This paper presents a simplified model of exchange rate behavior under a target zone regime. It shows that the expectation that authorities will defend the band exerts a stabilizing effect on exchange rate behavior within the band, even when the authorities are not actively intervening. The extent of stabilization can be related to three factors: the sensitivity of the current exchange rate to expected depreciation, the volatility of the process driving exchange rate "fundamentals", and the credibility of the commitment by authorities to defend the target zone. The exchange rate is assumed to be determined by two factors: "fundamentals", which evolve exogenously over time, and its own expected rate of change. The model shows that the target zone leads to more stable exchange rate behavior within the zone than free floating. The extent of stabilization can be related to the volatility of the fundamentals driving the exchange rate, the sensitivity of the exchange rate to expected appreciation or depreciation, and the credibility of the government commitment to defend the target zone. The model shows that the target zone has a stabilizing effect on the exchange rate within the band. The S-curve is always flatter than the 45-degree line, implying that fluctuations in the fundamental x are less than fully reflected in the exchange rate. This stabilization takes place even when the authorities are not actively defending the band: they only have to act when the fundamentals reach the levels x or x. The market recognizes this, and this realization creates regressive expectations that stabilize the rate. The key is that g(x) reaches a maximum of s at x. Thus we may differentiate (11) to get ∂s/∂x|_x = 1 + rA[e^{rx} - e^{-rx}] = 0, implying A[e^{rx} - e^{-rx}] = -1/r. At the same time, s = x + A[e^{rx} - e^{-rx}]. Thus we have x - s = 1/r = (γσ²/2)^{1/2}. The stabilizing effect depends on two parameters: γ, which is the sensitivity of the exchange rate to the expected rate of depreciation, and σ², the volatility of the fundamentals. The paper also considers the case of imperfect credibility. If the authorities do defend the band, they will now have a fully credible target zone; if they do not, the system will revert to free floating. The consequence of such a test is shown in Figure 3. The market knows that one of these two events will occur, and since any expected jump would yield an infinite expected rate of appreciation, x is defined implicitly by the requirement that the expected jump equal zero. The main substantive result is that the expectation that the authorities will defend a target zone will exert a stabilizing influence on exchange rate behavior inside the zone. The extent of this stabilizing influence depends on the sensitivity of the current exchange rate to exchange rate expectations, the volatility of the underlying determinants ofThis paper presents a simplified model of exchange rate behavior under a target zone regime. It shows that the expectation that authorities will defend the band exerts a stabilizing effect on exchange rate behavior within the band, even when the authorities are not actively intervening. The extent of stabilization can be related to three factors: the sensitivity of the current exchange rate to expected depreciation, the volatility of the process driving exchange rate "fundamentals", and the credibility of the commitment by authorities to defend the target zone. The exchange rate is assumed to be determined by two factors: "fundamentals", which evolve exogenously over time, and its own expected rate of change. The model shows that the target zone leads to more stable exchange rate behavior within the zone than free floating. The extent of stabilization can be related to the volatility of the fundamentals driving the exchange rate, the sensitivity of the exchange rate to expected appreciation or depreciation, and the credibility of the government commitment to defend the target zone. The model shows that the target zone has a stabilizing effect on the exchange rate within the band. The S-curve is always flatter than the 45-degree line, implying that fluctuations in the fundamental x are less than fully reflected in the exchange rate. This stabilization takes place even when the authorities are not actively defending the band: they only have to act when the fundamentals reach the levels x or x. The market recognizes this, and this realization creates regressive expectations that stabilize the rate. The key is that g(x) reaches a maximum of s at x. Thus we may differentiate (11) to get ∂s/∂x|_x = 1 + rA[e^{rx} - e^{-rx}] = 0, implying A[e^{rx} - e^{-rx}] = -1/r. At the same time, s = x + A[e^{rx} - e^{-rx}]. Thus we have x - s = 1/r = (γσ²/2)^{1/2}. The stabilizing effect depends on two parameters: γ, which is the sensitivity of the exchange rate to the expected rate of depreciation, and σ², the volatility of the fundamentals. The paper also considers the case of imperfect credibility. If the authorities do defend the band, they will now have a fully credible target zone; if they do not, the system will revert to free floating. The consequence of such a test is shown in Figure 3. The market knows that one of these two events will occur, and since any expected jump would yield an infinite expected rate of appreciation, x is defined implicitly by the requirement that the expected jump equal zero. The main substantive result is that the expectation that the authorities will defend a target zone will exert a stabilizing influence on exchange rate behavior inside the zone. The extent of this stabilizing influence depends on the sensitivity of the current exchange rate to exchange rate expectations, the volatility of the underlying determinants of