ESTIMATING GHANAS EQUILIBRIUM REAL EXCHANGE RATE: 19652004

1
ESTIMATING GHANAS EQUILIBRIUM REAL EXCHANGE
RATE 1965-2004

• Curtis E. Youngblood
• Lawrence K. Apaloo

2
OVERVIEW

• Relevance of the RER
• Update analysis of the ERER using annual data
  from 1965-2004
• Use single equation error-correction model to
  estimate the ERER
• Calculate misalignment of observed RER from its
  equilibrium
• Estimate short-run dynamics of the model to
  explain misalignment

3
Relevance of the RER

• RER is one determinant of a countrys
  competitiveness in international markets
• Powerful price signal to producers and consumers
  of exportables, importables, and nontraded goods
• Overvalued RER makes imports cheaper than they
  should be and reduces the profitability of
  producing exports and import-substitutes
• To determine whether the RER is fairly valued,
  have to know what its equilibrium value is

4
Modern View of the Equilibrium RER

• Simultaneously achieves
• Internal balance, i.e., clearing of the domestic
  market for nontradable goods and services
• External balance, i.e., a current account balance
  consistent with long-run sustainable capital
  inflows
• Changes over time in response to permanent
  changes in the fundamentals that determine it
• Under purchasing power parity, the ERER remains
  constant

5
Fundamental Determinants for Ghana

• Terms of Trade
• From African Development Indicators, updated
  through 2004 by applying changes in TOT
  reported in various IMF documents
• Permanent improvement appreciates the ERER as
  long as income effect dominates substitution
  effect
• Net Capital Inflows
• Measured as negative of the trade balance plus
  changes in reserve assets (includes private
  remittances and grants)
• Permanent increase appreciates the ERER
• Commercial Policy (Openness)
• Proxy for import and export taxes and other trade
  restrictions
• Measured as ratio of exports plus imports to GDP
• Permanent improvement favoring increased
  integration with world economy depreciates the
  ERER

6
Other Fundamentals

• Theory suggests additional fundamentals
• Domestic absorption ( of GDP)
• Share of public expenditure on nontradables in
  total public expenditure
• Total public expenditure ( of GDP)
• Technological change (used by Morrissey et al.)
• For most part, these have not been found
  significant for Ghana

7
Measuring the RER

• Need a multilateral RER to capture effects of
  changes in exchange rates important to Ghana US
  dollar, UK pound, euro
• First compute bilateral real exchange rates for
  these currencies
• BRER(i, t) E(i, t)xP(i, t)/P(t)
• E(i, t) is nominal exchange rate in cedis per
  foreign currency unit for currency i at time t
• P(t) and P(i, t) are CPIs for Ghana and country
  i, respectively, at time t
• Next, calculate weighted average of bilateral
  RERs
• MRER(t) w1BRER(US, t) w2BRER(UK, t)
  w3BRER(eu, t)

8
CHOICE OF WEIGHTS TO CALCULATE MRER

• Cocoa, gold, oil priced in dollars
• Accounted for 45 of merchandise trade in first
  half of 2003
• Direction of Trade Statistics (DOTS) method masks
  this reality
• e.g., imports from and exports to Nigeria
  accounted for 20 of total merchandise trade,
  mostly oil
• Weighting the naira by 20 would grossly
  overstate its importance
• Large divergence in RERs when dollar/pound and
  dollar/euro rates change significantly (2002
  onwards)

9
(No Transcript)
10
Sustainable Changes in Fundamentals

• Only permanent, or sustainable, changes in the
  fundamentals affect the ERER
• Use Hodrick-Prescott method to decompose values
  of each fundamental into a growth component and a
  residual cyclical component

11
RER and Fundamental Determinants
12
Terms of Trade

• Secular (permanent) decline of 50 between 1982
  and 2004
• Would tend to depreciate ERER

13
Trade Policy (Openness)

• Openness has increased dramatically since 1983
• Would tend to depreciate ERER

14
Capital Inflows

• Increased from zero in 1983 to 20 of GDP in 2004
• Inflows have nearly doubled as of GDP from 1999
  to 2004
• Would tend to appreciate ERER

15
Estimating the ERER

• First, capture any long-run relationships between
  the MRER and the fundamentals of the ERER by
  running a cointegrating regression
• Plug permanent (sustainable) changes in the
  fundamentals into the regression to obtain an
  ERER index
• Account for non-uniqueness of the constant term
  and rescale the time path of the ERER index
  calculated above
• Result is the ERER

16
Cointegrating Regressions

• Signs of coefficients significant and consistent
  with theory across all studies (Elbadawi defines
  RER as foreign currency price per cedi)
• Increases in TOT and KFLOW appreciate ERER
• Increases in openness depreciate ERER
• Test statistics reject null of non-cointegration

17
Adjusting the ERER Index

• Solution of the theoretical model that generates
  the cointegrating relationship above does not
  yield a unique intercept for mathematical reasons
  (Elbadawi, 1994)
• Have to normalize the ERER index to determine the
  position of the ERER in a given year
• Apply resource balance criterion (as in
  Elbadawi) scale the ERER index so that its
  average equals the average of the observed RER in
  years in which the resource balance is close to
  its equilibrium value (within /- 1)

18
(No Transcript)
19
(No Transcript)
20
Comments on Misalignment

• Severe overvaluation of late 70s and early 80s
  is apparent
• Overvaluation in latter half of the 90s caused
  in part by BOGs inappropriate use of the
  exchange rate as a nominal anchor (precursor to
  the speculative assault of 1999/2000)
• Assault restored equilibrium (no overshooting)
• Overvaluation approaching 20 in 2003 and 2004

21
Short-run Error Correction Model (ECM)

• In long run, RER will tend towards its
  equilibrium
• In short run, divergences are likely because of
• transitory or cyclical changes in the
  fundamentals
• policies that impede convergence (e.g., excessive
  deficits and domestic credit expansion, exchange
  rate pegs, etc.)
• An ECM incorporates the long-run (cointegrating)
  relationship between the ERER and its
  fundamentals and short-run influences that cause
  the observed RER to deviate from equilibrium

22
Error-Correction Model of the RER

• All variables significant and have expected signs
• Good explanatory power
• Permanent changes in fundamentals embodied in
  error correction term, CIRES-1
• 28 of divergence between observed RER and ERER
  will be closed in one year after 5 years 81 of
  the gap will be closed
• Slower rate of convergence than found in earlier
  studies, but very similar to Morrissey study

23
Error-Correction Model of the RER (continued)

• TOT above sustainable level by 8 in last 2 years
• tends to appreciate the MRER temporarily
• Openness below its sustainable level by 9 in
  last 3 years
• tends to appreciate the MRER temporarily
• Large transitory capital inflow (6 of GDP above
  permanent level) in 2001 perhaps working its way
  through the system
• would tend to appreciate MRER temporarily