Vous êtes sur la page 1sur 9


1.The macroeconomic balance approach addresses the requirements to achieve internal

and external balances simultaneously. Under the macroeconomic balance (MB) approach
the exchange rate are perceived to be in equilibrium when the associated underlying currant
account coincide with the equilibrium CA position for each country. A key distinction of the
MB approach from other approaches to assessing exchange rate is that it focuses explicitly on
the current account—or more precisely on whether the outlook for current account at
prevailing REERs is consistent with the CA norm. The notion of a CA norm masks its
underlying substance as the norm implies consistency of the current account with the existing
saving-investment balance, or, as it was presented in the early days of the MB approach, with
the capital account.1

2.The fact that sum of current accounts for the world should equal zero enforces global
consistency in the MB approach. In a multilateral framework, the sum of CA position
implied by the configuration of equilibrium exchange rates is globally consistent, or, in
practical application, that a simultaneous movements in all exchange rates from their
prevailing to their equilibrium levels would not significantly affect the convergence of the
global current account to zero. Thus, the equilibrium REER for each country under the MB
approach should ultimately be consistent with the sum global all current accounts of all
countries equal to zero. The deviation of the REER of an individual country from the
equilibrium would signal then that the country does not contribute to the global
macroeconomic balance.

3.The CGER MB approach consists of three steps.2 First, an equilibrium and multilaterally
consistent relationship between CA balances and a set of fundaments is estimated with panel
econometric techniques. Second, a CA norm is computed for the country under consideration
based on the relationship estimated in step one as a function of the medium-term levels of
fundamental for this country. Third, the REER adjustment that would close that gap between
the CA norm and the UCA is calculated using the relevant elasticities. The coefficients
estimated in stage one capture medium-term tendencies in the co-movements of the CAs with
the underlying fundamentals in 54 developed and transition economies that represent [80]
percent of global trade. It can be argued that this is a representative set of countries that are
ultimately responsible for global balances, as economies with a larger international presence
has larger multilateral effect of REER of other countries.

4. The coefficients in the CGER MB approach can not be applied directly applied to a
LIC. The coefficients obtained in step one of the CGER estimations reflect co-movements of
the CAs with the underlying fundamentals in developed and transition economies.
Isard and Mussa in Exchange rate Assessment (1998) for a historical discussion.
Lee et al (2008).

Application of these coefficients to the medium-term values of fundamentals of a LICs with

the view to estimate the CA norm would create an internal methodological conflict and
would impair the interpretation of the results. Also, a direct estimation of the coefficients for
a panel of LICs or their large groups (i.e. nonoill exporters and oil exporters) poses the
problem of multilateral consistency as for a meaningful estimation of the CA norm the sum
of CA in the panel should converge to zero. Finally, a direct estimate of the CA norm by
relating is a certain set of fundamentals essentially collapses the MB approach into the ERER
approach with the only difference that the CA, and not the REER is the dependent variable.3

5.Data requirements under the MB approach are essentially the same as under the ES
approach. With the notable exception of the CA norm, all other variables needed for the
estimations have already been discussed in section C. An important difference is that the MB
is dynamic concept as it allows to estimate a time-bound trajectory of the exchange rate
adjustment needed to bring the UCA in line with its norm, whereas the ES approach is
essentially static addressing the question of what the exchange rate adjustment is needed
today to bring the UCA in line with the benchmark NFAP. Therefore, the UCA evaluated
under the ES approach at a fixed point of time (2007), under the MB approach should
presented as an annual medium-term projection. In 2008-2013, the UCA differs from the
baseline CA by the magnitude of temporary factors that affected it in 2005-2007 but still have
visible impact in the medium term. Beyond the medium term, the UCA coincides with the
baseline CA as the lagged effects of the temporary factors die off. Macroeconomic
assumptions and elasticities needed to calculated the change in the exchange rate are the
same as in the ES approach. Finally, the same sensitivity tests can be applied to the results.

6.This paper proposes a simple and transparent non-econometric method to evaluate

the REER under the MB approach and applies it to WAEMU.

A. Inverted DSA

7.Several features make DSAs particularly appropriate for the MB approach. First,
DSA are focused on underlying policies, not only on sustainability of the debt. When the
debt-to-GDP ratio is on a non-explosive path, the solvency condition is automatically met. In
fact, the DSA assesses whether the policies summarized in the path for the primary fiscal
balance and in the non-interest CA are consistent with a sustainable debt path. Second, DSAs
looks at debt dynamics rather that at debt levels. Under the DSA, a country is solvent is the
NPV of the non-interest CA is equal or exceeds the NPV of its external debt. Third, DSA
capture the notion of the use of underlying, rather that the current, levels of key variables as
it explicitly requires that the standard templates be adjusted to the circumstances of a
particular country, for example by shortening or extending the averaging period and
excluding non-representative and one-off events, which may distort the outcome. by CA as
Finally, although produced by country teams, DSAs are based on macroeconomic
These and other difficulties are the reason that the MB CGER methodology for LICs is currently not available
and there are very few applications in the literature Chinn and Ito (2005)

projections, which are derived from multilaterally consistent global assumptions and the
latest WEO projections for oil prices, nonfuel commodity prices, interest rates, and exchange
rates, which assures their multilateral consistency.

8.A standard DSA presents a number of scenarios, of which only the baseline is
appropriate for the MB analysis. The baseline presents the evolution of keys
macroeconomic variables, which is consistent with a sustainable level of external debt. In
principle, the baseline scenario should reflect an unbiased projection of debt sustainability
derived from a comprehensive and realistic macroeconomic framework prepared in
conjunction with the debt analysis. The “alternative scenarios” included in a standard DSA
are not appropriate for the MB approach, as they do not reflect a comprehensive and
consistent alternative macroeconomic scenarios, but rather illustrate the impact on debt
sustainability of a mechanic change in one (the lower level of concessionality scenario) and
several (the historical scenario) variables.

9.The appropriateness of the baseline scenario for the under the MB approach depends
on whether it reflects current policies. This can be deduced from the underlying
assumptions, which usually clearly spelled out in the DSA. Since DSA are based on WEO
projections for the global environment they by construction incorporate the assumption of
unchanged current policies and unchanged REER, al lest for the 5-year WEO horizon.
However, country-specific assumptions can include measure that are expected to be taken in
the medium term, but have not been neither approved not inherently imminent. In practice,
very rarely DSAs analyze debt sustainability conditional on assumptions of sharp shift in
fiscal policies, rapidly scaled up aid, ort massive productivity growth, as any of such
assumptions require explicit justification and may not pass the reality check.

10.The status of LICs in the HIPC process broadly ensures that DSA are indeed based
on current policies. DSAs for pre-decision point countries are supposed to explicitly show
that on current policies the trajectory of external debt is not sustainable and key
macroeconomic variables deteriorate without a sizeable debt relief. For countries in the
interim period, the baseline includes the interim debt relief, which is also representative of
current policies as fiscal policies and CA projections, at least until the assumed completion
point, factor in debt service payments reduced by the amount of the interim debt relief.
Finally, for the post completion point countries, DSA already incorporate debt relief in the
baseline scenarios and assume new debt accumulation at a certain reasonable rate discussed
and agreed with the authorities today, i.e. they are again predicated on current policies.

11.For countries with unsustainable external debt, the baseline DSA scenario is not
usable for the purposed of the ES approach. In these cases, a full-fledged alternative
scenario based on an alternative macroeconomic framework but still on multilaterally
consistent global WEO assumption is usually available as part of DSAs.

B. Derivation of the CA norm

12.A proxy for the CA norm can be derived directly from the DSA baseline scenario. A
standard DSA calculates two variables directly relevant for the evaluation of the CA norm—
the debt stabilizing non-interest CA and the primary fiscal balance. These are the CA and
primary balance required to keep the debt-to-GDP ratio constant if all other relevant variable
included in the DSA remain at their last year level. The level of these variable explicitly
addresses the country’s solvency condition, i.e. establishes the level of the non-interest CA
and primary balance needed to keep its external debt constant as a share to GDP.

13.The CA norm can be extracted by inverting a standard DSA. A DSA answers the
question whether the level country’s external debt is sustainable and what it the resulting
level of CA deficit and primary balance a country can afford to run in the medium and long
run to help preserve debt sustainability. An inverted DSA answers a reverse question—what
level of the CA and the primary balance are needed to stabilize the debt-to-GDP level at
certain sustainable level. For a given time-path of reserve accumulation and non-debt
creating capital account flows, the CA that stabilizes external debt at non-explosive and
sustainable path may be viewed as a dynamic CA norm.

complications arise
1. Debt-accumulating with such derivation of the CA norm. First, since the CA

is financed not only by debt creating inflows, but also by non-debt creating inflows, selecting
the debt creating flows that would Debtstabilize external debt, my not be sufficient to ensure

sustainability of the CA. However, in the case of most LICs, WAEMU included non-debt-
Figure 2. Debt-reducing country
creating flows represent relatively minor and a relatively stable share in terms of GDP of all
Percent of GDP

balance of payment flows, in particular in the projected period. Debt In WAEMU, such flows so
not exceed 2.5 percent of GDP and does not change much through the projected period.
CA norm
Therefore, such flows may be treated as a constant at a steady state. Second, the DSA uses as
Percent of GDP

an input the baseline CA from the macroeconomic

Baseline CA framework. Therefore the CA reflects the
assumptions of the framework but does not includes no feedback from the `
DSA. Third, the
baseline CA used in theT ime DSA inevitably reflects the overall debt
Baseline CAsituation, with differs
substantially in countries with a sustainable level of external debt and countries with a large
CA norm
debt overhang. Finally, the norm level, but not the trajectory, will be different for debt-debt-
accumulating countries (Figure 1) and debt-reducing countries (Figure 2). T ime
WAEMU includes

15.The CA norm can be derived by inverting a standard DSA in three steps. First,
identify a DSA scenario with a sustainable debt profile. For post-debt relief countries this is
usually the baseline scenario. For pre-debt relief countries, this is usually an alternative
scenario derived from an alternative macroeconomic framework, which includes debt relief
needed to achieve sustainability. Second, calculate the CA balance that stabilizes the debt-to-

GDP ratio4. The DSA template provides all needed variables, although it calculates the non-
interest CA that stabilizes the debt-to-GDP ratio. By adding back the projected interest
payments and subtracting the change in external debt for the relevant year one can arrive at
the CA needed to keep external debt constant in terms of GDP. Third, since the current
account can be financed by both debt-creating and non-debt-creating flows, add to the CA
calculated in the previous step all non-debt creating flows, assuming they are roughly stable
as a share to GDP.

16. Application of the three step approach to WAEMU allowed to establish the CA
norms. Current baseline DSA scenarios were used for post debt relief WAEMU countries
with highly sustainable debt levels (Benin, Burkina Faso, Mali, Niger and Senegal).
Alternative DSA scenarios derived from a comprehensive alternative macroeconomic
frameworks were used for pre-debt relief countries (Cote-d’Ivoire, Guinea Bissau, and Togo).
For the purposes of the estimation of the CA nor, the first group was treated as debt-
accumulating group of countries. The second group was heterogeneous as external debt of
Cote d’Ivoire (the largest WAEMUU economy) is not substantially above the sustainability
levels and in fact converges to sustainability under the baseline scenario in the long run.
Debts of Guinea-Bissau and Togo are clearly unsustainable and they were treated as a group
of debt-reducing countries. Debt-stabilizing CA was derived by weighting debt-stabilizing
CAs of individual countries by their relative PPP-based GDP shares in the WAEMU region.
Finally, non-debt creating flows were added to arrive to the CA norm.

17.The resulting CA can be interpreted as a dynamic CA norm for a country. This

derived CA norm is multilaterally consistent as it is based on WEO global economic
assumptions. The norm is derived from a coherent macroeconomic framework for the
country and thereof is consistent wit its fundamentals. It also reflects external stability
because it is derived from the CA needed to stabilize external debt and debt-creating flows.
Finally, the norm also reflects domestic stability because it is consistent with a sustainable
primary fiscal balance.

18.The dynamics of the medium-term and long-term CA norms are distinctly different.
As should be expected in the medium term, the CA nor does not exhibit any particular trend
(Figure 3), hovering in the range of 4 to 5 percent of GDP. In the long run, the CA norm
clearly indicates a reduction in the CA deficit (Figure 4), suggesting that overall lower CA
deficit is needed to ensure along term macroeconomic stability.

Long-run, constant balance that stabilizes the debt ratio assuming that key variables (real GDP growth,
nominal interest rate, dollar deflator growth, and non-debt inflows in percent of GDP) remain at their levels of
the last projection year.

Figure 3. WAEMU: medium-term CA norm Figure 4. WAEMU: long-term CA norm

(Percent of GDP)
(Percent of GDP) 0
2007 2012 2017 2022 2027
2007 2008 2009 2010 2011 2012 2013 -1
-4 -5

-5 -6

-6 -7

19.Two characteristics of these CA norms are worth mentioning. First, the norm is a
projected trajectory of the CA, which may no be necessarily an optimal trajectory. For post-
debt relief countries, which can in principle afford accumulate new external debt to the
threshold of sustainability at a faster pace than in the baseline, the CA norm may be too low,
as these countries can afford higher CA deficits. For pre-debt relief countries, which need to
reduce external debt at least to the sustainability threshold, this trajectory my be too high as
these countries may need to run lower CA deficits or even surpluses. Second, the CA norm
changes over time. This reflects mainly accumulation or reduction of debt-creating and non-
debt creating liabilities in each year, as well as the fluctuations of the underlying CA.
However, after temporary factors die out and the underlying CA converges to its long-term
baseline projections, the dynamics of the norm starts reflecting mainly the assumption on
debt-creating flows.

20.The remaining steps to calculate REER misalignment are essentially the same as
under the ES approach discussed in the previous section, with two caveats:
• The MB approach uses both the actual (2007) and projected (2008-2027) values of the
UCA, whereas the MB approach focuses only on the 2007 UCA. The medium-term
projections of the UCA deviates from the baseline CA, as during the next five years
temporary factors are still active and affect CA. The long-term projections of the UCA
coincide with the baseline CA as all temporary factors die off (Figure 5).
Figure 5. WAEMU: baseline and underlying
current accounts •
Export and import elasticities are
2027 assumed the same for the medium
-3 2007 2013
Percent of GDP

and long term. Although this is a
-5 strong assumption, as elasticities in
-6 the long run should be generally
-7 higher, it can be addressed by way of
Baseline Underlying
a sensitivity test. In any case,
applying for consistency the
elasticities used in the ES approach, the Marshall-Lerner condition is met as the sum of
price elasticities is greater that unity.

21.Finally, applying the elasticity to the gap between the CA norm and the underlying
CA allows to establish the degree of REER misalignment for each projected year.
Distinction between medium-term and long term can be again made to usefully capture the
dynamic characteristics of the MB approach. In the medium term, although numerically the
REER can be viewed as overvalued, the misalignment does not exceed on average 5 percent
and is erratic from year to year. The magnitude of misalignment fluctuates drastically from 8
percent in 2007 to 0.2 percent in 2008, then increases and falls again.

Figure 6. WAEMU: Medium term REER misalignment

(Percent left scale, percent of GDP right scale)

2007 2008 2009 2010 2011 2012 2013
CA norm
Underlying CA

22.In the long run, accumulated misalignment accumulates and becomes clearly
pronounced. On current trends and with current policies, REER overvaluation may exceed
10 percent in 2013 and continue growing thereafter reaching 15 percent.

Figure 7. WAEMU: Long-term REER misalignment

(Percent, left scale; percent of GDP, right scale)

10 Overvaluation

CA norm
2007 2012 2017 2022 2027
Underlying CA

23.Therefore, based on the MB approach there are early indications that the REER
follows the trend, which may eventually misalign it with fundamentals. In the medium
term, the MB approach points only to a modest overvaluation of 1-5 percent, which is at the
margin of the confidence intervals and can not be established beyond reasonable doubts.
However, in the long run the compounding exchange rate overvaluation may become
substantial and misalign the REER with fundamentals by 10-15 percent presenting a risk for
external stability.


Exchange Rate Assessments: CGER Methodologies

Author/Editor: Lee, Jaewoo ; Milesi-Ferretti, Gian Maria ; Ostry, Jonathan David ; Prati,
Alessandro ; Ricci, Luca Antonio
Series: Occasional Paper No. 261

Methodology for Current Account and Exchange Rate Assessments

Author/Editor: Isard, Peter ; Faruqee, Hamid ; Kincaid, G. Russell ; Fetherston, Martin
Series: Occasional Paper No. 209
Published: December 28, 2001

Exchange Rate Assessment: Extension of the Macroeconomic Balance Approach

Author/Editor: Isard, Peter ; Faruqee, Hamid
Series: Occasional Paper No. 167
Published: July 24, 1998