سودآوری بانک مرکزی از مداخله در بازار ارز

This paper assesses the profitability of foreign exchange market intervention by the Bank of England for the period 1973–1995. Profitability appears to an objective measure of the success of intervention policy, but we show that the tendency of the central bank to “lean against the wind” in its foreign exchange interventions means that profitability calculations will be heavily biased by the amount of cumulative intervention arbitrarily considered. It follows that profitability calculations will be strongly affected by the start–end dates used for calculation purposes. This cumulative intervention bias undermines the usefulness of using the profitability criterion as the sole measure for evaluating intervention policy. Nonetheless, the large sums of public money that can be staked on foreign exchange market interventions suggests the need for open reporting of the profits/losses involved.

In a classic 1953 essay “The Case for Flexible Exchange Rates,” Milton Friedman proposed an appealingly simple criterion by which the success or failure of interventions by the authorities in the foreign exchange market should be judged. Specifically, he argued: …it would do little harm for a government agency to speculate in the exchange market provided it held to the objective of smoothing out temporary fluctuations and not interfering with fundamental adjustments. And there should be a simple criterion of success-whether the agency makes or loses money. (Friedman, 1953, p. 188) The profitability criterion is intuitively appealing and seemingly objective criterion with which to assess central bank intervention. Moreover, it also has the advantage of being easily quantifiable. However, using data for the United Kingdom, we show that a theoretical argument developed by Carrado and Taylor (1986) showing that ex post profitability measures will inevitably be biased by the amount of cumulative intervention arbitrarily considered constitutes an important criticism of the profitability criterion and has empirical relevance. This paper also helps to reconcile the conflicting results of existing studies, some of which find central bank intervention to have been profitable, such as Argy (1982) and the Bank of England (1983) and others, which have found it to be loss making such as Jacabson (1984) and Taylor (1982).

In evaluating intervention policy, the profitability criterion while intuitively simple and appealing and having the advantage of being measurable suffers the problem of being biased by the amount of cumulative intervention arbitrarily considered. This is linked to the tendency of central banks to “lean against the wind” in their intervention policies. Reported profits/losses are usually a function of the amount of net cumulative intervention. If net cumulative intervention is low, one is likely to observe the central bank making a profit, while if net cumulative intervention is large one is likely to observe the central bank making a loss. Ex post profitability measures are biased to profits or losses depending upon how much cumulative intervention is arbitrarily allowed for. Since the profitability measure is not a useful criterion for assessing central bank exchange market intervention, this suggests the need for an alternative yardstick. However, any alternative assessment criterion will inevitably be plagued by problems of quantification. Studies such as Mayer and Taguchi (1983) and Williamson (1983) have suggested that exchange rate intervention can be regarded as being successful if it moves the exchange rate towards some “equilibrium” value but any such measures will crucially be dependent upon a satisfactory concept of the equilibrium exchange rate. An alternative approach would be along the lines of Edison, Miller, and Williamson (1987), which is based on simulating the effects of intervention on the exchange rate and then tracing the effects of this in a macroeconomic model on the objective function of the authorities. Whether such alternative measures can be considered less arbitrary will depend upon how much agreement economists are likely to achieve over the correct model governing the economy. This suggests we may be a long way from having a truly objective criterion for assessing the actions of the authorities in the exchange market.