تجزیه و تحلیل تجربی بر نقش سفته بازان بزرگ در بحران ارز

Corsetti et al. (2004) demonstrate that the presence of a large speculator in the foreign exchange market makes the remaining traders more aggressive in their speculative attacks. We conduct an experiment designed to test their theoretical predictions and also use the experiment to analyze an additional aspect that has not been previously covered in the literature: namely, whether the entry of a large speculator and the exit of the same speculator have the same effect in magnitude on the probability of a successful speculative attack. We obtain two main findings. First, the results support the main conclusion of Corsetti et al. (2004) that the presence of a large speculator makes other small speculators more aggressive. Second, the results suggest that the effect of the entry of a large speculator on the probability of successful speculative attacks is larger than that of the exit of the same speculator.

A common view underlying accusations against large speculators, like George Soros and Julian Robertson, is that they can exercise a disproportionate influence on the likelihood and severity of currency crises. Examples of crises in which this common view was debated include the ERM crisis of 1992, in which George Soros was named “the man who broke the Bank of England”, and the Asian crisis of 1997, in which he was accused of being included among “the anarchists, self-serving rogues, and international brigandage” by the then Prime Minister of Malaysia, Mahathir Mohamad.1 The literature has not yet reached a consensus on the question whether this common view behind these accusations is correct. Corsetti et al. (2004) raise this question theoretically while Brown et al. (1998), Fung and Hsieh (2000), Fung et al. (2000), and Corsetti et al. (2002) present it empirically. Corsetti et al. (2004) provide a theoretical prediction that the presence of a large speculator makes all other speculators more aggressive in their attacks on a currency peg. Unfortunately, empirical studies in this area provide no clear answer. In this paper, we use an experimental analysis to contribute to the literature. The experimental analysis provides an interesting alternative to the empirical analysis, because data constraints may make it difficult to empirically discern the effects of the presence of a large speculator in speculative attacks on a currency peg. Data constraints arise because large speculators’ personal funds are typically registered in so-called tax havens and they do not have to disclose data as regulations are far less stringent. As a result, it is difficult to obtain sufficiently detailed data to empirically determine the role of a large speculator in these attacks. The contribution of this paper is twofold. First, we conduct an experiment designed to test the theoretical predictions of a speculative-attack model following Corsetti et al. (2004). Second, we also use the experiment to analyze an additional aspect that has not been previously covered in the literature: namely, whether the entry of a large speculator and the exit of the same speculator have the same effect in magnitude on the probability of a successful speculative attack. Our main findings can be summarized as follows. First, the results of the experiment support the main conclusion of Corsetti et al. (2004) that the presence of a large speculator causes other speculators to be more aggressive in their attacks. Second, the results also suggest that the effect of the entry of a large speculator is larger than that of the exit of the same speculator. Policymakers should take the second result into consideration if they impose or remove limitations on speculative position holdings for the large speculator. The paper is organized as follows. Section 2 reviews the literature. Section 3 explains the speculative-attack model used in our experiment. Section 4 describes the experimental design. Section 5 presents the results. Section 6 discusses an implication of the results that may be related with a possible regulatory aspect. Section 7 concludes the paper.

This paper reports the results of an experiment designed to test the theoretical predictions of Corsetti et al. (2004). The results of the experiment support the main conclusion of Corsetti et al. (2004) that the presence of a large speculator makes other small speculators more aggressive. Moreover, the results suggest an asymmetric effect between regulating and deregulating the speculative position holdings limitation, which was not a concern of Corsetti et al. (2004). This may provide us with some suggestions for constructing a dynamic model by extending the original model. For future research, we are planning to conduct experiments to test other related theoretical papers. For example, Bannier (2005) theoretically highlights the role market sentiment has on the impact of a large trader in a currency crisis. Bannier (2005) shows that the large trader makes market responses more aggressive only if the market believes economic fundamentals to be weak (i.e., market sentiment is pessimistic) while the presence of the large trader decreases the ex ante probability of a crisis under optimistic market sentiment. Conducting experiment on Bannier (2005) would be useful to investigate whether or not this is related with the asymmetric effect discussed in Section 6.24 Likewise, Corsetti et al. (2006) investigate the role of the official creditor (the IMF or an international lender of last resort) as a large player in the world economy. It is very likely that it would also be challenging to estimate these models because of data constraints. Experimental analysis would then be able to provide some useful information to accompany the literature.