Asymmetric stabilizing impact of international reserves

ABSTRACT

This article studies international reserves’ nominal exchange rate stabilizing impact in emerging markets and developing countries, with a particular focus on its nonlinearity and asymmetry across different states of the economy. Using the fixed-effects and dynamic panel threshold models, we find the reserves to short-term debt threshold ratio after which the marginal stabilizing effect of reserves begins to fall during tranquil times. Such diminishing returns, however, do not appear to exist even at the excessive level of reserves during the global financial crisis, partly justifying precautionary demand for international reserves. These results call for extending reserve pooling or swap arrangements to enhance efficiency of reserve management by holding adequate, rather than excess, international reserves with an access to emergency lending during the crisis.

KEYWORDS:

• Diminishing returns
• exchange rate volatility
• global financial crisis
• international reserves

JEL CLASSIFICATION:

• F31
• F33

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Rodrik (2006) estimates the cost of excess reserves in developing countries to be around 1% of their GDP in 2004.

² See Bayoumi and Eichengreen (1998), Cady and Gonzalez-Garcia (2007), and Devereux and Lane (2003) for modelling the nominal exchange rate volatility. International reserves and their asymmetric stabilizing impacts were not examined in these studies.

³ See Table A1 in Appendix A for data sources. Following the fine classification of Ilzetzki, Reinhart, and Rogoff (2017), a country belongs to the category of peggers in a given year if it takes a de facto peg or pre announced horizontal band with margins of no larger than ±2%.

⁴ Appendix B presents the list of sample countries.

⁵ Similarly, Fratzscher (2009) reports that a majority of emerging market economies experienced a sharp currency depreciation against the US dollar in the period July 2008–January 2009 despite the fact that they were not the origin of the crisis.

⁶ We find no evidence of decreasing returns at the lower than 60th percentile during tranquil periods from both FE OLS and system GMM estimations.

⁷ Admittedly, countries are likely to have different views on the optimal hoarding of international reserves and the advantages of reserve pooling, as precautionary and mercantilist motives have different weights in their policy objective functions.