Exchange Rate Movements in Indonesia Determinants
Exchange Rate Movements in Indonesia Determinants
Exchange Rate Movements in Indonesia Determinants
net/publication/265534141
Article
CITATIONS READS
7 548
2 authors, including:
Sahminan Sahminan
Bank Indonesia
20 PUBLICATIONS 93 CITATIONS
SEE PROFILE
All content following this page was uploaded by Sahminan Sahminan on 26 June 2015.
2008
Exchange Rate Movements in Indonesia:
Determinants, Effects, and Policy Challenges
September 2008
In this paper we provide an overview analysis of the role of the exchange rate in the
Indonesian economy. In particular, we have explored three main issues. The first is
how the exchange rate is determined in Indonesia, the second is why the exchange
rate matters to the Indonesian economy, and the third concerns the policy challenges
presented by exchange rate fluctuations in the post-crisis period. We find some
evidence to suggest that non-market factors continue to play a role in influencing the
exchange rate, although many of the vulnerabilities typically associated with rupiah
movements appear to have declined.
The views expressed do not reflect the views of Bank Indonesia or Reserve Bank of Australia.
Corresponding author: Sahminan Sahminan, Bank Indonesia, Jl. M.H. Thamrin No.2, Jakarta.
Email: [email protected].
2
1. Introduction
Emerging market economies are often reluctant to allow their currencies to fluctuate freely
(e.g. Calvo and Reinhart 2000), because of the potential for sharp exchange rate movements to
exacerbate inflationary pressures and financial sector vulnerabilities. Although Indonesia
adopted a floating exchange rate regime in the wake of the 1997 crisis, the value of the rupiah
continues to be at least partly determined by non-market forces. Indeed, it appears that the
central bank has been prepared to intervene in order to prevent sharp currency movements in
both directions, while buying foreign exchange on a more sustained basis in an attempt to
foster export competitiveness through the maintenance of a low, stable exchange rate.
In some instances the Indonesian authorities have openly acknowledged their role in managing
the exchange rate, particularly when intervening to support the currency in periods of sharp
depreciation. In these cases intervention often appears to be carried out under the pretext of
“keeping volatility in check” (see section 2.2.1). Indeed, it seems to be widely accepted that
intervention is a valid policy tool, which can be used, “possibly in conjunction with monetary
policy, to stabilise market expectations, calm disorderly market, and limit unwarranted short-
term exchange rate movements because of temporary shocks” (Warjiyo 2005). This may at
least partly reflect public pressure on the central bank to manage the exchange rate at an
appreciated level. If it is true that “the public monitors exchange rate developments every
day”, then it is perhaps not surprising that the behaviour of the exchange rate may be regarded
by some policy-makers as “a key indicator of central bank performance” (Hutabarat 2006).
That said, there are a number of other possible reasons for why (particularly emerging market)
monetary authorities will prefer some degree of control over the exchange rate, even when
operating an inflation targeting regime. First, high exchange rate pass-through into domestic
prices may mean that exchange rate movements have important implications for monetary
policy. In this respect, inflation targeting is not necessarily inconsistent with exchange rate
intervention, if this intervention is only carried out in cases where it is compatible with the
central bank‟s inflation forecast. Second, the risks attached to currency depreciation will
increase in cases where there is a large amount of external debt held by various sectors of the
economy, leading to the potential for currency mismatches on domestic balance sheets. Of
particular concern is the potential for fiscal dominance, whereby a large stock of foreign
currency denominated government debt may lead to political pressure on the central bank to
limit any declines in the domestic currency. Third, relatively underdeveloped financial
markets may mean that dollar pegs are used to provide informal hedging for importers and
exporters, in response to trade being largely invoiced in US dollars and „original sin‟ problems
of capital market incompleteness that make hedging difficult (McKinnon and Schnabl 2003).
Finally, sharp reversals of capital flows may cause fear-driven momentum in exchange rate
movements, leading to possible overshooting and potentially damaging volatility.
3
A possible conclusion is that a managed float designed to accumulate reserves and resist real
appreciation could be preferable in some conditions over the polar extreme of free floating, for
countries that are not prepared to accept the constraints of a hard peg (Hernandez and Montiel
2001). That said, there are some circumstances where attempting to actively limit movements
in the exchange rate may not be the optimal policy response. Allowing nominal exchange rate
appreciation as a response to sustained capital inflows, for instance, helps to contain the local-
currency prices of imported goods and insulate the domestic supply of money and credit,
mitigating inflationary pressures in the domestic economy (IMF 1997). As the „impossible
trinity‟ dictates that a central bank cannot indefinitely control both the nominal exchange rate
and the money market rate in an economy open to capital flows, a floating exchange rate
should also allow the authorities to retain more control over domestic monetary policy. More
generally, allowing the nominal exchange rate to fluctuate in a wider range introduces
uncertainty and increases perceived exchange risk which should help to deter purely
speculative capital flows.
This paper touches on some of these policy issues in its investigation of the causes and effects
of exchange rate movements in Indonesia. We first consider fundamental exchange rate
determinants and measures of exchange rate flexibility, in conjunction with a more detailed
analysis of the August 2005 depreciation and policy response, so as to obtain a broad picture
of the factors that have driven exchange rate movements since the 1997-98 crisis. We then
consider the effects of exchange rate movements on prices, the external sector, and balance
sheets in Indonesia, and conclude by elaborating on some of the policy challenges presented
by exchange rate fluctuations in the post-crisis period, in the context of Indonesia‟s inflation-
targeting regime.
4
In estimating our models we confine ourselves to the post-crisis period, using monthly data
from January 2000 to December 2006, so as to exclude the sharp exchange rate movements
associated with the Asian crisis. We use an error correction framework incorporating interest
rate differentials, the terms of trade, changes in net foreign assets, and measures of country
risk and global risk as potential exchange rate determinants. Results can be found in
Appendix A.
In the long run model with variables in levels, we find a negative relationship between the
level of country risk and both the real and nominal exchange rate. Country risk, which is taken
from the International Country Risk Guide and based on a composite of political, financial,
and economic risk factors, is commonly used by investors as a proxy for default or regulatory
risk. Our finding confirms that political turmoil and policy uncertainty have had marked
effects on investor confidence and market sentiment since the turn of the decade. For instance,
the weakening of the rupiah in April-July 2000 was mainly due to political uncertainty and
social unrest, which were quickly reflected by downgrades to Indonesia‟s sovereign long-term
and short-term debt. Uncertainty stemming from accusations of presidential graft also put
downward pressure on the rupiah during January-April 2001, as did the further debt
downgrades and forecasts of lower growth which again reduced investor and corporate
confidence later in the year. The Bali bomb blast of October 2002 was another example of a
political event which triggered panic buying of US dollars.
Long run results for the (non-oil) terms of trade also conform to our priors and contrast with
those of Sahminan (2005). Under a floating exchange rate regime, movements in a country‟s
terms of trade should be offset by corresponding exchange rate movements, cushioning the
effect on income and output. For instance, a decline in the price of exports relative to imports
will – assuming inelastic quantity responses – lead to a decrease in the supply of foreign
currency relative to demand, resulting in an exchange rate depreciation that should in turn
improve the competitiveness of export and import-competing industries. We find that the level
of the terms of trade positively affects both the real and the nominal exchange rate. On the
other hand, contrary to expectations changes in the terms of trade are found to have a
significant and negative impact on changes in both the nominal and real exchange rate.
Among the determinants tested in the short run models incorporating the error correction term,
only changes in net foreign assets have a significant effect on the exchange rate with the
expected sign, indicating that capital inflows – reflected by a decrease in net foreign assets –
lead to a rupiah appreciation (and vice versa). Indeed, while persistent current account
surpluses may have placed sustained upward pressure on the exchange rate since the crisis, the
model results suggest that fluctuations in capital flows appear to be a significant driver of
shorter-term exchange rate movements. However, endogeneity may well complicate this
relationship: part of the reason that shorter term portfolio flows are viewed as creating
instability is because any initial depreciation caused by a withdrawal of capital may itself
prompt further withdrawals.
5
Interest rate differentials may also affect capital flows into and out of a country, with
corresponding implications for the domestic currency. This was the case in Indonesia during
2004, for instance, when declining interest rate differentials due to monetary policy tightening
in the US triggered short-term capital outflows, while higher oil prices and concerns about the
presidential election in July added to downward pressure on the rupiah. However, we find
little evidence of a robust relationship between interest rate differentials and the exchange rate.
There are a number of reasons why this may be the case. First, higher interest rate differentials
will not necessarily attract foreign investment if they reflect correspondingly large differences
in the levels of risk associated with such investment. Second, while higher interest rates may
attract debt inflows, they may actually deter equity inflows through their dampening effect on
the broader economy. Third, endogeneity may again be an important concern: it is possible
that causality may run from exchange rates to interest rates as well as from interest rates to
exchange rates. For instance, in April and May 2003 Bank Indonesia explicitly signalled that
the strengthening currency gave it more leeway to spur the economy by lowering interest
rates, which it said had room to fall. Such two-way causality will be present if the authorities‟
decision to alter domestic interest rates is in any way affected by movements in the currency,
an issue that will be returned to later.
To clarify these causality relations we run Granger causality tests on the exchange rate and the
explanatory variables in our model. We find that the nominal rupiah exchange rate Granger
causes the interest rate differential and country risk, while the real effective exchange rate
Granger causes the interest rate differential, country risk, and net foreign assets. Hence there is
evidence to suggest that interest rates, capital flows, and perceptions of country risk may all
themselves be affected by exchange rate movements, despite being treated as exogenous in the
model. Moreover, rolling regressions show that many of the coefficient estimates are highly
unstable, indicating that different factors matter at different times, and further reflecting the
difficulties inherent in estimating a model of this nature. Indeed, three-year rolling regressions
suggest that only the coefficient of country risk is stable over time, with most other
coefficients changing in both magnitude and sign over the period.
We have already noted that if interest rates and exchange rates are co-determined, this will
create problems for our particular modelling approach. More generally, if the exchange rate is
in any sense treated as a target of policy in Indonesia, then models that include traditional
market-oriented determinants are much less likely to be successful. Indeed, policy responses
to exchange rate movements – such as intervention in foreign exchange markets – may in fact
prevent the exchange rate moving towards a market-determined equilibrium. Our next step,
therefore, is to examine whether and how much the exchange rate has been managed in the
post-crisis period.
Here we attempt to determine the extent to which exchange rate management can explain
these stylised facts. Given the difficulty of obtaining actual data on exchange rate intervention,
one approach is to use changes in central bank reserves as a proxy and, on this basis, to
directly determine the extent to which the central bank appears to be intervening to influence
the exchange rate. Another approach, closely following Baig (2001), is to compare the post-
crisis behaviour of the rupiah and other nominal variables with their pre-crisis behaviour. In
doing so, it may be possible to determine whether adjustments to achieve external and internal
balance are occurring more through the exchange rate (instead of through other nominal
variables), or whether there is instead any evidence to suggest that Indonesia has reverted to a
de facto peg against the US dollar.
Figure 1
Asian Currencies Against US$
December 1996 = 100
Index Index
110 110
100 100
South Korea
90 90
80 80
Malaysia
70 70
60 60
Thailand
50 50
Philippines
40 40
30 Indonesia 30
20 20
10 10
1997 1999 2001 2003 2005 2007
7
Figure 2
Asian Nominal Effective Exchange Rates
2000 = 100
Index Index
120 120
South Korea
Malaysia
110 110
100 100
Thailand
90 90
80 80
Philippines
70 70
Indonesia
60 60
50 50
1998 2000 2002 2004 2006 2008
55 5000
50 6000
45 7000
Rp per USD (RHS)
40 8000
35 9000
30 10000
Reserves (LHS)
25 11000
20 12000
2000 2001 2002 2003 2004 2005 2006 2007 2008
8
Figure 4
Foreign Exchange Reserves and the
Exchange Rate
12-month rolling correlations
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
-0.2 -0.2
-0.4 -0.4
2000 2001 2002 2003 2004 2005 2006 2007 2008
Of course, there are a number of caveats associated with treating reserves as an indicator of
intervention. Perhaps most importantly, the motivation underlying reserve accumulation may
be purely precautionary. Building a large stockpile of reserves assets may be viewed as a way
to reduce vulnerability to external shocks and forestall a sell-off of the domestic currency.
Taking advantage of a favourable price by accumulating reserves when the rupiah exchange
rate is high should not then be viewed as a signal of intervention. Also, fluctuations in reserves
can reflect valuation adjustments, debt repayments, and other factors that do not necessarily
represent market intervention. Moreover, forward market intervention is not fully captured by
the gross reserves figures (Baig 2001).
Despite these caveats, there is anecdotal evidence to suggest that an interventionist motive
explains at least part of the movements in reserve assets. This intervention is often publicly
justified by the need to constrain „volatility‟ in the rupiah. In practice, this often appears to
imply intervening to prevent sharp exchange rate depreciations from continuing further. For
instance, in July and August 2007 the rupiah depreciated by over 4 per cent due mainly to the
sub-prime mortgage turmoil in the US, the impact of which spread to emerging markets as
global investors reassessed their holdings of risky assets. Along with the Federal Reserve
decision to cut interest rates, which triggered the return of capital inflows into rupiah assets,
Bank Indonesia also intervened in the foreign exchange market to reduce downward pressure
on the rupiah. At the end of August, the Governor of Bank Indonesia reported that a rupiah
worth 9000 per US dollar was “best for our economy” and that BI would “remain in the
market to keep volatility in check”.1
Instances of shorter-term intervention to prevent sharp appreciations in the exchange rate can
also be pinpointed. For example, in the months prior to the depreciation discussed above, the
9
rupiah appreciated quickly, supported by higher yields on rupiah instruments relative to other
instruments in the region. Together with the rising appetite of global investors for emerging
market assets, these yield differentials helped fuel substantial foreign capital inflows in the
first half of 2007. In this case, to prevent further exchange rate appreciation, Bank Indonesia
stepped into the market and bought foreign exchange.
Moreover, although BI has shown an inclination in recent years to intervene to prevent sharp
exchange rate movements in both directions, there are some signs that BI has been
deliberately buying foreign exchange in the market over the longer term to maintain a low,
stable rupiah for the benefit of exporters. Notably, although Indonesia‟s reserve accumulation
has been relatively subdued until recently, in the past two years it has picked up dramatically,
with reserve holdings doubling in size since the low reached in August 2005 (Figure 5).
Notably, this two-year period corresponds closely to the period of relative flatness in the
rupiah (refer to Figure 1), which has continued despite significant capital inflows, trade
surpluses, and a broadly depreciating US dollar.
2 The reserve accumulation described above (and, in the absence of full sterilisation, the implied growth in base money) may have itself
partly contributed to these inflationary pressures.
10
Figure 5
100 200
Malaysia (LHS)
75 150
50 100
Thailand (LHS)
25 Indonesia (LHS) 50
Philippines (LHS)
0 0
2000 2001 2002 2003 2004 2005 2006 2007 2008
Figure 6
Asian Real Effective Exchange Rates
2000 = 100
Index Index
Thailand
160 160
Philippines
140 140
Indonesia
120 120
100 100
Korea Malaysia
80 80
60 60
40 40
1998 2000 2002 2004 2006 2008
11
Figure 7
Asian Currency Volatility*
% %
PHP per
US$
1.5 1.5
SKW per US$ THB per US$
MYR per US$
1.0 1.0
0.5 0.5
0.0 0.0
1999 2001 2003 2005 2007
* 22-day average of absolute daily percentage change against the USD
However, a more volatile exchange rate does not by itself mean that an exchange rate regime
can be characterised as freely-floating. The volatility of interest rates and reserves might also
be expected to show a corresponding decline in economies that shift away from targeting the
exchange rate through monetary policy and intervention and towards a floating regime.
Reserves volatility should fall as the monetary authorities are no longer bound to intervene via
changes in reserve holdings to maintain the exchange rate at a predetermined level. Instead,
balance of payments adjustments to capital flows (for instance) can occur through changes to
the exchange rate. In regimes where the exchange rate is the nominal anchor, on the other
hand, changes to interest rates may also be needed to help ensure that capital flows are
consistent with balance of payments equilibrium at the desired exchange rate. Indeed, with the
exception of Malaysia, the interest rates of the East Asian countries were substantially more
volatile than those of floating countries during the pre-crisis period of fixed exchange rates
(Baig 2001).
Figure 8 suggests that the volatility of both reserve holdings and interest rates in Indonesia has
generally fallen in the post float period, suggesting that exchange rate movements are reducing
some of the adjustment pressures formerly borne by other nominal variables. That said,
reserves volatility rose sharply in the second half of 2005, and is now at around its pre-crisis
levels.
12
Figure 8
Volatility of nominal variables
12-month average absolute percentage changes
% %
24 6
Interest rate
20 5
Exchange rate
16 4
Reserves
12 3
8 2
Inflation
4 1
0 0
1994 1996 1998 2000 2002 2004 2006 2008
The 12-month period beginning with the market turmoil and rupiah depreciation of August
2005 provides a particularly interesting case study of exchange rate movements in Indonesia.
In August 2005 the rupiah depreciated by over 10 per cent against the US dollar (Figure 9).
13
Figure 9
Rupiah per US dollar
Daily
8000
Rp 8000
Rp
8500 8500
9000 9000
9500 9500
10000 10000
10500 10500
11000 11000
Jan-04 Jan-05 Jan-06 Jan-07 Jan-08
Figure 10
Balance of Payments
CEIC data, quarterly
US$ US$
bn Changes in bn
reserves
Net capital
4 4
flows
2 2
0 0
-2 -2
Current account
balance
-4 Net errors and -4
omissions
-6 -6
Mar-04 Dec-04 Sep-05 Jun-06 Mar-07 Dec-07
The primary causes of the depreciation were a weaker balance of payments, and negative
sentiment towards fiscal sustainability due to a sharp increase in oil prices. These rising oil
prices combined with increasing domestic demand to trigger a reversal in the current account
from surplus to deficit (Figure 10). In particular, the oil price increase meant the fuel import
bill of the state-run oil company Pertamina rose sharply, but the increase was not followed by
a correspondingly higher volume of oil exports due to the prevailing constraints in oil
production.
Negative sentiment towards fiscal sustainability due to the subsequent increase in the cost of
fuel subsidies put further downward pressure on the exchange rate. Increases in the US federal
funds rate (from 2.25 per cent in January 2005 to 4.25 per cent in December) also created an
environment of uncertainty in world financial markets, raising the risk premium demanded on
14
investments in emerging markets. The short-term capital inflows (particularly bank and money
market flows) that had provided a source of foreign exchange in the second half of 2004
sharply reversed, while foreign direct investment inflows were only marginal (Figure 11).
Indeed, there is some evidence that perceptions of increased risk quickly became self-fulfilling
in the third quarter of 2005, with mounting depreciation expectations stimulating further
purchases of foreign exchange.
Figure 11
Net Capital Flows
CEIC data, quarterly
US$ US$
bn Net portfolio bn
Net FDI inflows debt inflows
4 4
2 2
0 0
-2 -2
-4 Net other -4
Net portfolio investment inflows
equity inflows
-6 -6
-8 -8
Mar-04 Dec-04 Sep-05 Jun-06 Mar-07 Dec-07
Figure 12
Indonesian Nominal Variables
Monthly
Rp %
Rp per USD
11000 20
3-month interest
rate (RHS)
10000 15
9000 10
8000 5
Inflation y-o-y
(RHS)
7000 0
6000 -5
2000 2001 2002 2003 2004 2005 2006 2007 2008
In response, fiscal measures were taken to reduce fuel subsidies, while Bank Indonesia‟s
response to the ensuing surge in inflation and the external pressure on the rupiah was to
increase the policy rate from 8.5 per cent in July to 12.75 per cent by the end of the year
(Figure 12). Together with an increase in reserve requirements designed to absorb excess
liquidity from the banking system, this tightening in monetary policy helped to restore
investor confidence.
15
At the same time, various policies aimed at exchange rate stabilisation were also instituted.
Through the 27 April 2005 policy package, Bank Indonesia increased the scale of foreign
exchange intervention, raised the maximum interest rate under the guarantee-scheme on
foreign exchange deposits, and required commercial banks to maintain a net open position in
foreign currency on an intraday basis. Following the August depreciation, BI adopted a further
series of policy measures to curb speculation in the foreign exchange market. The 30 August
2005 policy package launched foreign currency swaps with 1-7 days maturity as an instrument
of open market operations, provided 3-6 month swap facilities with the option of extension for
hedging by investors, and prohibited margin trading of the rupiah against foreign currency. BI
also began to more intensively monitor non-underlying foreign exchange transactions by
banks, and limited rupiah transactions and the provision of foreign exchange credits by banks
to non-residents (Bank Indonesia 2006).3
In line with the growth in the balance of payments surplus, official reserves expanded to reach
$42.6 billion at the end of 2006. It is difficult to deduce how much of BI‟s reserve
accumulation over the period was intended to enhance reserve coverage (the precautionary
motive for reserve holdings), and how much was intended as foreign exchange intervention.
At September 2005 reserves had hit their lowest level since late 2002, suggesting a
precautionary justification for increasing reserve holdings. However, by mid-2006 there were
some concerns that rupiah appreciation had adversely affected exports, especially in the
manufacturing sector. In particular, the real effective exchange rate reached its highest level in
five years in May. Hence both motivations may have played a role. The close correlation
between international reserve movements and movements in the exchange rate over the period
3 A number of broader measures were also adopted. Bilateral swap arrangements and Asian swap arrangements with the ASEAN countries
plus Japan, China and Korea were signed to bolster access to foreign currency reserves. Bank Indonesia and the Government also agreed to
establish a mechanism for dollar demand management in Pertamina. In addition, the 5 July 2005 policy package required State-Owned
Enterprises (SOEs) to repatriate their export revenues. See Bank Indonesia (2006) for more details.
16
suggests that the changes in central bank reserves can at least partly be attributed to
intervention (Figure 13).
Figure 13
Foreign Exchange Reserves and the
Exchange Rate
Monthly
65
US 7500
Rp
60 8000
Rp per USD
55 (RHS) 8500
50 9000
45 9500
35 10500
30 11000
25 11500
2004 2005 2006 2007 2008
As a whole, the 2005-06 period draws attention to a number of key themes. For instance, it is
clear that, at least in this period, market sentiment and yield differentials were important
drivers of „hot‟ capital flows in Indonesia. More generally, while the sharp exchange rate
decline in August 2005 can clearly be traced to balance of payments pressures, there is some
evidence to suggest that this depreciation was also at least partly responsible for the shift to
current account surplus and the subsequent recovery that took place in 2006. At least in this
case, it appears that externally-driven exchange rate movements had an equilibrating impact
on the external balance. Finally, this episode also reveals the authorities‟ willingness to
employ a combination of policy measures to combat the effects of exchange rate depreciation,
including monetary tightening, foreign exchange intervention, capital controls, and other
measures aimed at exchange rate stabilisation. However, it remains unclear whether these
measures were designed solely to mitigate the inflationary threat associated with currency
weakness, or whether the exchange rate was also targeted during this period for reasons
independent of its pass-through effect on inflation.
17
3.1 The effect of the exchange rate on inflation
A major issue for monetary policy in emerging markets is that exchange rate pass-through to
domestic prices tends to be high. Changes in the exchange rate may affect domestic price
inflation through their impact on the prices of traded final goods, the prices of imported
intermediate goods, and inflation expectations (Ho and McCauley 2003). A priori, we might
expect the strength of these effects to be particularly strong in the case of emerging market
economies such as Indonesia, which have a high-inflation history and have only recently
shifted to an inflation targeting regime. If exchange rate depreciations do tend to cause
significant increases in consumer prices, policy-makers may attempt to more actively manage
the exchange rate, even within the context of an inflation-targeting regime. Indeed, targeting
the exchange rate as a (particularly important) inflation determinant is not necessarily
incompatible with treating low and stable inflation as the overriding objective of monetary
policy.
There are a number of studies which have provided estimates of exchange rate pass-through to
domestic prices in Indonesia, most of which find that pass-through in Indonesia is higher than
pass-through in other Southeast Asian countries. Using quarterly data from 1985 to 2000,
Sahminan (2005) finds complete exchange rate pass-through to import prices in Indonesia.
Yanuarti and Hutabarat (2006) find that a 1 per cent exchange rate depreciation increases
domestic prices in Indonesia by 0.25 per cent, which is larger than the inflationary effect of
exchange rate movements in Malaysia, the Philippines, and Thailand. Bank Indonesia‟s
SSMX model also generates estimates of exchange rate pass-through to domestic prices. The
SSMX model estimates that a 1 per cent exchange rate depreciation leads to a 0.48 per cent
increase in import prices and a 0.16 per cent increase in consumer prices. More recently, Ito
and Sato (2006) show that exchange rate pass-through into CPI in Indonesia is larger than in
Korea, Thailand, Singapore, and Malaysia, and that exchange rate pass-through in Indonesia
has not declined significantly since the crisis (in the period to August 2005).
Kurniati (2007) uses an error correction framework to model exchange rate pass-through in
two stages: the first round effect of exchange rate changes on import prices, and the second
round effect of import price changes on consumer price inflation. She finds that short-run pass
through to import prices has declined significantly in the post-crisis period compared to the
pre-crisis period (from 0.38 to 0.21), while long-run pass through is relatively unchanged. She
also finds that both short-run and long-run exchange rate pass-through to consumer prices
have fallen in the post-crisis period. These declines in pass-through are attributed to a decline
in the import to GDP ratio, a more stable history of inflation outcomes and a decline in
inflation expectations.
If it is true that the responsiveness of consumer price inflation to exchange rate changes has
fallen, there may be less justification under an inflation targeting regime for an aggressive
policy response to any given exchange rate shock. However, Kurniati also suggests that
domestic prices appear to respond proportionally more to large exchange rate changes (such as
18
those experienced during the crisis) than small changes, suggesting a potential asymmetry in
exchange rate pass-through. As a result, she does not necessarily rule out the need for a
broader policy response if the exchange rate experiences a “high and permanent” change.
That said, compared with other emerging market economies in Asia, the Indonesian economy
remains one of the least integrated into world trade. This in turn makes Indonesia relatively
more insulated from the external sector consequences of exchange rate fluctuations. Actual
trade openness (as measured by the ratio of exports plus imports to GDP) is around 50 per
cent, compared to an average of 130 per cent for ASEAN countries, and unlike its neighbours
Indonesia‟s trade integration with the global economy has not increased since the mid-1990s
(IMF Selected Issues 2007).
Nevertheless, the traded sector still forms a substantial proportion of the Indonesian economy,
and hence it is important to assess how and to what extent this sector is affected by
movements in the exchange rate. Previous research is somewhat inconclusive on this issue.4
Astiyah and Santoso (2006), for example, find that a real exchange rate depreciation does not
improve the trade balance in the short-run, while in the long-run, the positive impacts of a real
exchange rate depreciation on export volumes are offset by an increase in imports. Husman
(2005), on the other hand, provides contrary evidence that the Marshall-Lerner condition holds
for Indonesia‟s trade balance with US, Japan, Korea, Taiwan, and Germany in the period from
1993-2004. Nonetheless, the estimated coefficient on the real exchange rate is relatively small,
indicating that non-exchange rate factors (particularly on the supply side) may have a greater
role in determining the trade balance.
Here we confine our attention to simple aggregated models of export and import volumes.
Although this approach involves a number of limitations, our aim is simply to obtain a broad
4 The part on previous research that follows is directly cut from Warjiyo (2005)
19
picture of the impact of exchange rate movements on the real economy.5 We use a standard
aggregated trade model which expresses import volumes as a function of domestic GDP and
the real effective exchange rate (REER), and export volumes as a function of foreign GDP and
the REER. Quarterly data from 1995 to March 2008 were obtained on imports and exports
(using constant price series from the national accounts), domestic and foreign GDP (the latter
being a weighted average of Indonesia‟s eight major trading partners), and the real effective
exchange rate (Figure 14).
Figure 14
REER and Trade Volumes
2003 = 100
Index Index
180 180
Import volumes
160 160
140 140
120 120
100 100
REER
80 80
Export volumes
60 60
40 40
1995 1997 1999 2001 2003 2005 2007
We use a standard two-step error correction procedure to derive the relevant elasticities. In the
first step we regress levels of real imports (exports) on the REER and levels of domestic
(foreign) GDP, with all variables in logs. In the second step we estimate dynamic equations in
first differences including the lagged residuals from the first-stage regressions – the error
correction terms – as explanatory variables. Lags of REER differences and GDP differences
are also included in the second step regressions, with Akaike‟s information criterion used to
select the lag structure.
Results of these regressions are set out in Appendix B (Tables 1 and 2). As expected, the long-
run elasticity of real exports is negative with respect to the REER. Contrary to standard
theory, however, the equivalent coefficient in the equation for real imports is also negative
(and larger than the exports coefficient) indicating that a decline in the REER is also
associated with an increase in imports. Similarly, the short-run REER elasticities are generally
negative for both exports (as expected) and imports (contrary to expectations).
The apparently negative relationship between import volumes and real exchange rate
movements over the post crisis period contradicts the expected price effect of exchange rate
changes. One possible explanation is Indonesia‟s dependence on imported inputs in the
production of exports. If imported inputs comprise a high proportion of the total inputs
5 For instance, aggregation biases may mean that the trade models used here underestimate the response of trade volumes to relative prices,
20
required to produce exports, then to the extent that a depreciation of the currency increases
export demand (through the standard price effect), it will also increase the demand for
imports. We might also expect a production lag before exporters are able to use imported
inputs to increase export production. Hence, if this explanation is correct, imports should
increase before and in anticipation of an increase in exports. To test this, we run pairwise
Granger causality tests on the first differences of export and import volumes. Results are
reported in Appendix B (Table 3), and suggest that changes in import volumes do indeed
Granger cause changes in export volumes over the full sample period and in the post-crisis
sample. On the other hand, there is no evidence to suggest that changes in export volumes
Granger cause changes in import volumes.
The degree of exchange rate pass-through to rupiah-denominated import and export prices
may also affect the above relationships. Specifically, if export prices are more closely linked
to the exchange rate than import prices, then the positive effect of the exchange rate
depreciation on import volumes will be amplified. This is because, for any given depreciation,
the margin between the local-currency price of exports and the cost of imported inputs will be
relatively larger, while the price effect on imports for consumption will be relatively smaller.
We test this by running regressions of the first difference of import and export wholesale price
indices on lagged differences of themselves and contemporaneous and lagged differences of
the exchange rate (note that lagged differences of the exchange rate are not found to be
significant in any instance). Our results, reported in Appendix B (Table 4) suggest that rupiah-
denominated export prices are almost twice as responsive to exchange rate movements as
rupiah-denominated import prices are, and this result is robust across different specifications.
Overall, these findings suggest that exchange rate movements may have a relatively low real
impact on Indonesia‟s external sector, which in any case accounts for a relatively small
proportion of the Indonesian economy. As further support for this result, we also find no clear
relationship between the real exchange rate and the US dollar denominated trade balance over
the post-crisis period. Our explanation pertaining to Indonesia‟s dependence on imported
inputs is in line with previous empirical studies that have cited the imported input effect
(amongst other factors) as an explanation of why exchange rate depreciations can in fact be
contractionary in terms of their overall effect on the Indonesian economy (e.g Hutabarat
2006). At very least, offsetting cost and revenue exposures to exchange rate movements in
Indonesia‟s external sector appear to reduce the impact of these movements on trade and
output. This may also imply that any exchange rate response to an external shock needs to be
relatively large in terms of size and/or duration if it is to stimulate the traded sector adjustment
required to establish external balance.
because they do not account for large differences in response across sectors or individual goods.
21
Countries that finance themselves by issuing foreign currency-denominated debt are selling
investors protection against the risk that the domestic currency will depreciate, and, in the
process, increasing their own vulnerability to exchange rate movements. If a substantial
portion of debt is denominated in foreign currencies, as is often the case in emerging
economies, the resulting currency mismatch between foreign-currency debts and revenues
means that an exchange rate depreciation can lead to an increase in the domestic currency cost
of repayments without a commensurate increase in ability to pay.6
Prior to the 1997 crisis the private sector had a number of incentives to borrow from abroad,
resulting in substantial currency mismatches on corporate balance sheets. Tight monetary
policy, a lack of domestic liquidity, and the absence of a domestic bond market all motivated
firms to source their funding offshore. Importantly, currency risk appeared low as movements
in the exchange rate were constrained by a policy of tightly managed depreciation (of around
5 per cent per year) against the US dollar. While corporate revenues were primarily domestic,
the exchange rate peg meant that the ensuing currency mismatches were not generally viewed
as cause for concern, and short-term external debts were rolled over relatively easily.7
However, as the rupiah depreciated during the 1997 crisis, firms were suddenly confronted
with a large increase in the rupiah value of their external debts. As most of their liabilities
were short-term in nature, they had little time to restructure their debts, adding a maturity
mismatch problem to that of the currency mismatch. These problems quickly flowed through
to the banks, which had matched the currency composition of their assets and liabilities by
extending credits in foreign currency, but whose balance sheets nevertheless deteriorated. This
was because the credit quality of banks‟ foreign currency assets (which were mainly loans to
the corporate sector) became severely impaired, even though the valuation effects of the
depreciation may have been reflected on both sides of their balance sheets. The government‟s
finances were similarly adversely affected. Taxation revenues from the private sector
declined, while expenditures on safety net programmes and subsidies aimed at alleviating the
burden of the crisis on the poor increased. As a result, the government‟s ability to service its
external debt also weakened significantly.
Prima facie, it appears that both public and private sector balance sheets have become less
vulnerable to exchange rate depreciations in the post-crisis period. The level of external debt
outstanding as a proportion of GDP is one indicator that suggests that declines in the exchange
rate may now affect the debt servicing burden less adversely than previously (Figure 15).
Unfortunately, as is the case for most emerging market economies, the lack of available data
makes it difficult to conduct a fully-fledged balance sheet analysis in the Indonesian case. In
particular, data on private sector balance sheets (particularly the currency composition of
assets) are generally unavailable or inaccurate.
6. Also, if an increase in the domestic interest rate is used to stabilise the exchange rate, domestic currency borrowers will instead be
adversely affected.
7. This discussion of the crisis draws heavily on Kusumaningtuti (2004).
22
Figure 15
External debt outstanding
Per cent of GDP
% %
140 140
120 120
100 100
80 80
60 60
40 40
20 20
0 0
1998 2000 2002 2004 2006
As a result, the authorities may have strong incentives to prevent the exchange rate from
falling sharply. However, the data show that public sector external debt has remained
relatively flat in absolute terms since the crisis, suggesting no increase in vulnerability to
exchange rate movements (Figure 16). The government has also gradually increased its use of
domestic debt finance, while attempting to facilitate the development of a domestic bond
market, both of which should assist in mitigating balance sheet risks (Bank Indonesia, 2006).
Figure 16
External debt outstanding by sector
USD USD
bn Financials bn
160 Corporates 160
Government
140 140
120 120
100 100
80 80
60 60
40 40
20 20
0 0
1998 2000 2002 2004 2006
Figure 17
Commercial Banks' Foreign Currency
Denominated Credits and Deposits
Rp bn Rp bn
200 200
Deposits
150 150
100 100
Credits
50 50
0 0
Net
-50 -50
-100 -100
1995 1997 1999 2001 2003 2005
24
Figure 18
Share of Credits and Deposits Denominated in
Foreign Currency
Commercial banks, per cent of total
% %
40 40
Share of credits denominated
in foreign currency
35 35
30 30
25 25
20 20
Share of deposits denominated in
foreign currency
15 15
10 10
1995 1997 1999 2001 2003 2005
25
current and capital account. Allowing the rupiah to appreciate may therefore help to
reduce inflation by reducing the price of imported goods and reducing the pressure on
the money supply.
26
Appendix A
Table 1: Results of Unit Root Tests
NER REER
Interest Rate Differential -0.008* 0.009*
(0.004) (0.003)
Terms of Trade 1.041* 0.222
(0.369) (0.335)
Country Risk -1.096* -1.828*
(0.457) (0.411)
Global Risk -0.1337 -0.072
(0.125) (0.142)
Net Foreign Assets 0.025 0.243*
(0.106) (0.090)
27
Table 3: Estimation Results for the Variables in Difference
29
Table 3: Granger causality tests
Changes in export volumes (dxvol) and import volumes (dmvol)
Test sample
Mar 1995 to Mar 1998 to
Mar 2008 Mar 2008
Yes Yes
dmvol dxvol
(0.059) (0.082)
No No
dxvol dmvol
(0.772) (0.884)
Note: p-values in parentheses. X Y means “X Granger-causes Y”, or, more formally, the rejection
of the null hypothesis “X does not Granger-cause Y”, using a Granger causality test with 4 lags and
assuming a 0.10 p-value threshold.
30
References
Astiyah, S and MS Santoso (2005), „Nilai Tukar dan Trade Flows‟ (Exchange Rate and Trade
Flows), Bank Indonesia Working Paper 02/2005.
Bank Indonesia (2005), „Foreign exchange intervention and policy: Bank Indonesia
experiences‟, in BIS Papers No. 24 (May), „Foreign exchange market intervention in emerging
markets: motives, techniques and implications‟.
Calvo, G and C Reinhart (2002), „Fear of Floating‟, Quarterly Journal of Economics, 117, pp.
379-408.
Frankel, J and SJ Wei (1994), „Yen Bloc or Dollar Bloc? Exchange Rate Polices in the East
Asian Economies‟, in Ito T and A Krueger (eds.), Macroeconomic Linkage: Savings,
Exchange Rates, and Capital Flows, Chicago, University of Chicago Press.
Hernandez, L and P Montiel (2001), „Post-Crisis Exchange Rate Policy in Five Asian
Countries: Filling in the “Hollow Middle”?‟, International Monetary Fund Working Paper
01/170.
Ho, C and RN McCauley (2003), „Living with flexible exchange rates: issues and recent
experience in inflation targeting emerging market economies‟, BIS Working Papers No. 130.
Husman, JA (2005), „Pengaruh Nilai Tukar Riil terhadap Neraca Perdagangan Bilateral
Indonesia: Kondisi Marshall-Lerner dan Fenomena J-curve‟ (Impact of Exchange Rate on
Indonesia‟s Bilateral Trade Balance: Marshall-Lerner Condition and J-Curve Phenomenon),
Bank Indonesia Working Paper 04/2005.
IMF (2007), „Indonesia - Staff Report for the 2007 Article IV Consultation‟, IMF Country
Report No. 07/273, August 2007.
31
IMF (2007), „Indonesia – Selected Issues‟, IMF Country Report No. 07/273, August 2007.
IMF (1997), „Indonesia - The Implications for Monetary Policy of Dealing with Large Capital
Flows‟, Background Paper for 1997 Article IV Consultation Discussions.
Ito, T and K Sato (2006), „Exchange Rate Changes and Inflation in Post-Crisis Asian
Economies: VAR Analysis of the Exchange Rate Pass-Through‟, RIETI Discussion Paper
Series 06-E-018.
McKinnon, R (2001), „After the Crisis, the East Asian Dollar Standard Resurrected: An
Interpretation of High Frequency Exchange Rate Pegging‟, HKIMR Working Paper No.
04/2001.
McKinnon, R and G Schnabl (2004), „The East Asian Dollar Standard, Fear of Floating, and
Original Sin‟, Review of Development Economics, 8(3), pp. 331-360.
Mohanty, MS and P Turner (2005), „Intervention: what are the domestic consequences?‟, in
BIS Papers No. 24 (May), „Foreign exchange market intervention in emerging markets:
motives, techniques and implications‟.
Sahminan, S (2005), „Estimating Equilibrium Real Exchange Rates of the Rupiah‟, Bank
Indonesia Working Paper 08/2005.
Sahminan, S (2005), „Exchange Rate Pass-Through into Import Prices in Major Southeast
Asian Countries‟, Bank Indonesia Working Paper XX/2005.
Siregar, R and RS Rajan (2004), „Exchange Rate Policy and Reserve Management in
Indonesia in the Context of East Asian Monetary Regionalism‟, Centre for International
Economic Studies Discussion Paper No. 0403.
32