تجارت بازخورد و اثرات متقابل همبستگی در بازار ارز: شواهد بیشتر

This paper tests for presence of feedback trading, asymmetric behavior and autocorrelation linkages in several industrial and emerging economies' exchange rates, with respect to the US dollar, as well as the Euro. The issue is examined via the means of a GARCH-augmented feedback model for the period of 1990 to 2003. The empirical results indicate presence of feedback trading and/or asymmetric behavior in both types of economies' exchange rates but absence of such behavior in the Euro. Presence of asymmetric behavior implies that market traders rely on central banks to intervene so they can realize short-term profits. Furthermore, evidence of volatility persistence in several exchange rates implies inefficiency in those markets. Finally, there are instances where the first-order autoregressive parameter is positive and statistically significant in the exchange rates of both industrial and emerging economies but not in the Euro. For the latter currency, lack of asymmetric behavior and feedback trading implies a credible currency in the eyes of foreign exchange traders.

Some investors attempt to identify trends in past stock prices and base their portfolio decisions on expectations derived from such trends. Such behavior is termed feedback trading. Feedback trading can either be positive or negative and each type presents some concerns for every market participant. The concern about positive feedback trading is that it makes financial asset prices to overreact to new information and can be considered desirable or unpleasant. For instance, investors with positive feedback strategies can be regarded either as destabilizers (or noise traders) or stabilizers. The first could occur because their sales contribute to the fall of the market and their purchases add to market advances. Therefore, if such trading tends to destabilize economies, it can have a serious impact on the emerging economies and the benefits from the liberalization of their markets can be diminished. Moreover, in several instances such strategies generate volatility in returns and create bubbles which may lead to market crashes when they burst. The second possibility could take place if trades of such investors are related to changes in risk premiums or permanent price changes. Financial market liberalizations of emerging economies usually attract more and diverse investors, which results in market appreciations and more inflows of capital. Such trading represents an important aspect of the functioning of financial markets since it reduces the risk of market crashes and eases the flow of transactions among participants. Negative feedback trading exists when investors ‘buy low and sell high’, that is, engage in selling stocks following price increases and in buying stocks following price declines. Negative feedback trading can be the result of profit-taking as markets advance or from investment strategies that specify a target wealth in a portfolio. In the context of the foreign exchange market, positive feedback trading exists when traders buy/sell after a depreciation/appreciation of the exchange rate, whereas negative feedback occurs when traders buy/sell following an exchange-rate appreciation/depreciation. Essentially, positive feedback traders tend to stabilize the currency since their strategies typically follow hedging opportunities and extensive use of stop-loss orders. By contrast, negative feedback trading tends to emerge from the traders' efforts to realize profit as the exchange-rate appreciates (or depreciates) thereby driving the exchange rate's value away from its long-run value. These situations can be plausibly assumed because since emerging markets have fully opened up their markets to foreign investors and, additionally, their returns have become more closely correlated with the returns of developed economies. Another consequence of the presence of a sufficient number of feedback traders in the foreign exchange market is the autocorrelation of returns, which is related to the extent of predictability in the foreign exchange returns. Recent evidence suggests that autocorrelation patterns are present and are multifaceted, at least in stock markets [e.g., see LeBaron (1992), for such dependencies in US stock returns, and Campbell et al. (1993), on the relationship between trading volume and stock return autocorrelation]. Regarding the foreign exchange market, Allen and Taylor (1990) indicate that most traders consider trend patterns at least as relevant as the market fundamentals in the determination of exchange-rate expectations in the short run. Similarly, Frankel and Froot (1987) find evidence of extrapolative expectations and credit it to the use of trend chasing (or charts) by professional traders. Vitale (2000) finds that noise trading in the foreign exchange market may be used to exploit expectations and exchange rates in order to achieve an informational advantage and ultimately a profit opportunity. Finally, Laopodis (2004) finds evidence of noise trading in the foreign exchange markets of both industrial and emerging economies. A related issue concerns the degree of efficiency in the foreign exchange market. Specifically, if volatility is persistent then the currency market in question is not efficient since news should be fully and immediately incorporated in the exchange rate and not take a long time to be assimilated. A final and an equally important issue this paper seeks to address is the possibility of asymmetric behavior in feedback trading. Specifically, under the presence of asymmetry, is positive feedback trading more intense during exchange-rate appreciations than during exchange-rate depreciations? In other words, do informed traders have an informational advantage over the noise traders or is it the other way around? Evidence on these issues is mixed for the equity and foreign exchange markets (e.g., Sentana and Wadhwani, 1992, Frankel and Froot, 1987, Aguirre and Saidi, 1999 and Vitale, 2000). A related issue concerns the credibility in the foreign exchange market, regarding a particular currency, under presence of feedback trading. For example, positive feedback trading connotes that noise traders expect further appreciations/depreciations to take place, based on past appreciations/depreciations, and thus engage in sell/buy actions. In other words, these traders believe that a currency can be sustained only when it has depreciated but not when it has appreciated, therefore suggesting that asymmetric behavior is prevalent in the market under exchange-rate appreciations. Overall then, the purpose of this paper is to further empirically determine whether feedback trading exists in the foreign exchange market for several developed and emerging economies and whether, in particular, noise trading results in stabilization or destabilization in the foreign exchange market. In this respect, the paper adds to the insights found by Laopodis (2004) and extends his analysis by investigating more issues as well as the Euro's behavior. These issues will be addressed in this paper via the use of a positive feedback trading model augmented with a GARCH specification for the variance of the foreign exchange returns. The analysis includes an analysis of the specifics of the exchange rates and extends the empirical analysis by investigating the structural changes in these financial markets as well as the introduction of the Euro currency. Finally, a subperiod analysis is included for the purpose of examining the impact of specific financial crises on the nature of feedback trading in these markets. The rest of the paper is organized as follows. Section 2 lays out the methodological design of the study and specifies the empirical model. Section 3 contains the data description and presents some preliminary statistics. Section 4 presents the main empirical findings and discusses them, while Section 5 summarizes and concludes the study.

This paper examines the possibility of noise trading and autocorrelation patterns of 7 developed and 10 emerging economies' exchange rates with respect to the US dollar for the 1990 to 2003 period. The paper also considers the Euro's behavior from 1999 to 2003. More specifically, the paper seeks to empirically determine whether the presence of feedback trading strategies is a distinguishing feature of an emerging economy or it is a common element to both developed and emerging economies. Additionally, we examine whether there exist asymmetric behavior, volatility persistence, and credibility in the foreign exchange market following such trading. These questions are tested via the use of a GARCH-augmented feedback model. The results reveal an evidence of feedback trading and/or asymmetric behavior in both types of economies. Presence of asymmetric behavior implies that market traders rely on central banks to intervene so they can realize short-term profits and such actions, both by smart money and noise traders, could be destabilizing in the foreign exchange market. However, we found an absence of feedback trading and/or asymmetric behavior in the case of the Euro, which suggests that the currency is viewed as credible in the eyes of foreign exchange participants. Further, for several exchange rates we found high volatility persistence, which implies inefficiency in these currency markets. This result also implies that during volatile periods, deviations from a long-run value in an exchange rate are likely to increase since informed and noise traders exert a greater influence on the exchange rate. Finally, there are instances where the first-order autoregressive parameter is positive and statistically significant, in the currencies of both developed and emerging economies. This finding implies the presence of the so-called bandwagon effect, whereby past currency movements are followed by expectations of currency movements in the same direction.