ABSTRACT
The reason for this difficulty is the problem of correlation with other unfortunate developments with which it coincides. If the central bank has made a large loss, one has to ask why. Perhaps it was because it tried (or was instructed to try) to defend the exchange rate, and saw its attempts fail. Losses on intervention could exceed gains (measured in devalued home currency) on remaining foreign exchange reserves, and, if they do, the central bank could make an overall loss in the year, and be driven into negative capital (assuming it had none, or very little, to start with). In such circumstances, when the domestic private sector and banks may be feeling acute strain from increased local currency interest payments to overseas lenders, raising tax rates at such a juncture could, in Shakespeare’s words, pile Pelion on Ossa. Unwelcome correlations could well arise, too, in the aftermath of a domestic banking crisis (which might accompany a foreign exchange crisis, but could also occur on its own).7 The central bank may have chosen, or been forced, to rescue one or more ailing banks, or offer guarantees on loans or deposits. Such actions could create a large hole in the central bank’s balance sheet. And at a most injudicious moment to try to balance a gravely damaged public sector budget, by increasing tax rates.8 Furthermore, this observation implies that central bank capital and commercial banks’ capital are really substitutes, not complements. The more capital the latter have, the less likely any will be brought to the brink of folding, and the less need the central bank has for a large capital base to guard itself against the risk and costs of rescuing them. Hence it is silly to suggest that the central bank should be subject to the same kind of Basel rules as retail banks. The correlation, cost and uncertainty arguments suggest that the level of the central bank’s capital may well not be an irrelevance in practice. They are reinforced by other observations. One is that debt enforcement is always challenging, even for central banks, and that arrangements for it vary greatly between different countries.9 A second is that a central bank in deep trouble, forced into exploiting revenue from seigniorage, will eventually find its monopoly challenged by steadier currencies.10 Third, the monetary policy regime in force can exert critical influence on whether budgetary problems translate into default risk, as Uribe (2006) demonstrates. Fourth, the full financial consequences of a central bank’s operations can only really be gauged in the long run. There could be big profits in a run of short periods, but the prices of the assets that it buys and sells can be volatile. So the flows may be lumpy and erratic, and losses are always a possibility in a later period, or indeed at any time. Hence the need for capital to be able to absorb them as and when they occur. But our fifth observation is perhaps the most telling: Parliament cannot bind its successors. Politicians have brief tenure. They are apt to change their minds. They cannot issue copper-bottomed, fully dependable commitments. So having some capital of its own insures the central bank against the mercuriality of government, and makes its own commitments more credible. It offers some measure of independence.11 This said, however, it must be admitted that the central bank’s capital would indeed be irrelevant if two conditions held: if governments could fully commit,12 and if markets functioned perfectly.
