Holding international reserves has benefits for countries, as they serve as pledgeable collateral for international borrowing. Deciding the appropriate level of reserves is, for a country, akin to the process a company goes through when deciding its level of cash. There are benefits in that creditors will demand a lower risk premium to more liquid companies. There are costs, most importantly the opportunity cost of investing the money in more risky and less liquid alternatives. For countries, there is an additional problem. The higher the level of reserves, the more exposed they are to capital losses that would happen if the exchange rate were to appreciate.
So, what is the appropriate level of reserves, and has the recent surge in reserve accumulation by developing countries been a good thing or a bad thing?
This IMF paper tackles the issue:
This paper examines the (quasi-)fiscal impact of the (opportunity) cost of international reserves. It proposes a conceptual framework, with particular emphasis on two hitherto somewhat neglected aspects: a more appropriate measure of gross opportunity cost, and potential savings from lower external debt spreads that countries "buy" by holding reserves. The framework is then applied to 100 countries over 1990-2004. The results suggest that a turning point has been reached in recent years: while most countries made money on their reserves during 1990-2001, most have been losing money during 2002-04.