A new way of modeling the dynamics of an exchange rate target zone is presented. In the presence of a single upper (resp. lower) target boundary, the exchange rate is precisely represented as the sum of a free float and a short (resp. long) position in a call (resp. put) option with strike price at the boundary. To model a target zone (with two boundaries), a natural approach consists in describing the exchange rate dynamics as the combination of the two, namely the sum of free float together with a long position in a put written on the lower boundary and a short position in a call option written on the upper boundary, respectively. We show that this first order approximation leads to significant mispricing (as much as 20%) and must be iterated, leading to an infinite sequence of compounded 'mirrored' option prices. We analyze basic properties of such mirrored nested options analytically, describe how to calculate them numerically, and show why it is crucial to take into account higher order corrections in realistic target zones. We argue that this analogy to option prices allows for conceptually simple generalizations that describe different target zone arrangements. We apply our methodology to the estimation of the fundamental value of the Hong Kong dollar that is hidden by the target zone peg to the US dollar. We also estimate the implied maturity and explain how this parameter serves as direct proxy for target zone credibility.

This paper extends the Krugman target zone model by including a realignment mechanism. Various properties of that realignment mechanism are discussed. The movement of the exchange rate is governed both by a Wiener process on fundamental and by a Poisson jump process with endogenous realignment size. The realignment mechanism is such that (except in cases where a speculative attack occurs) no jump in fundamental is needed to accompany the jump in the exchange rate. A risk neutral valuation of currency options is constructed. Some properties of option values under realignment risk are illustrated by numerical results.

Currency target zones have been under scrutiny for the past three decades, which led to the development of two broad classes of quantitative models: Phenomenological ones that explicitly take into consideration the market's perception of the bounded exchange rate, and more mechanical ones that rely on put and call options. Until now, the two models have only been compared qualitatively. Here, we derive, for the first time, a quantitative link between these two approaches. Specifically, we show how the former approach has to be generalized in order to recover the second one. This mapping lets us relate the phenomenological parameter of the first approach to economically well-known quantities.