The management of foreign exchange reserves

BIS Economic Papers No 38
July 1993

Introduction

Questions about the management of foreign exchange reserves are likely to acquire increased prominence among the range of issues facing many central banks. Basic questions concerning the amount and form of reserves are particularly pressing for newly-established central banks, notably in the states of the former USSR. In most other formerly centrally-planned economies as well as many developing countries. central banks will need to consider the reserve implications of moving towards increasing currency convertibility and changes in exchange rate arrangements. More generally still, the marked increase in the scale of private international capital flows and the large amplitude of recent swings in the exchange rates of the three major international reserve currencies - the US dollar, the Deutsche Mark and the yen -underline the importance of careful assessment of reserve needs and composition. The purpose of this paper is to examine the main factors that should be taken into account in two basic aspects of foreign exchange reserves management: what amount and what form of reserves should be held. For the most part, the paper consists of a distillation of academic work -theoretical and empirical - on the demand for and currency composition of reserves. An important limitation of such analysis, as well as of this paper, is that the focus of attention is mainly on small economies, rather than on countries large enough to significantly influence foreign variables or whose own currencies serve as important international reserve assets or as a medium of exchange in international transactions. Nor does the paper address the special questions raised by particular exchange rate arrangements such as the European Exchange Rate Mechanism. The specific institutional arrangements that underpin this mechanism - notably the provisions for joint intervention and extensive mutual credit facilities - raise issues that are well beyond the scope of this paper.

The next section of this paper identifies three principal motives for holding reserves: a transactions demand, an intervention-related or precautionary demand and, third, a wealth-related or portfolio demand. Of these, the precautionary demand for reserves is generally regarded as by far the most important for most countries.

In Section Ill, the main factors influencing the desired or appropriate level of reserves are examined. These include the nature of the exchange rate regime, the external exposure of the economy, the flexibility of the economy in adjusting to or privately financing external payments imbalances, the extent of constraints on international trade and capital flows and, lastly, the opportunity cost of holding reserves. With respect to the relationship between reserve levels and the magnitude of balance-of-payments disturbances this paper argues that, logically, it is more appropriate to assess reserve needs or reserve adequacy on the basis of the ratio of reserves to current account variability than on the basis of the more commonly used ratio of reserves to imports. Evidence on this issue, together with the implications of following alternative "rules of thumb" in setting reserve levels, is examined for a group of sixty countries over the period 1979-91.

Section IV turns to the issue of the currency composition of reserves. Two traditional approaches are discussed: the mean-variance portfolio approach and the transactions-based approach. A simple empirical test does not support the proposition that either of these approaches dominated in practice, in either industrial or developing countries, over the period 1975-92.

Also outlined is an alternative intervention-oriented approach to reserve currency composition, which takes the timing of reserve use into account in choosing the optimal reserve portfolio. Data covering the period 1988-92 are used to illustrate the potential implications of this approach for a variety of industrial countries.