A managed exchange rate regime breaks down and a period of free floating commences whenever the stock of foreign exchange reserves falls below some critical threshold level. A government faced by an exchange rate crisis will clearly be tempted to borrow in order to replenish its stock of foreign exchange and so defend its exchange rate. In Discussion Paper No. 95, Research Fellow Willem Buiter considers how such borrowing affects the survival of exchange rate regimes. Buiter extends recent literature on the collapse of exchange rate regimes by allowing explicitly for the government budget constraint and for the interest cost of servicing the public debt.

Under a managed exchange rate regime, Buiter notes, reserves are lost whenever domestic credit expansion (DCE) exceeds the growth of money demand. Domestic credit expansion is equal to the interest cost of servicing the public debt minus government borrowing, plus the 'primary government deficit' (spending on goods and services minus taxes net of transfers, excluding interest payments on the public debt).

A government that borrows to replenish its stock of reserves will obviously have available a larger stock of foreign exchange. If the interest rate earned on the additional reserves is less than the interest rate paid by the government on its newly incurred debt, however, the government's budgetary position will worsen as a result of the decision to borrow. If it does not reduce its primary deficit, the increased interest payments the government now faces as a result of the borrowing will force it either to increase the rate of domestic credit expansion or to engage in further borrowing. Sooner or later, the need to remain solvent will eliminate the option of further borrowing and domestic credit expansion will increase. By borrowing without a commensurate reduction in its primary deficit, a government therefore purchases a once-off increase in the level of its foreign exchange reserves at the expense of an ultimately higher rate of decline in the stock of these reserves.

Buiter analyses a very simple example of this borrowing in which there is a once-off open-market sale of government debt, after which the government does not resort to further borrowing. How does this affect the exchange rate regime? Buiter finds that, in the absence of uncertainty about future DCE or the growth of money demand, this borrowing will have an ambiguous effect on the timing of the exchange rate collapse. If the borrowing occurs just before the collapse would otherwise have occurred, the collapse is postponed. But if the borrowing takes place long before the collapse would have occurred in the absence of borrowing, the collapse is brought forward! Buiter also establishes that the borrowing always increases the magnitude of the speculative attack that brings about the final collapse of the managed exchange rate regime, as measured by the size of the final loss of reserves that brings the stock of reserves down to its critical threshold level.

Suppose we make the more realistic assumption that there is uncertainty concerning future DCE or the growth of money demand. Buiter finds that, in this case, borrowing lowers the probability of an early collapse, but increases the probability of collapse over a longer time horizon. Borrowing increases the length of the expected interval before collapse, provided that the interest rate on the public debt is not extremely high.

The policy message of the paper is familiar but important, Buiter contends. Borrowing additional reserves to defend the exchange rate has the effect of 'buying time'. This might be useful if a delay in implementation increases the quality of the package of fiscal cuts necessary to create a viable managed exchange rate regime. In the absence of a fiscal correction, borrowing lowers the likelihood of an early collapse and lengthens the expected interval preceding the collapse. Because borrowing raises the likelihood of a later collapse, however, viability (in the sense of the long-run assured survival of the managed exchange rate regime) can be achieved only by lowering the government's need for revenue from 'seigniorage', i.e. from recourse to monetization and an inflation tax. This requires a fundamental fiscal correction, in the form of a reduction in the primary deficit.


Borrowing to Defend the Exchange Rate and the Timing and Magnitude of Speculative Attacks
Willem H Buiter

Discussion Paper No. 95, February 1986 (IM)