بحران ارز و گرایش نرخ پیش فروش:مدارک و شواهد اقتصادهای در حال ظهور تحت معافیت

This paper investigates the effects of two financial crises (the 1997 Asian currency crisis and the 2000 Turkish financial crisis) on the forward discount bias in 14 emerging-market economies using a robust two-stage procedure. This unique sample of less researched currencies displays: (i) high persistence in forward discount equations; and (ii) varying variance ratios between changes in exchange rates and the forward premium. The findings provide new insights into the forward discount puzzle: financial crises exert considerable power on the forward discount bias and uphold the forward rate unbiasedness hypothesis (FRUH) by reverting the negative sign into positive.

A substantial amount of research has been conducted on the forward discount bias, a puzzle in which the forward rate inversely predicts the future spot exchange rate; see the review paper by Engel (1996).1 Specifically, the forward discount bias refers to the case where the slope coefficient in the regression of the change in the future spot rate on the current forward discount is statistically negative rather than unity. This phenomenon contradicts the forward rate unbiasedness hypothesis (FRUH) and violates the theory of the uncovered interest parity (UIP). The FRUH stipulates that the forward rate should be an unbiased predictor of the future spot rate while the UIP denotes that the expected future change in the spot rate is determined by the interest rate differential between two countries. This violation implies that currency excess returns are predictable by the forward discount. In fact, this anomaly has been economically exploited as a profitable trading strategy by professional currency traders through carry trades (in which one borrows in low-yield currency and invests in high-yield currencies) since the beginning of the 21st century. See Phillips and Snow (1998) for an early paper and Villanueva (2007b) and Clarida et al. (2009) for more recent studies. This phenomenon has directed a considerable amount of research on verifying the potential explanations. Among these, several major competing explanations have been advanced to explain the anomaly, with a focus on the major currencies of the advanced economies. None has succeeded in fully accounting for the violation of FRUH. This paper intends to address the forward discount bias from a different perspective. The purpose of this paper is to investigate the effect of two financial crises (specifically the 1997 Asian currency crisis and the 2000 Turkish financial crisis) on the forward discount bias in the 14 emerging economies with and without crises via the proposed two-stage analysis. This paper differs from previous studies in several ways. First, this paper explores the effect of two major financial crises in emerging markets on the forward discount anomaly, which has only scarcely been explored in previous studies. Second, this paper employs a systematic two-stage analysis from simple to complex approach which has not been conducted before. Third, the results from this paper are expected to shed new insight on the forward discount puzzle and related carry trades in light of whether real economic gains are justified or sustainable. This paper contributes to the current literature in at least three ways. First, for the emerging countries, the results support the findings of Flood and Rose (2002) that the forward discount works better during crisis periods than non-crisis periods. The improvement lies in two aspects: For one, this paper is able to accurately differentiate crisis periods from non-crisis periods by utilizing Bai and Perron, 1998 and Bai and Perron, 2003 endogenous multiple structural breaks and consequently make more reliable inferences than Flood and Rose (2002). Conversely, the latter simply lumped the 1990s together as the crisis period and ran the difference regression and made inferences that UIP works better in crisis periods than other periods. Furthermore, they did not include non-crisis periods in the paper to justify their comparison between crisis and non-crisis periods. Rather, they compared their results to prior literature in general (i.e. treating prior literature as non-crisis periods). Hence, their inference could be inaccurate. Second, this paper adds a new dimension to the forward discount that structural changes associated with financial crises have a dominating effect on the forward discount compared with any other structural breakpoints. Third, the difference regression in this paper also supports the findings of Frankel and Poonawala (2010) that the forward discount in emerging markets are less biased than that in major currencies. While prior literature has mostly examined the major currencies with virtually all estimates of the forward discount as being negative, this paper goes beyond Frankel and Poonawala (2010) by focusing on the crisis effect in the emerging economies on the forward discount by employing a sequentially searching multiple structural breakpoint model.2 Our results provide new evidence as follows. First, financial crises exert a powerful effect on the forward discount bias: under proper identification of the breakpoint, the modified regressions uphold FRUH by reverting the negative coefficient into positive in the crisis countries during the crisis periods. Second, structural breaks associated with financial crises exhibit a dominating (magnified) effect on the forward discount bias surrounding crisis breakpoint than any other structural breakpoints. Third, the crisis effect on the forward discount bias is larger for the crisis countries than for the non-crisis countries. The findings of this paper have several important economic implications. For currency traders, the results confirm the strategy of trading the forward discount bias in non-crisis periods and unwinding carry trade positions in crisis periods, as evidenced by the yen–dollar carry trade in the most recent global financial crisis that began in late 2007. Furthermore, knowing that trading forward discount bias is not only statistically significant but economically important in the surging carry trades that attract speculative capital over the last two decades (Hochradl and Wagner, 2010), central banks of high-carry currencies can unexpectedly lower interest rates to avoid sudden unwinding of carry trades that may trigger currency crash and fend off speculative attacks, while governments of funding currencies can tighten credit policy to reduce speculative capital and maintain currency stability before crisis. The remainder of the paper is organized as follows. Section 2 provides literature review. Section 3 describes the data and proposes the two-stage analysis to examine the effect of financial crises on the forward discount bias in 14 emerging economies. Section 4 presents the empirical results and analysis. Section 5 summarizes and concludes.

Our analysis of the forward discount bias in the 14 emerging currencies via the difference regression with endogenous multiple structural break models has provided interesting results that financial crisis exerts a magnified impact on the forward discount bias. Our results can be summarized as follows. First, once the structural breaks are accounted for, seven (four from the Asian group and three from the non-Asian group) of the 10 cases in the emerging markets where the FRUH did not hold in the original regression without breaks now uphold the FRUH in the difference regression with breaks during the crisis periods. Second, there exists a dominating effect of financial crises on the forward discount because only the regime changes brought about by financial crises could alter the conventionally negative coefficients into positive and preserves the unitary slope hypothesis (supporting the FRUH) during the crisis period while all other breakpoints do not possess that power.5 The structural breaks associated with financial crises therefore possess global effect while other structural changes brought about by government interventions or exogenous shocks exert local impact. Third, the crisis effect on the forward discount is more conspicuous and larger in size for the crisis countries (as in the cases of Indonesia, Malaysia, Thailand, and Turkey) than for the non-crisis countries (as in the cases of Hong Kong, India, and Mexico). In the former cases, the hypotheses of α = 0, β = 1, and Et (ɛt+1) = 0 cannot be rejected, satisfying all three criteria for the FRUH during the crisis periods, while the sign or the magnitude of the forward discount in the latter cases does not change. Our findings have important implications for policy makers and global investors. For policy makers, it is insightful to know that structural breaks induced by financial crises could reverse the sign and size of the forward discount and lead to its correct direction. For international investors including those who are active in carry trades, it is crucial to know that during crisis the forward discount may actually predict the future spot rate in the right direction. In this paper the emphasis is on the misspecification for not properly taking into account the structural breaks. Gospodinov (2009) has dealt with this but did not consider structural breaks as we do. One useful analogy of our research methodology rests within the recent carry trade opportunities. Borrowing from the low-interest-rate countries and investing in high-yield countries may not generate the expected excess returns because the high-interest rate countries may indeed depreciate their currencies and render the converted proceeds from abroad less valuation than in normal times. A reexamination of the UIP condition under carry trades by Clarida et al. (2009) suggests that the negative coefficient on the forward premium in low volatility currencies turns into positive in high volatility currencies. Our results are along the same lines but with a more elaborate treatment of the structural breaks.