سیاست های پولی در طول حملات سوداگرانه: آیا عوارض نامطلوب میان مدت وجود دارد؟

This paper extends the currency crises model of Aghion, Bacchetta and Banerjee (2000, 2001, 2004) in different directions. Our main result is that a tight monetary policy can have adverse effects beyond the short term and can potentially cause a currency crisis in the medium term, even in cases when the interest rate defense is successful and prevented a currency crisis in the short-run. In addition, we add a risk premium and find that this increases the likelihood of a crisis, can help explain contagion, and that prospective capital controls will increase the likelihood that such controls will be needed as an emergency measure

The unexpectedness and the severity of the Southeast Asian crisis demonstrates the importance of understanding the underlying mechanisms of currency crises as well as the optimal policy response to an emerging crisis. One of the findings in the immediate aftermath was the inadequacy of the existing models of currency crisis to explain the events. Neither the first-generation models relying on inconsistent economic policy nor the second-generation models where the government faces a trade-off between maintaining the fixed exchange rate and a desire to boost output and employment by pursuing a more expansionary monetary policy are consistent with empirical evidence.1 The Asian countries had an impressive economic performance the years preceding the crisis. After the crisis, following the depreciation, they all experienced deep recessions in which output fell at unprecedented rates. These empirical observations motivated the development of a third-generation of currency crisis models.2 Common within this latter approach is that they focus on stock rather than flow variables. There is no common denominator of which stock variables to focus on or the transmission mechanism, however. One strand of the literature emphasizes vulnerabilities in the financial sector Chang and Velasco, 2000, Chang and Velasco, 2001, Corsetti et al., 1999b and Corsetti et al., 1999a while another strand focus on adverse impacts of foreign currency denominated liabilities on the balance sheets of the corporate sector Aghion et al., 2000, Aghion et al., 2001, Aghion et al., 2004, Bergman and Hassan, 2008, Cho and Kasa, 2008, Krugman, 1999, Miller et al., 2005 and Miller et al., 2006. One of the appealing aspects of these frameworks is that the expectation of a depreciation can be self-fulfilling and thus enabling the existence of multiple equilibria. One key issue within the third-generation framework concerns the role of monetary policy. Krugman (1999), Aghion et al., 2000, Aghion et al., 2001 and Aghion et al., 2004, Cho and Kasa (2008), and Miller et al., 2005 and Miller et al., 2006 all suggest that a tight monetary policy both prevent and resolve a currency crisis. However, there are also studies suggesting that a loose monetary policy and a depreciated currency is the optimal response, see for example Furman and Stiglitz (1998), Gertler, Gilchrist, and Nataluccim (2007), Céspedes, Chang, and Velasco (2004), Christiano, Gust, and Roldos (2004) and Bergman and Hassan (2008).3 In this paper we extend the Aghion et al., 2000 and Aghion et al., 2001 (ABB) model in different directions.4 Our main focus is on the question whether the short-run policy recommendation of raising the interest rate can have adverse medium term effects such that it causes a currency crisis in the medium term even in cases when the interest rate defense was successful in preventing a currency crisis in the short-run. We show that adverse medium-term effects are present and that they can potentially reverse the effects of tight monetary policy, i.e., causing a currency crisis in the medium term even though the policy was successful in preventing a crisis. This feature of our model provides an explanation to the empirical observation that tight monetary policy may work for some countries but will not prevent a crisis for other countries. Second, we add a risk premium on domestic assets and show that the existence of a risk premium increases the likelihood of a crisis both in the short-run and in the medium term. More importantly, the inclusion of a risk premium also allows us to discuss contagion and prospective capital controls. If we assume that the risk premium is time-varying and that negative shocks in one market affect subsequent returns in other markets leading to increases in the risk premium, our model suggest an explanation to why healthy economies could be vulnerable to speculative attacks. We also find that an expectation of future capital controls (an increased risk premium on domestic assets) can increase the likelihood that such controls will be needed. In particular, there is a trade-off between having the option of a beneficial emergency use of capital controls and the adverse effect on the risk premium stemming from the possible future imposition of such controls. The remainder of the paper is organized in the following way. The next section presents the essential features of the ABB (2000, 2001) model including some aspects not covered in ABB’s original article. In the following section we focus on the medium term properties of the model where we also consider the effects of monetary policy. As a special case, we also add a risk premium to the model and use the risk premium to explain contagion. The next section focuses on the effects of prospective capital controls. The final section summarizes our main results.

This paper re-examines the Aghion et al. (2001) model and extends the analysis in different directions. Our main contribution is that we explicitly examine the medium-term implications of the model and conclude that we cannot exclude the possibility of neither multiple equilibria nor crisis in the medium term. This finding is important both for the understanding of crises as well as for the policy recommendations drawn from the model. Specifically we find that for some characterizations of the economy the short-term policy recommendation of raising the interest rate can have an adverse effect beyond the short term that could potentially cause a crisis in the medium term. Even though an interest defense is successful in preventing a currency crisis in the short-run, it may cause a currency crisis in the medium term. This has the implication that policy makers must carefully evaluate the characteristics of their economy before bringing the different monetary policy advice to bear. This feature of the model is also consistent with empirical observations that interest rate defenses have been successful in some cases but unsuccessful in other. We then extend the model by assuming that agents are risk avert such that a risk premium on domestic assets is added to the UIP relation. Not surprisingly this has the impact of making the economy more vulnerable to crisis, but in addition a risk premium can help explain contagion of healthy economies previously characterized by the “good” scenario if the risk premium adjusts to negative shocks in other markets. Furthermore adding a risk premium provides a means for a theoretical inspection of the effect of prospective capital controls. We find that an expectation of future capital controls, by raising the risk premium, can increase the likelihood that such controls will be needed. This means the policy maker faces a trade-off between having the option of a potentially beneficial emergency use of capital controls and the adverse effect on the risk premium stemming from the possible future imposition of controls. An implication of this finding is that it could be optimal for the authorities to relinquish this option if this can be done credibly.