Some Evidence that a Tobin Tax on Foreign Exchange Transactions may Increase Volatility
More than two decades ago, James Tobin suggested imposing a tax on all foreign exchange transactions (Tobin, 1978). Similar proposals for imposing transaction tax on trading of other securities (see Schwert and Seguin, 1993 for a review) are also often made by eminent economists (Stiglitz, 1989, Summers and Summers, 1989). One of the putative benefits of a transaction tax is that this may decrease the volatility of prices. The intuitive rationale behind this, believed to be first articulated by Keynes in 1936, is that a transaction tax would hurt the speculators disproportionately more because they tend to trade much more frequently. An implicit assumption in this argument is that speculative trading is on average destabilizing which in turn causes prices to be more volatile. A contrasting view is offered by Milton Friedman who argued (Friedman, 1953) that rational speculators may in fact help stabilize prices. The relative merits of these opposing views can only be judged by analyzing this issue empirically.
The empirical evidence on the effect of transactions taxes on volatility of prices is rare. Umlauf (1993), using Swedish stock market data from the 1980s, shows that the introduction of, or an increase in Swedish tax, led to an increase in volatility of stock prices; Jones and Seguin (1997) show that the reduction in the commission portion of the transactions costs in 1975 led to a decrease in volatility (and increase in volume) of stock prices. However, there appears to be no empirical evidence in the extant literature on the effect of a transaction tax in the foreign exchange market on the volatility of exchange rates. In this paper, we provide some evidence. Our results, suggest that a Tobin tax on foreign exchange transactions may, in fact, lead to an increase in the volatility of exchange rates which is exactly the opposite to the claim made by Tobin and proponents of his suggestion such as Jeffrey Frankel who recently resurrected Tobin's argument (see Frankel, 1996).
We estimate the effective transactions costs in the foreign exchange market for the period 1977 to 1999 using foreign currency futures data. Our approach in estimating the transactions costs in the foreign exchange market makes a contribution in two ways. First, we show that previous approaches for estimating the transactions costs in the foreign exchange market (Frenkel and Levich, 1975, 1977, 1979 and McCormick, 1979) had some methodological problems arising from the incorrect use of bid and ask price data and from the use of non-synchronous data in the spot and forward markets. Second, we measure transactions costs faced by the marginal investors that set prices in the foreign exchange markets which is unlike the previous approaches which, if implemented correctly, would measure transactions costs in foreign exchange market that are faced by commercial customers of banks. While the estimates of transactions costs useful for judgements about the impacts of alternative exchange rate regimes on the levels of trade, might be those incurred by commercial firms, the estimates of cost relevant for determining prices, in contrast, are the smaller costs incurred by large commercial banks who are likely to be marginal investors determining prices in the foreign exchange market. We estimate that transactions costs over the last two decades on average were no more than one-twentieth of one percent, and in the last decade may have fallen to as low as one-fiftieth of one percent.
We then, using our approach, construct time series of monthly estimates of effective transactions costs for four currencies, the British Pound, the Deutsche Mark, the Japanese Yen and the Swiss Franc. We also construct time series of monthly volatility (i.e., standard deviation) of foreign currency futures returns and monthly volume (i.e., number of futures contracts traded) for the four currencies. Using regression analysis, we document that volatility is positively associated with the level of transactions costs and that volume is negatively associated with the level of transactions costs. Thus our results are consistent with the notion that an increase in transactions costs does indeed lead to a reduction in volume of trading as one might expect, but its effect on volatility is exactly opposite of what proponents of Tobin tax would have liked to see.