metla.fi 
ABSTRACT 
This  paper presents  long-run  exchange  rate elasticities of  Finnish newsprint  and pulp  export  
prices  in the United Kingdom  (UK) and Germany  over  the period  1980-1994. The 
elasticities were estimated for a markup  pricing  model using Johansen's cointegration  
method. The  magnitudes  of  the elasticity  estimates obtained  by  imposing  price  homogeneity  
indicate that pass-through  of exchange  rate changes  to export  prices  has been incomplete  in 
both markets.  For  newsprint,  the pass-through  elasticity was  0.60 in the UK and 0.46 in 
Germany,  which  means that  about one  half of  the changes  in the exchange  rates  is reflected 
in prices  as  expressed  in the destination countries' currencies and about a half in prices  
measured in Finnish markka.  Instead, in  pulp  exports  the pass-through  elasticities differed 
from each other between the two destination countries more clearly  than in newsprint  
exports  indicating  a different pricing  behaviour of Finnish pulp  exporters in the UK and 
Germany  when exchange  rate changes.  Pass-through  was small in the UK pulp  market 
(0.07) as compared  to Germany  (0.68).  
Keywords: Exchange  rate pass-through,  international trade, newsprint, wood pulp,  
Johansen's  cointegration  method,  United Kingdom,  Germany  
2 
I. INTRODUCTION  
A distinctive feature in the economies of the Nordic countries and  Canada is  the large  share 
of forest industry  in the  total manufacturing  and exports  of goods.  In Finland,  forest 
industry  products  account for about one third of the total export  earnings.  Cyclical  
fluctuations of the forest industry  are  detectable in the overall performance  of  the Finnish 
economy. Finnish currency  (markka)  has  been devaluated several  times,  which may have 
improved  the profitability  and competitiveness  of Finnish  export industry  in relation to  its 
main competitors,  e.g., Sweden and Canada. However,  empirical  evidence  on  the effects  of  
devaluations on the exports  of forest industries is  largely  missing.  Knowledge  on the 
exchange  rate  effects is important  in analysing  possible  implications  of Finland's 
participation  in the third stage of the European  Monetary  Union (EMU). Adjustment 
difficulties to  the new  market environment may  arise for  Finnish pulp  and paper trade in the 
euroarea,  if the exchange  rate  has  had an important  effect on Finnish  export pricing  and 
exports.  Examining  the effects  of exchange  rate changes  on  product  prices  provides  also  
information on  producers'  pricing  behaviour in the European  pulp  and  paper  markets.  
Exchange  rate effects are  examined in the present  study  using  the concept  of pass  
through  (PT),  which can be defined as  the extent  to which exchange  rate  changes  are  
reflected  in the export price.  PT is  said to  be  complete  (PT=l)  when the exporter  does not  
adjust  prices  in his  home  currency,  and  exchange  rate  changes  are  fully  reflected  in foreign  
currency-expressed  export prices.  By contrast, there is  no PT (PT=O), if the home 
currency-expressed  export  prices  fully  adjust  to  exchange  rate  fluctuations.  PT may  also  be 
incomplete  (partial), which  means  that exchange  rate  changes  are  partly reflected in foreign  
currency  prices and  partly  in the exporters'  home currency prices  (0<  PT< 1). 
PT also produces  information on  the degree of price competition in the markets 
(e.g., Athukolara 1991).  In the earlier literature, incomplete  PT  is  commonly  attributed to  
imperfect  competition  and the use  of market power. Thus, the magnitude of PT indicates 
the effects  of  depreciation/appreciation  of the markka  on  Finland's price  competitiveness  in 
forest products'  markets. 
The interest to  study  PT originates  from the contemporary debate on  the efficiency  
of exchange  rate  policy  in balance of payments  adjustment.  In most  previous  studies  PT is  
found to be incomplete  (see,  e.g.,  the survey  by  Menon 1995). The rate  of  PT is also found 
3  
to  differ between countries  indicating  different pricing  behaviour. Some studies  indicate that 
PT  for  total export  and import  prices  tends  to  be higher  for smaller more-open economies 
than for larger  less-open  economies (Japan,  USA)  (e.g.,  Kreinin 1977, Spitaeller  1980).  
However,  results from the multicountry  studies indicating  the PT based on the size and 
opennes  of the country  are  conflicting  (see  e.g.,  Menon 1995, Khosla and Teranishi 1989). 
The estimates  for PT have also  been found to  vary  widely  by  the industry  (e.g., 
Feenstra et al. 1996), which means  that it is  necessary  to study  the concept at the 
commodity  level. Most studies  concern  aggregate exports or  imports,  while more recent  
research  analyse  PT also  at the  industry-level  (see,  e.g., Hooper  and Mann 1989,  Menon 
1995). The  response  of  export  price  to  changes  in exchange  rates  depends  on several 
factors that affect through  the impact  that exchange  rates  have on marginal costs and 
markups  of  price  over marginal  cost.  The adjustment  of  markups  in response  to  exchange  
rate  change  is referred to  as pricing  to market (Krugman  1987). Studies that examine 
pricing  to  market (PTM)  find that the degree  of PT varies across  export  markets.  They  
suggest that firms price discriminate across  export  markets by  varying  the degree  of 
exchange  rate  PT (e.g.,  Athukolara and  Menon 1995). PTM approach  has  been applied  by  
e.g., Knetter (1989,  1993),  and Pick and Park  (1991)  in estimating  pass-through  elasticities. 
Studies  concerning  PT of  forest  products  are few. However, the opinion  that changes  
in exchange  rates  play  a  key  role  in international trade in forest products  is  widely  held, and 
a  devaluation is  expected  to increase exports  of forest products  (e.g.  McCarl and Haynes  
1985).  From the previous  studies  can  be mentioned Vesala (1992),  who  studied exchange  
rate  PT  for  Finnish  paper exports.  PT  was  found to  be  higher  for exports  to  Europe  (in  the 
range 0.66-0.69)  than for  exports  to the USA (in  the range 0.16-0.30),  which may be due 
to the larger  domestic paper production  in the USA than  in Europe.  Uusivuori and 
Buongiorno  (1991)  found incomplete  PT in forest products  exports  from the USA  to 
Europe  and Japan  for most of the product  categories.  According  to Alavapati  et al.  (1997)  
exchange  rate PT was  small in  Canadian pulp  export  price  indicating that Canadian 
exporters  do not  have significant market power in pulp exports.  The PT estimate of  Menon 
(1993)  for Australian imports  of  paper was 0.45. The results  of  the previous studies indicate 
a larger  PT for wood industry  products  than  for paper industry  products  (Uusivuori  and 
Buongiorno  1991,  Menon 1993 and Hänninen 1998). 
4 
Previous models have often been bivariate,  where a price  variable has been regressed  
on the exchange  rate  using ordinary  least squares (OLS) method (e.g., Kreinin 1977, 
Knetter 1989, 1993, Pick  and Park 1991, Spitaeller  1980, Uusivuori and Buongiorno  1991). 
Using  OLS in the estimation of PT has  been criticised (e.g.,  Menon 1995), because the data 
are usually  nonstationary  (trended) and thus  the OLS estimates  may  suffer  from "spurious  
regression"  problems.  Multivariate approaches  in estimating  PT  have commonly  applied  a 
markup  pricing  equation  (see  e.g.,  Dornbusch 1987,  Hooper  and Mann 1989,  Athukolara 
and  Menon 1994 and 1995,  Hung  et  al. 1993, Kuismanen 1995,  Naug  and Nymoen  1996,  
Vesala 1992,  Hänninen 1998).  The models  of  these studies are  based on the assumption  that 
the markup  is  variable and affected,  for example,  by  the foreign  competitor's  price and 
foreign  demand pressure.  In the study  of  Rockerbie  (1992)  estimable equation  for  analysing  
PT was  derived from the reduced form of an export  demand and supply  model. 
The analysis  of the present  study  is based on a markup  model for  Finnish exports  of 
the two most homogenous  paper industry  products,  i.e. newsprint  and pulp  to the UK and 
Germany.  The Johansen (Johansen  1988, Johansen and Juselius  1990, 1992)  multivariate 
cointegration  method for nonstationary  time series  data is  applied  in the estimation.  This 
method has not  yet  been  widely  used in PT studies (see,  however,  Menon 1993, Naug and 
Nymoen  1996,  Alavapati  et al. 1997,  Hänninen 1998),  although  it has  many advantages  
compared  to the traditional cointegration  methods (e.g., Engle  and  Granger 1987).  For  
example,  in multivariate models it is  possible  to test  the number of  cointegration  vectors  
among the variables and to test the coefficient restrictions in the same cointegration  
framework. 
II. THEORETICAL MODEL  
A model  for  estimating PT for  Finnish paper and  pulp  exports  to  the UK  and Germany  is  a  
markup model that  is  previously  used in analysing  PT e.g., by  Hooper  and Mann (1989),  
Athukolara (1991)  and  Hung  et  al.  (1993).  In  deriving  the price  equation,  it is  assumed  that 
a representative  Finnish exporter  produces  exclusively  for imperfectly  competitive  UK or  
German markets, employs  a  constant  returns  to  scale  technology  and unit production  cost,  
Cf.  The representative  Finnish firm maximizes profit  by  taking  the  competitors'  price  and 
5 
supply  of  newsprint  (or  pulp)  as  given  and by  setting  the price  in Finnish  markkaa (Pf)  as a 
constant  markup  over  unit production  costs (Cf).  With Xf  denoting  exports,  the exporter's  
profit  is  Vf defined as  
Profit maximization yields 
where T|  is  the price  elasticity  of  demand. According  to  Hung et  al.  (1993),  a  more  general  
case,  in which  the competitor's  price  affects  the exporter's  price,  can  be presented  using  the 
concept  of  a  variable markup,  defined by  assuming  that the coefficient T)  depends  on,  among 
other things, price competitiveness  in the export market. The competitiveness  can be 
described as  the relative price  (P0 ER)/Pf where (P0  ER) is  the competitor's  price in the 
exporter's  currency and ER  is  the exchange  rate  (FIM/GBP  in the UK  and HM/DEM in 
Germany).  Thus the price  elasticity  of  demand is 
From (2)  and  (3)  the pricing  behavior  of  the profit  maximizing  exporter  can  be described as 
a variable markup  over  the unit cost: 
The  variable markup,  (j),  depends  on the relative price:  
where 0  is  the relative  price  elasticity  of  markup  and 0 >O.  The constant  markup  is  obtained 
if 0 =  0 and (j)'  =  r|/(ri-l). The second equality  in (5) derives from the log-linear  
approximation  of  the nonlinear function <{>. Substituting  (5)  into (4)  and taking  a logarithmic  
(1) Vf=(Pf-Cf)Xf 
(2) Pf=Cf Tl/(T1-1) 
(3) r\=r\((P„ ER)/Pf)  
(4) Pf  =if Cf  
(5) <|>  =  <|>((/>0  ER)/Pf)  =  V((Po  ER)/Pf  )
9
 
6 
transformation, a relation for  the price  of Finnish newsprint  (or pulp)  in the export  market 
is obtained: 
where 7 =l/(1+0), 8 is  a  constant  and u  is  the disturbance term that captures  all other 
possible  factors.  Symbol  7 (0<  7 <  1) indicates the PT  coefficient (PT  =  -(3 (pf  -  er)/der) 
= 7). The other symbols  are the same as above. Lower-case variables denote logs  of 
corresponding  upper-case variables. The export  price, pf;  should be homogenous  of  degree  
zero  in the exchange  rate  and  the competitor's  price,  and  the equality  restriction  is  imposed  
on  the coefficients of pa 
and er  in the  estimation. 
If  the coefficient of  Finnish cost,  7,  is  unity  in (6),  Finnish export price  in markkaa 
(FIM) is  determined solely  by  Finnish  production cost. Thus,  the markup is  held constant  
after a  change  in the exchange  rate. In this case,  the changes  in er  (FIM/GBP,  FIM/DEM),  
will completely  pass  through  on the Finnish  export price  denoted in foreign  currency and 
PT=l. If 7 = 0,  the changes  in  er  (FIM/GBP,  FIM/DEM)  are absorbed  by  the variable 
markup  and  PT=O.  Thus,  export  prices  in foreign  currency  do not  change  (net  of  any  effect 
exchange  rate  changes  may  have on  prices  through  variations in input costs). This  indicates 
perfect  competition  in the market.  Between the two  extremes, PT may  be incomplete  (0<  
PT<l).  In this case,  the changes  in er will change  the export price, both in FIM and in 
foreign  currency  and exchange  rate  changes  are partially  absorbed into  the variable 
markup.
1
 
III. DATA 
The  data of  the study  are quarterly, seasonally  unadjusted  and cover  the  period  from the 
first  quarter of  1980 to the fourth quarter of 1994 (N=6o).  The Finnish export  prices  (pf) 
'
 The PT may  also  be  affected by factors other than these included in the model (6).  If Finnish 
production  costs include an important  imported  component, it is  evident that  the  degree of  PT of 
exchange  rate  changes  is  less than one,  even  if 7=l. In the  Finnish pulp  and paper  industry,  the 
inputs  are  for  the  most  part of  domestic origin,  which indicates that this  effect is  not  very important. 
(6) pf  =8 +  (1-  7)(er+  pa)  +  7cf  +  u 
7 
are described by  the average FOB (free on  board)  export  unit values of newsprint (CN 
48.01) and pulp  (CN 470321 and CN 470329)  in Finnish markkaa (FIM) and they  are  
obtained from the customs  statistics  of Finland (The  Board of Customs).  The foreign  
competitor's  prices  (p0) are the average CIF  (including  cost,  insurance and freight)  unit 
values (GBP/metric  tn  and DEM/tn)  of  total  newsprint  and pulp  imports  (excluding  imports  
from Finland)  to  the UK  and  Germany.
2
 The source  of  this  data are  the customs  statistics  of 
the UK (CSO) and Germany  (Statistisches  Bundesamt).  Exchange  rates  (er) between the 
Finnish markka and GBP and DEM are quarterly  averages and they  were  obtained from the 
Bank of  Finland. The producer  price  index  (1990=100,  for  commodity  group SIC 341)  
obtained from Statistics  Finland is  used to  describe Finnish production  costs (cf)  for the 
pulp  and newsprint industry,  because better cost  data  are  not  available. This may cause  
uncertainty in estimating  PT,  which  must  be taken into account  in drawing  conclusions. 
The UK  and Germany  are important  markets for Finnish newsprint and pulp,  
accounting  for about 40 %  of  newsprint  and 60 % of pulp  exports  from Finland.  Finnish  
average market  share  from  total newsprint  imports  has  been about 28 % in the UK and 22 
% in Germany during  the period  studied. Finland has  lost  market share in both countries  to 
especially  Sweden. This development  has  been offset by  the acquisition  of  production  
capacity  owned by  Finnish forest  industry  companies  in the UK  market during the 1980 s
and  in Germany  during the 19905. In  pulp  imports  to  the UK  and  Germany  Finnish average 
market shares have been around 15 %. 
In Figure  IA, the export prices  of Finnish pulp and newsprint  denoted in the 
importers'  currency in the UK  and Germany are  graphed  together  with the exchange  rates.  
It can be seen that prices  (in  GBP  and DEM)  and exchange  rates  fluctuate against  each 
other,  which suggests that there has  very likely  been some degree  of exchange  rate  PT  in 
both markets; markka depreciations  have lowered the Finnish price and markka 
appreciations  have  raised it. 
The long-run  development  of  the exchange  rates  FIM/GBP  and FIM/DEM differed 
from each other in the 1980 s.  The Finnish markka appreciated  with respect to  GBP  and 
2
 The  product  codes of newsprint  are  for  British imports  SITC 641.1 and for  German imports  CN  
4801  0010  including  item CN  4801 650 for year 1980, item CN  4801 790  for  years 1981-87 and 
item CN 4801 0090 for 1988 onwards.  Product codes  of pulp (bleached sulphate  pulp)  are  for 
British imports  SITC 251.51 and  SITC 251.52 and for German imports  CN 470321 and CN 
470329.  
8  
depreciated  with respect  to  DEM. In the early 1990 s  the  exchange  rates  have behaved 
similarly  and the large  devaluations of  the markka in 1991 and 1992 can be seen in the 
graphs.  The markup  over  production  costs {pf -  cf  ) in Figure  IB indicates that the 
profitability  margin  has  fluctuated over  time. All in all,  the development  of the variables of  
the study  indicates that studying  the degree  of PT is important  in modelling  the  
international trade of pulp  and paper. 
IV.  ESTIMATION METHOD AND RESULTS 
The  present study  uses  Johansen's cointegration method, which is suitable for  estimation 
purposes  when the variables are nonstationary.  Compared  to  the traditional cointegration  
methods (e.g.,  Engle  and Granger  1987) Johansen's method has  several advantages.  Unit 
roots  are  explicitly  incorporated  in the specification,  and the method takes into  account  
short-run dynamics in estimating  the cointegration  vector. Additionally,  linear restrictions  of 
the coefficients,  e.g., price  homogeneity,  can  be tested in the same cointegration  framework.  
The Johansen's method is briefly summarized below,  but  a comprehensive  description of 
estimating cointegration  vectors  and testing  hypotheses  can  be found in Johansen (1988)  
and (1995),  see also Johansen and Juselius  (1990,  1992). The method uses  a statistical 
model that is  a VAR process  of order  k,  where p  is the number of  variables. 
The VAR model can be reparameterized  in error  correction (EC)  form as:  
where Axt is a  1(0) vector of  the first-differenced variables, /x  is a  vector  of constants, Dt is  
a  seasonal  dummy  and  kis  the lag  length  (k  = 1,...,  N).  In (7)  constant  fi  can  be  restricted  to  
the cointegration  space, if  no linear trend is  present  in the data series,  n is the matrix of 
long-run  coefficients of the lagged  levels. It can  be decomposed  into a  matrix of loadings,  
a, and a matrix of cointegrating  vectors,  p, i.e., n=ap'.  The number of cointegration  
vectors  that exist  among the variables  is  determined by  estimating  the rank of  the matrix  IT 
Johansen (1988)  formulates two likelihood ratio tests for the cointegration  rank, a  
(7) Axt -  rjAx;.! r£_jA%£+i  + +/*  + + et> t-\,...,T, 
9 
maximum eigenvalue  test and a trace  test,  from  which the trace  test  results  are reported  
here. 
Provided that the data series are  cointegrated,  it may be possible  to estimate a 
cointegration  vector  the coefficients of which  describe the long-run  equilibrium  relationship  
implied by  equation (6)  for each destination country  and product. The estimations are  based 
on the VAR model given  by  equation  (7)  with four individual equations.  The suitable lag  
length,  k, of  the VAR model was  determined by  diagnostic  testing  of the data in each case.  
In addition,  the Schwarz (SC) and Hannan-Quinn  (HQ) information criteria (using  
likelihood ratio tests  for  sequential  decreases  in the number of  lags  from 5  to  1  in each  of 
the systems)  were  applied  in  the determination of  the lag  length  of  each of  the VAR models 
(see Doornik and Hendry  1994). 
When the number of  cointegration  relations, r,  is  determined it is  possible  to test  
hypotheses  on the long-run  matrix, IT=aP'. The present  study  tests restrictions on the 
coefficients  (3y  of the cointegration  vectors. Homogeneity  between Finnish export  price,  pf,  
and the other  nominal variables, was tested by restricting  the respective coefficients 
accordingly.  The markup  relationship  was  also tested by  examining  if the estimate for  7 (see  
equation  6)  could be  restricted as unity  in the cointegration  vector.  
Before proceeding  with cointegration  estimation, the individual time series were  
tested for  nonstationarity  using  Augmented  Dickey  Fuller tests  (Dickey  and Fuller 1979).
3
 
Test results  in Table 1 indicate that all the time series are nonstationary  and integrated of 
order one.  The cointegration  approach  is  thus suitable  for modelling  the long-run  exchange  
rate  PT  in British and  German markets  of  pulp  and  paper.
4
 
3
 Also  Phillips-Perron  (Phillips  and Perron 1988) modification of  the  unit root  tests  were  performed.  
As  the results  were found to be uniform regarding  the  null  hypothesis of nonstationarity, the  results  
are not  presented here (available from  authors upon request). 
4
 The VAR-models were tested for stability by using  Chow-tests and examining the recursive  
residuals of  the four models systems.  According  to the  resulting  recursive  constancy statistics,  there 
were  no problems  with the stability of  the  models. Regarding  the cointegrating  relationships,  also  
their stability  was  examined with  plotting  recursive  eigenvalues  of  individual  long run  vectors.  No  
signs  of  instability  was  detected here  either. 
10 
Exchange  rate  effects  in Finnish  newsprint  prices  
Results  for  Johansen's  cointegration  rank  test for  newsprint  price  model of  the UK  market 
are  presented  in Table 2. The VAR model with seasonal dummies, restricted  constant  and 
two lags  was found to be sufficient to  remove  residual autocorrelation of individual 
equations  (diagnostic  test results  are presented  in Appendix 1A). The cointegration  rank 
was found to be one  using the trace test with and without the degrees  of freedom 
correction. 
Also in the German  data, a  VAR model with seasonal dummies,  restricted constant  
and two  lags  was  found to be  a  sufficient description  of  the data generating  process  (see  
Appendix  1  A).  According  to  the cointegration  rank  test (Table  2),  the test  value for  r  = 0  is  
only  slightly  lower than  the  critical value,  so  we can reject  r  =  0 and accept  r  = 1. 
Economic theory  requires  homogeneity  of prices  following equation (6). The 
homogeneity  in nominal variables  pf,  cf  and (p0 + er)  implies  that the coefficients of  pa and 
er  should be equal  and that the coefficient of cf should  equal the difference between the 
coefficients of pf and er in the cointegration  vector.  This assumption  was  tested by 
restricting the coefficients of the UK  and German cointegration  vectors accordingly.  Long  
run  price homogeneity  could be accepted  (Table  3)  and the restricted cointegrating  vectors 
can be written  as 
where pf  is  Finnish price  (FIM/tn),  er  is  exchange  rate,  cf  is  Finnish  production  cost  and  p0 
is competitors'  price (GBP/tn,  DEM/tn).  The  relation  (8)  indicates that from a 10 percent  
change  of the exchange  rate, 6 percent would in the long  run be reflected in the Finnish 
export price  in pounds  sterling  and 4 percent  in the markup  over  domestic  costs. If the 
exchange  rate  changes  were  fully  reflected in GBP  prices,  PT  would be complete  and  prices  
in FIM would remain unchanged.  This hypothesis  was  also tested,  but the restriction  in the 
long-run  newsprint price  relation could not  be  accepted  in the UK  (Table  3).  For  Germany  
the relationship  (9)  indicates that in  the long-run  equilibrium  less  than  one  half of an 
exchange  rate  change  is  reflected  in DEM and  over  a  half in FIM prices.  Consequently,  the 
(8) UK: (3': 1.00/?/= +  o.4o(er+po)  +  0.60  c/ +  1.85, 
(9) Germany: p'  : 1.00/?/= +  o.s4(er+p o)  +  0.46  c/ + 1.32, 
11 
markup  assumption,  i.e. the restriction of production  cost coefficient (7  from equation  6)  to  
one (Table  3)  could not  be accepted  for Germany  either. 
Exchange  rate  effects  in Finnish pulp  prices 
For the pulp  price  model of the UK, the VAR model with seasonal dummies, restricted  
constant  and four lags  was  found  to  be a  sufficient  description  of  data generating  process  
(see  Appendix  IB).  Johansen's cointegration  rank  test  in Table 4 indicate the cointegrating  
rank to be one (at  1 %  level)  or two  (at  5  % level).  However,  since  the degrees  of freedom 
corrected trace  test (not  reported  here) indicated only  one  cointegrating  vector,  we  proceed  
according  to the theoretical model and identify  the  pass-through  coefficient through  
imposing  price homogeneity  in the UK cointegration  vector  (Table 5). 
Also in the German data, a VAR  model with four lags,  seasonal dummies and 
restricted constant  was  found to  be a  sufficient description  of the data generating  process  
(see  Appendix  IB).  According  to the cointegration  rank test (Table  4),  the trace  test result 
for r< 2 is  above the critical value at 5  % level. However, when using  the degrees  of 
freedom corrected version of  this test  (not reported), the result  r=l  is  obtained. We proceed  
accordingly  and make the coefficient restrictions under rank r=l also on the German 
cointegrating  vector.  The markup  pricing  relations in both countries can  be written as  
where the symbols  are  the same as  above. In the UK cointegrating  vector the homogeneity  
restriction  was  accepted  and the coefficient for long  run  exchange  rate  pass-through  of 
0.07  is  obtained for  the pulp  price  (Table  5).  Even  the zero  restriction  for  the PT  coefficient 
could be imposed  for pulp price  in the UK, indicating  that  the Finnish exports  of pulp  have 
been insensitive to exchange  rate  changes.  Instead,  the restriction of markup  relation to 
express  complete  pass  through  (i.e.  PT=l)  is  rejected  in the UK data. 
For  Germany,  the likelihood ratio test  statistic  for  the restricted  price  relationship  was  
rather close to 5  % critical value, so we imposed  the restriction.  According  to  (11)  close to 
(10)  UK: P' : I.oop/=  +O.93(er+/>0)  +  0.07c/ +  0.18  
(11)  Germany: p': I.oopf=  +o.32(er+p o)  +  0.68  c/  +  2.25,  
12 
70 %  of an exchange  rate  change  in the German  market was  thus reflected in DEM  prices  
and over  30 % in prices  denoted in FIM in the long-run  equilibrium.  Consequently,  the 
markup assumption,  i.e. the restriction  of production  cost  coefficient to one (Table  5) 
could not  be accepted  for  German  pulp  price.  If the  magnitude  of PT coefficient is  
interpreted  in terms of market competition, the low (even  zero)  value of it reflects  that the 
UK pulp  market  is more  competitive  than German  market. 
V. CONCLUDING  REMARKS  
This paper examines the  long-run  exchange  rate  pass-through  (PT)  of  Finnish pulp  and 
newsprint  export  prices  in the UK  and Germany.  According  to  the estimation results,  pass  
through  effects  have been  incomplete for  both  product  groups and markets. Thus,  changes  
in the exchange  rates  are  partly  reflected in Finnish pulp and paper export  prices  in foreign  
currency and partly  in the markup  over  production  costs. For  newsprint,  the elasticity  
estimates were relatively  close to each other in both destination countries (in  the range 
0.46-0.60), while for pulp,  the UK  elasticity  (0.07)  was  small as  compared  to  Germany  
(0.68).  
The estimated  PT coefficients  indicate that devaluations have boosted  Finnish export 
demand of  paper in the UK  and  German markets,  and  that the competition  between Finland 
and the other exporters  in these markets  has  been imperfect.  Moreover,  Finnish paper  
exporters  appear to  have pursued  a  midway  pricing  strategy  aimed at maintaining  market 
shares  and profitability  as  the exchange  rate  changes  (see  also Athukorala and Menon 
1995).  In contrast, the pricing  of Finnish pulp  exports  has  been more  clearly  destination  
specific  as  compared  to paper. 
The results of this study  agree with Vesala (1992),  whose PT estimate for Finnish 
paper exports  to  Europe  was  0.66.  For  newsprint,  the present  study  estimated a  slightly 
lower PT coefficient in Germany  than in the UK, which may be related to  the smaller 
market share  in Germany  than in the UK. Germany  is also a large newsprint  producer,  
whose  domestic  producers  may  have market  power  in relation to  the foreign  importers.  One 
of  the underlying  reasons  for  incomplete  PT  in the UK  and  Germany  may  be  the location of 
Finnish paper production  capacity  in these countries. According to Gron and Swenson 
13 
(1996),  firms' export  prices  are unlikely  to change  one-for-one with exchange  rate  
movements  if they  have production  capacity  across  borders. This may also have affected the 
magnitudes  of  the PT estimates for pulp,  which seem not  be connected with Finnish  market 
shares  in the same way  as  those of  the newsprint.  Other things  related to, for  example,  the 
use  of  pulp  as a  rawmaterial of  paper  and a  possibility  to transfer pricing  of  pulp  in the 
paper mills owned  by  Finnish forest industry  in the customer  countries may have effects  on 
the PT  estimates of  pulp. 
The implication  of  this  study  for  Finland's  participation  in stage three of  EMU is  
that some adjustment  difficulties could arise in pulp  and paper industry,  because the 
exchange  rate  between the Finnish markka and the main importing  countries' currencies  
have affected exports.  Along with euro, Finland and Germany  will have a common 
currency,  while with respect  to the UK  the exchange  rate  will be determined between euro 
and GBP. Basing  on the PT elasticities  the effects  of EMU membership  on the competition  
for market share are,  however,  likely  to be smaller in Finnish pulp  and paper industry  than in 
Finnish sawnwood industry.  When Finland joins  the EMU, national exchange  rate  policy  can 
no longer  be used to improve  Finnish price  competitiveness. Hence,  other  means must be 
found to adjust  to future disturbances caused by  shocks  in demand and world market  prices. 
ACKNOWLEDGEMENTS 
We thank Kari  Heimonen, Jari Kuuluvainen,  Juuso Vataja  and an anonymous referee for 
helpful  comments  on  the earlier versions of  this paper. 
14 
REFERENCES 
Alavapati, J.R.R.,  Adamowicz,  W.L. and  Luckert,  M.K. (1997)  A  cointegration  analysis  of 
Canadian wood pulp  prices.  American Journal  of  Agricultural  Economics,  79,  975- 
896. 
Athukolara, P. (1991)  Exchange  rate  pass-through:  The case  of Korean exports  of 
manufacturers. Economic letters,  35,  79-84. 
Athukolara,  P. and Menon, J. (1994)  Pricing  to  market behavior and exchange  rate pass  
through  in Japanese  exports.  Economic  Journal,  104,  271-81. 
Athukolara, P.  and Menon, J. (1995)  Exchange  rates  and strategic  pricing:  The case of 
Swedish  machinery  exports.  Oxford  Bulletin of  Economics and Statistics,  57, 533- 
545. 
The Board of  Customs. Foreign  Trade. 1980-1994. Helsinki 
CSO, Overseas  Trade  Statistics  of  the United Kingdom.  1980-1994. London. 
Dickey,  D. and Fuller,  W.A. (1979)  Distribution of the estimators for  autoregressive  time 
series with a  unit root.  Journal of  American Statistical  Association,  74,  427-431.  
Doornik,  J.A. and Hendry,  D.F. (1994) PcFiml 8.0. Interactive Econometric  Modeling  of 
Dynamic  Systems.  Institute of Economics and Statistics,  University  of Oxford. 
International Thomson Publishing,  London. 436 p.  
Dornbusch,  R.  (1987)  Exchange  rates  and  prices.  The American Economic Review,  77,  92- 
106. 
Engle,  R.F. and Granger,  C.W.J.  (1987)  Co-integration  and error  correction: 
representation,  estimation and testing.  Econometrica, 55,  251-176. 
Feenstra,  R.C., Gagnon,  J.E. and Knetter,  M.M. (1996)  Market share and exchange  rate 
pass-through  in world automobile trade. Journal of  International Economics,  40,  
187-207. 
Gron,  A.  and Swensson,  D.L. (1996)  Incomplete  exchange-rate  pass-through  and imperfect  
competition:  The  effect of local production.  The American Economic Review,  86,  71- 
76. 
Hooper, P. and Mann, C.L. (1989)  Exchange  rate pass-through  in the 1980 s: The case  of 
U.S. imports  of  manufactures. Brookings  Papers  on Economic Activity,  1, 297-337. 
15 
Hung, W., Kim, Y. and Ohno, K. (1993)  Pricing  exports:  a  cross-country  study.  Journal of 
International Money  and Finance,  12, 3-28. 
Hänninen,  R. (1998)  Exchange  rate changes  and the Finnish sawnwood  demand and  price  in 
the UK market.  Silva Fennica,  32,  61-73. 
Johansen,  S. (1988)  Statistical analysis  of cointegration vectors. Journal of Economic 
Dynamics  and Control,  12,  231-254. 
Johansen, S. (1995)  Likelihood-based inference in cointegrated  vector  autoregressive  
models. Oxford  University  Press  Incorporated,  New York.  267 p.  
Johansen, S. and Juselius,  K. (1990)  Maximum likelihood estimation and inference on 
cointegration  -  with applications  to the demand for money. Oxford  Bulletin of 
Economics and Statistics,  52,  169-210. 
Johansen, S. and Juselius,  K. (1992) Structural tests in a multivariate cointegration  analysis 
of the PPP  and the UIP for  UK.  Journal of Econometrics,  53,  211 -244. 
Khosla,  A. and Teranishi, J. (1989) Exchange  rate  pass-through  in export  prices  - an 
international comparison.  Hitotsubashi Journal of Economics,  30,  31- 48. 
Knetter, M.M. (1989) Price discrimination by  U.S.  and  German exporters.  The American 
Economic Review,  79, 198-210. 
Knetter,  M.M. (1993)  International comparisons  of pricing  to market  behavior. The 
American Economic Review,  83,  473-486. 
Kreinin,  M.E. (1977) The  effect of exchange  rate changes  on the prices  and volume of 
foreign  trade. IMF Staff  Papers,  47, 207-229. 
Krugman,  P.R.  (1987)  Pricing  to  market when the exchange  rate  changes.  In  Real-financial  
linkages  among open economies (Ed.)  Sven  W. Arndt and  J.  David  Richardson,  MIT 
Press,  Cambridge,  pp. 49-70. 
Kuismanen,  M. (1995)  Exchange  rates  and  import  prices  in Finland:  estimation of exchange  
rate  pass-through.  Bank  of  Finland Discussion  Papers  17.  33  p.  
McCarl,  8.A., and Haynes,  R.W. (1985)  Exchange  rate  influence on softwood lumber 
trade. Journal of  Forestry,  83,  368-370. 
Menon, J. (1993)  Exchange  rate  pass-through  elasticities for the Monash Model: A 
disaggregated  analysis  of  Australian manufactured imports.  Centre of  Policy  Studies,  
Monash University.  Working  Paper  OP-76,  July.  51  p.  
Menon,  J. (1995)  Exchange  rate  pass-through.  Journal of  Economic Surveys,  9, 197-231. 
16 
Naug, B.  and Nymoen,  R.  (1996)  Pricing  to  market in  a  small  open economy. Scandinavian  
Journal of  Economics,  98,  329-350. 
Phillips,  P.C.B. and Perron,  P. (1988)  Testing  for  unit root  in time series regression.  
Biometrika,  75,  335-346. 
Pick, D.H.  and Park,  T.A. (1991)  The competitive  structure  of  U.S.  agricultural  exports. 
American Journal of  Agricultural  Economics, 73, 133-141. 
Rockerbie,  D.W. (1992) Exchange  rates, pass-through  and Canadian export  
competitiveness:  an analysis  using vector  autoregressions.  Applied  Economics,  24,  
627-634. 
Spitaeller,  E.  (1980)  Short-run effects of  exhange  rate  changes  on  the terms of  trade and 
trade balance. IMF Staff  Papers,  27,  320-48. 
Statistisches  Bundesamt,  Aussenhandel Nach Waren  und Ländern,  1980-94,  Wiesbaden. 
Statistics  Finland, Bulletin of Statistics  (SVT),  1980-1995, Helsinki.  
Uusivuori,  J.  and  Buongiorno,  J.  (1991)  Pass-through  of exchange  rates  on  prices  of forest 
product  exports  from the United States to  Europe and Japan. Forest Science, 37, 
931-948. 
Vesala,  J.  (1992)  Incomplete  exchange  rate  pass-through  and hysteresis  in trade. A survey 
of  recent  theories  and  empirical study  of  export  pricing  of Finnish paper manufactures. 
Bank of  Finland Discussion  Papers  29. 76 p. 
17 
Figure.  1A. Nominal CIF  -  prices  of Finnish newsprint and wood  pulp  in the UK  (GBP/tn) 
and  Germany (DEM/tn)  with the exchange  rates,  FIM/GBP  and  FIM/DEM. 
Figure.  1B. Markup.  Relation between Finnish unit price  of exports  (FIM/tn)  for  newsprint  
and pulp  and production  price  index (1990=100)  for  paper and pulp  industry.  
18 
19 
Table 1. Results  for  unit root tests. 
1 For  critical values,  see  Dickey  and Fuller (1979).  
Note: Constant or  constant  and trend (T)  included in  test equation  and N=  number of  lags  
needed in test  equation  to  remove  autocorrelation. *(**) denotes rejection  of  null  
hypothesis  of nonstationarity  at 5  % (1  %) level. 
Table 2.  Results  for the cointegration  rank  test in British and German newsprint  models. 
Level First  difference 
Newsprint:  export  price  from Finland -1.78 (T,  N=0) -6.43** 
Wood pulp:  export  price  from Finland -1.65  (T,  N=0) -4.91** 
Production cost -2.62 (N=0) -AA1** 
Exchange rate,  FIM/GBP -2.34 (N=0) -6.57** 
Exchange  rate,  FIM/DEM -1.61 (T,  N=0) 
Newsprint:  competitor's  price  in UK -2.56 (N=0) -8.62** 
Newsprint:  competitor's  price  in Germany  -0.93 (N=l) -6.16** 
Wood pulp:  competitor's  price  in UK -3.16 (T,  N=0) -8.33** 
Wood pulp: competitor's  price  in Germany -1.53  (T,  N=0) -3.98** 
United Kingdom  Germany  
H0:r<i  Eigenvalues  Trace test Eigenvalues  Trace test 95% 
Xi statistic Xi statistic critical value  
r=0 0.43 59.8 0.39 52.5 53.1 
r<l 0.31 31.5 0.23 24.5 34.9 
r<2 0.17 13.2 0.15 10.2 20 
r<3  0.08 4.0  0.02 1.03 9.2 
20 
Table 3. Tests for  restrictions on  the newsprint  price  relations in the UK and Germany.  
*
 %2 critical value  for at 5  % level is  5.99 with two  degrees  of  freedom and 7.82 with 3 d.g.f. 
Symbols:  pf  denotes  the  export  price  from Finland,  er  is  the  exchange  rate, cy  is  Finnish production  
cost  and p„  is  the competitor's  price. 
Table  4. Results  for the cointegration  rank  test  in British  and German pulp  models. 
Pf er cr P" Constant  *2(2uy 
UK: 
Price 1.00 -0.40 -0.60 -0.40 -1.85 1.41 
homogeneity  [0.49] 
Complete 1.00 0 -1.00 0 18.8 
pass-through [0.000]  
Germany:  
Price 1.00 -0.54 -0.46 -0.54 -1.32 5.25 
homogeneity  [0.07] 
Complete 1.00 0 -1.00 0 9.14 
pass-through [0.03]  
United Kingdom Germany  
H
0:r<  i Eigenvalues  Trace test Eigenvalues  Trace test 95% 
A.i statistic Xi statistic critical value 
r=0 0.42 71.6 0.45 80.0 53.1 
r<l  0.34 40.7 0.36 46.9 34.9 
r<2  0.19 17.7  0.24 21.8 20.0 
r<3  0.10 5.9 0.11 6.6 9.2 
21 
Table 5.  Tests  for restrictions  on the pulp  price  relations in the UK  and Germany.  
*
 y2  critical value for at  5  % level is  5.99  with two  degrees  of  freedom and 7.82 with 3 d.g.f. 
Note: For  symbols,  see Table 3. 
Pr er Cf Pn Constant  7.2(2/3)* 
UK: 
Price 1.00 -0.93 -0.07 -0.93 -0.18 0.06 
homogeneity  [0.97]  
Complete  1.00 0 -1.00 0 19.93 
pass-through  [0.000] 
Germany:  
Price 1.00 -0.32 -0.68 -0.32 -2.25 6.74 
homogeneity  [0.03] 
Complete  1.00  0 -1.00 0  17.15 
pass-through [0.000] 
22 
APPENDIX lA. 
Misspecification  tests for residuals from Johansen's cointegration  estimation of  British  and 
German newsprint  models with two  lags,  seasonal dummies and restricted  constant. 
Note: Values in square brackets  are  marginal significance  levels and 
*
 indicates that the null 
hypothesis  is rejected  at  the 5 percent level. 
a>
 Autocorrelation of the residuals of  individual 
equations  and a whole system  was  examined  using  the F-form of the Lagrange-Multiplier  (LM)  test,  
which is  valid for  systems  with  lagged  dependent  variables. 
b>  Heteroskedasticity  was tested using  
the F-form of the LM test  against 4th order autoregressive  conditional heteroskedasticity.  
Normality  of  the  residuals of  individual equations  and a whole system was  tested with the Doornik-  
Hansen test  (Doornik and Hendry  1994). For  further  detail and  test  references, see Doornik and  
Hendry  (1994).  
Symbols: the export  price  from Finland,  er  is  the exchange  rate,  cj is  Finnish production  
cost  and p„  is the competitor's  price.  
Equation  Tests  for residuals and standard errors  
Autocorrelation 
a>
 Heteroskedasticity  b>  Normality c> Standard errors  
F
AR(4,40)  FARCH(4,36)  X
2
(2)  Oe 
UK: 
Apf 0.16 [0.96] 0.58 [0.68]  1.84 [0.40] 0.02 
Aer 0.86 [0.50]  0.16 [0.96] 7.36 [0.03] 0.03 
Acf  1.03  [0.40]  0.07 [0.99] 3.23 [0.20] 0.02 
Ap„ 1.30 [0.29] 0.05  [0.99] 47.19 [0.00] 0.01 
System:  VFAr(64,100)=  =1.17 [0.24] Vx
2 (8)=39.26 [0.00]*  
Germany: 
Apf 0.29 [0.89] 0.64 [0.64] 0.84 [0.65] 0.02  
Aer 0.73 [0.58] 2.19 [0.09] 5.23 [0.07]  0.03 
Acf 2.53 [0.06] 0.49 [0.74] 3.66  [0.16] 0.02 
Ap„ 1.77 [0.15] 5.48  [0.00]*  5.77 [0.06] 0.02 
System:  vfAR(64 >  1 00 ) =  1  -25  [0.16] V%
2
(8)=1 1.79 [0.17] 
23 
APPENDIX 1B. 
Misspecification  tests  for  residuals  from Johansen's  cointegration  estimation of  British  and 
German pulp  models with four lags,  seasonal dummies and  restricted  constant.  
Notes: Values in  square brackets  are marginal  significance  levels and 
*
 indicates that  the null 
hypothesis is  rejected  at the 5 percent level. 
a> Autocorrelation of the residuals of  individual 
equations  and a whole system  was examined using  the F-form  of  the Lagrange-Multiplier  (LM) test,  
which is valid for systems  with lagged  dependent  variables. Heteroskedasticity  was  tested using 
the F-form of the LM test against 4th order autoregressive  conditional heteroskedasticity.  
Normality  of the residuals of  individual equations  and  a whole system was  tested with  the  Doornik- 
Hansen test (Doornik and Hendry  1994). For further detail and  test references, see Doornik and  
Hendry  (1994). 
For symbols,  see  Appendix  IA. 
Equation  Tests  for residuals and standard errors 
Autocorrelation 
a>
 Heteroskedasticity  
b>
 Normality  
c> Standard errors  
F
AR(4,32)  FarCH(4.28)  X
2
(2)  
UK: 
Apf 0.07 [0.99] 0.88 [0.48]  9.06 [0.01]*  0.07 
Aer 3.25 [0.02]* 0.21 [0.93] 3.69 [0.16] 0.03 
Acf 1.60 [0.19] 0.38 [0.82] 3.49 [0.17] 0.01 
APo 1.38 [0.26]  3.71 [0.02] 39.4 [0.00]*  0.13 
System:  VFAR (64,68)  = 0.97  [0.55] VX
2
(8)=69.34 [0.00]* 
Germany:  
Apf 0.13 [0.97] 1.30 [0.29] 3.97 [0.14] 0.06 
Aer 0.69 [0.60] 1.15 [0.36]  5.83 [0.05] 0.03 
Acf  1.64 [0.18] 0.26 [0.90]  2.73 [0.26] 0.01 
Ap<> 2.55 [0.06]  0.35  [0.84]  1.02 [0.60]  0.04 
System:  VFAR( 64>68)=  1.07 [0.39] VX
2
(8)=22.78 [0.01]* 




ISBN 951-40-1650-5  
ISSN  0358-4283  
Hakapaino Oy 1998