Managing the Monetary Consequences of Reserve Accumulation in Emerging Asia

Abstract

The huge increase in international reserve holdings by Asian countries since the 1997 crisis has been one of the most important recent developments on the international financial scene. These buildups have contributed substantially to concerns about the creation of excessive global liquidity. How justified these concerns are depends heavily on the extent to which the reserve accumulating countries have been able to sterilize the effects on their domestic monetary aggregates. We use a unified theoretical framework to undertake dynamic estimations of the magnitude of sterilization and offset coefficients (which measure the degree of capital mobility) for a large set of Asian economies. We find that despite substantial capital mobility there has been a high degree of effective sterilization to date.

Keywords:

• Asia
• balance of payments function
• capital mobility
• monetary reaction function
• reserves
• sterilization

Acknowledgements

Excellent research assistance by Nicola Virigill is gratefully acknowledged. This paper was completed while the second author visited the Hong Kong Institute for Monetary Research (HKIMR). The author gratefully acknowledges the hospitality and support offered by the HKIMR Comments by Hans Genberg, other participants at an HKIMR seminar and an anonymous referee helped in improving the quality of the paper. The views expressed in this paper are those of the authors and do not necessarily reflect those of the HKIMR, its Council of Advisors, or the Board of Directors. Financial support from the Claremont Project on Asian Political Economy is gratefully acknowledged.

Notes

1. See Genberg et al. (2005) for more detailed descriptive data of reserve stockpiling in Asia.

2. There has been a growing body of literature exploring various aspects of the precautionary motive for reserve hoarding. See Aizenman et al. (2007), Garcia and Soto (2004), Jeanne and Ranciere (2006), Kim et al. (2005) and Li and Rajan (2006).

3. As noted by the Bank of Thailand (2005):

for a while after the 1997 crisis, accumulating foreign exchange reserve was another important consideration for exchange rate management. A strong stock of reserve helps to minimize external vulnerability and increase confidence in the economy, especially among foreign investors. (p. 280)

A similar point has been made regarding Korea:

Since adoption of the free floating exchange rate system, the Korean authorities have actively endeavoured to achieve stabilization of the foreign exchange market. In the process of overcoming the currency crisis, the Korean authorities tried to expand the nation's international reserves. (Rhee & Lee, 2005, p. 204)

The importance of intervention in order to accumulate reserves comes out clearly from the survey of selected central banks by Mihaljek (2005).

4. See, for example, Willett and Kim (2006). Of course, if an economy has been maintaining a fairly stable and rigid exchange rate peg like China and Hong Kong without there being opportunistic devaluations, it may be inappropriate to characterize that as being mercantilist. With regard to the supposed floaters in emerging Asia, there is evidence that they continue to actively manage their currencies vis-à-vis the US dollar even post-1998 (see Cavoli & Rajan, 2006b; Rajan, 2006; Willett et al., 2005; McKinnon & Schnabl, 2004). In addition, part of the change in reserves in US dollar terms arises from “revaluation gains” due to the depreciation of the US dollar against the major currencies in which reserves might be held, especially the Euro.

5. Our focus is on emerging Asian economies. Japan was not considered since it stopped large-scale intervention in 2004.

6. We have excluded Hong Kong from the empirical analysis as the theoretical framework to be used as motivation for the empirics (third section) is somewhat more appropriate for quasi-flexible/managed exchange rate regimes. In any event, Hong Kong does not appear to use monetary instruments to sterilize reserve buildup (Mohanty & Turner, 2005; Pang, 2005).

7. We exclude the period 1997:q2–1998:q2 as this is broadly the period of crises and acute monetary, exchange rate and financial instability in the region. The dating corresponds to Khalid and Kawai (2003) who identify July 1997 to 20 June 1998 as the currency crises period in Asia. (Indonesia is the possible exception to this.)

8. Needless to say that aggregates variations between countries. For instance, in ASEAN and the NIEs, the current account switch has largely been due to a collapse in investment rates. While investment rates in China have remained high, the current account surplus in China is primarily fuelled by rising savings rates (especially corporate savings) (see Kharas et al., 2006 and references cited within).

9. See Kharas et al. (2006) and IMF (2006) for recent discussions of capital flows in emerging Asia.

10. We discuss definitional details of NFA and NDA later in the paper.

11. For instance, see Celasun et al. (1999), Fry (1993), Kim (1995), Nyatepe and Coo (1995), Sarjito (1996), Rooskareni (1998) and Brissimis et al. (2002).

12. For instance in the case of Thailand the central bank has noted:

Since 2 July 1997, Thailand has adopted a managed-float exchange rate regime, replacing the basket-peg regime which had been in operation since 1984. The value of the baht has since then been largely determined by market forces. The Bank of Thailand manages the exchange rate by intervening in the foreign exchange market from time to time in order to prevent excessive volatilities in the markets, while fundamental trends are accommodated. In other words, movements in the exchange rates which are in line with the changes in economic fundamentals and financial development would only be smoothed and not resisted. (BOT, 2005, p. 276)

A broadly similar point is noted in the case of Korea:

Considering Korea's thin foreign exchange market as a small open economy, and the vulnerability of the won exchange rate to diverse external shocks and the changing global environment, the Korean authorities' efforts to stabilize the exchange rate could be regarded as inevitable. (Rhee & Lee, 2005, p. 205)

Also see the survey of selected central banks by Mihaljek (2005).

13. Even in the case of Singapore which is a basket pegger, Cavoli and Rajan (2007) show that the central bank loss function can be modeled in terms of output and inflation, the only difference being that they use the nominal effective exchange rate (NEER) rather than the interest rate as the policy instrument (also see Khor et al., 2004).

14. While one should ideally use NEER, we use bilateral US dollar rates as it substantially eases model solution. In any event, in so far as the bulk of trade in Asia is US dollar denominated, we probably do not lose much in terms of generality by using nominal bilateral rather than effective rates.

15. It is possible that a change in capital flows will cause a change in the money multiplier (for instance, see Rajan, 2007). More generally one might think about specifying a separate equation endogenizing the money multiplier (Jha & Rath, 2001).

16. More precisely, one would want to use a measure of broader fiscal stance, namely full employment primary fiscal balance. As Athukorala and Rajapatirana (2003) note:

(A) measure widely used to represent a fiscal policy stance in this type of analyses is the budgetary balance (measured as a ratio of GDP gross domestic product). But we believe that government expenditure is a superior indicator because in the context of an economic boom a country could well experience a “revenue surplus”, a reflection of a faster revenue growth compared to expenditure growth. Meaningful deficit comparison across countries should correct for such biases. Another problem with published data on budget deficits is that different definitions of taxation and borrowing can heavily skew the measured deficit.

17. In reality, the current account and capital account may not be completely independent. However, since the exact links are far from clear – for instance, high capital inflows could lead to trade liberalization to curb the inflationary effects or the domestic boom or does greater trade liberalization lead to intensified capital inflows – the assumption of independence may not be entirely inappropriate.

18. There are, of course, many other ways one could model the current account [see Edwards, (2002) for a review]. Given that our focus is on sterilization as opposed to the current account per se, a parsimonious modeling approach to the current account seems reasonable.

19. Given that we need daily data to compute a reliable within-the-quarter volatility measure, we use bilateral US dollar rates.

20. We also tried the M 1 money multiplier but the results did not change much.

21. Cyclical is defined as deviations from trend is based on the entire sample (i.e. pre- and post-crises). For a more careful analysis of output gaps in the East Asian economies in our sample (i.e. excluding India) as well as Hong Kong, see Gerlach and Yiu (2004) who discuss various ways of measuring the output gap in emerging Asian economies that are undergoing significant structural changes. They find that various output gap measures, including estimates derived from the HP-filter rule, are quite highly correlated.

22. The exchange rate is in logarithm term.

23. Additionally, higher inflation could in practice engender greater uncertainty, leading to reduced capital flows, though using forward looking inflation and cyclical income in the estimations leads to concerns about causality as discussed earlier.

24. We tried more than one period lags but they were not statistically significant.

25. This said, as highlighted in Table 3, most regional economies did not use fiscal consolidation as a means of curbing the inflationary consequences of capital inflows. The IMF (2005) makes a similar point when it notes of emerging Asia “(f)iscal policy was not used extensively to counter the aggregate demand impact of capital inflows” (p. 61).

26. Three caveats should be noted. (1) we use only foreign interest rates rather than interest rate differentials as the domestic interest rates are already captured in the ?NDA _t term (see Equation (10)). (2) The “c” term in Equation (9) is a measure of risk aversion. As discussed earlier, declining risk aversion is often cited as a reason for capital inflows to Asia. Since we lack good proxies for this variable, the term is kept constant and embedded within coefficients of the structural model. (3) The other oft-noted push factor, namely industrial country growth, is likely to be highly correlated with domestic country cyclical growth which is already included in the equation. Dasgupta and Ratha (2000) and Montiel and Reinhart (2000) discuss the determinants of capital flows into developing economies.

27. Of course there are exceptions. For instance, during an economic downturn there could be simultaneous use of expansionary monetary and fiscal policies, and vice versa during an upturn in economic activity.

28. Siklos (2000) pointed out a similar problem with the Hungarian–German interest rate differential and has argued that interest rates should not be difference stationary. The ADF results were confirmed by the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test.

29. Specifically, for Korea, Philippines, Taiwan, and Thailand we use the Money market rate. For Indonesia we use the Call money rate. For Malaysia we use the Interbank overnight money rate. For Singapore we use the 3-month Interbank rate.

30. Another factor includes interest earnings earned from foreign reserves accumulation. We ignore this issue.

31. Loosely speaking, this should correspond to the liquid part of foreign assets of the central bank. Ideally we should also be using only the liquid portion of central bank foreign liabilities. These data are not available.

32. In general, the monetary authority recognizes the end-year revaluation of foreign currency liabilities and assets in the Profit and Loss account of the income statement. Since the end-year income statement balance will be included in the equity (K) account of balance sheet, the change of NFA due to the revaluation effect can be offset by the change of equity so that the domestic monetary base will be unchanged. In other words, if NFA rises because of an increase in S _t, .

33. Of course, we could assume different scenarios such as Prasad and Wei (2005) do in the case of China. However, we have no basis of making such assumptions and given the number of other issues we deal with in this paper, it seems wise to refrain from making such ad hoc adjustments.

34. Korea, Singapore, and Taiwan do not report data for equity (K) account on International Financial Statistics (IFS), even though the balance sheet is balance. Therefore, for these three economies, Revaluation effect.

35. For instance, see Hu (2005) in the case of China.

36. See the analysis and references in Willett et al. (2005).

37. While some countries tightened capital controls during the crisis, average levels of controls are not noticeably after the crisis than before. See Li et al. (2006).

38. Detailed regression results are available on request.

39. The coefficients are initially estimated for the period between 1990:ql and 1995ql. Then whenever we add one more observation to estimate the following offset and sterilization coefficients we drop the earliest observation at the same time so as to ensure uniform sample size. For example, the second coefficient is estimated by using the sample period from 1990:q2 to 1995:q2. So each coefficient is estimated by using 5 year observations. To capture the effect of the currency crises, we add a crises dummy (equals one between 1997:q2 and 1998:q2) in the model.

40. See Kharas et al. (2006) for a discussion of capital flows in East Asia.

41. The World Bank (2005) and Mohanty and Turner (2005) discuss the latter two costs and Rodrik (2006) discusses the issue of opportunity costs. These costs need to be balanced against the likelihood that higher reserve holdings reduce a country's perceived international credit standing, hence lowering the country's risk premium.

42. Of course there is also the issue that increasing reserve holdings could encourage financial laxity in the reserve-issuing countries such as the US.