مطالعه رفتار نقل قولی اسپرد قیمت خرید و فروش فروشندگان ارز خارجی

Based on a questionnaire surveying dealers in the Taipei inter-bank foreign exchange market that was conducted in March 2001, I attempted to answer the question of who initiated the wider currency spread. It was found that the risk-averse dealers of small banks quoted wider spreads in order to conceal their inferior positions regarding information and inventory or to avoid market volatility risk. Some of the dealers of large multinational banks in major financial centers who normally quote conventional spreads were found to quote wider spreads in response to the request for quotations by small Taiwanese bank dealers who widened their spread quotes.

Foreign exchange dealers set conventional spreads and fear that widening spreads in a turbulent market might damage an equitable and reciprocal trading relationship, their market image, and potential trading opportunities. Non-conventional spreads can drive away potential trading opportunities and the dealers' ability to read the market. This is confirmed by a series of recent questionnaire surveys involving foreign exchange dealers in three East Asian financial centers, the U.S., and London foreign exchange markets by Cheung and Wong (2000), Cheung and Chinn (2001), and Cheung et al. (2004). In addition, studies performed by Bollerslev and Domowitz (1993) and Huang and Masulis (1999) have found that while market competition increased when large multinational bank dealers' market shares increased, bid–ask spreads consequently declined. By contrast, Bollerslev and Melvin (1994) found that an overwhelming majority of quotations involved a shift in the spread from previous quotations in Reuters FXFX banks' quotation data. A growing body of literature on the microstructure of foreign exchange markets suggests that foreign exchange dealers adjust bid–ask spread quotes to change inventory carrying costs at market close (for example, Boothe (1988), Bessembinder (1994) and Jorion (1996)), avoid transaction risk (Glassman (1987), Bessembinder (1994), Bollerslev and Melvin (1994), Wei (1994), Goodhart and Payne (1996), and Louis et al. (1999)), reduce overnight inventory risk (Chakrabarti (2000)), compensate for adverse selection costs (Perraudin and Vitale (1996) and Bossaerts and Hillion (1991)), avoid adverse selection at market open (Naranjo and Nimalendran (2000), Payne (2003), and Bjonnes and Rime (2005)), and affect trading volume (Glassman (1987), Bessembinder (1994), Hartmann (1999)). How do we reconcile these two strings of seemingly inconsistent findings? If dealers fear that widening spreads might drive away potential counterparts and continue to quote conventional spreads even in turbulent periods, then who initiates non-conventional spreads, and why? Why have so many authors found that dealers have changed their quoted spreads on many different occasions?1 Why then do small bank dealers frequently widen spreads? Bollerslev and Domowitz (1993) pointed out that risk-averse smaller bank dealers need to widen their spreads to control inventory at the end of their regional trading hours as well as to avoid trading losses due to their inadequate information on market order flow at the day's opening. In this paper, I discover that local bank and foreign bank dealers in Taiwan who deal with dealers in other major markets are inclined to quote wider spreads when important news is disclosed and market risk increases. The trading sizes are considerably small for local bank and foreign bank dealers in Taiwan compared with major bank dealers in other financial centers. Goodhart and Figliuoli (1991) reported wider spreads emanating from countries in the Far East at the beginning of European trading hours. Why do dealers in countries in the Far East quote wider spreads at the beginning of European trading hours? This is no doubt because those small bank dealers in the countries in the Far East eagerly square their overnight positions at the end of regional business hours which is also the beginning of European trading hours. They quote wider spreads in order to avoid receiving unwanted positions or revealing information on their own unfavorable positions. The evidence of this study indicates that when trading with dealers in other major financial markets, local bank and foreign bank dealers in Taiwan are inclined to quote wider spreads when important news is disclosed and market risk increases. The foreign bank dealers in Taiwan widen spreads when price volatility is high or when their counterparts widen their spreads which may signal possible market price changes or unknown information. Those dealers of larger banks, who normally quote conventional spreads, also widen spreads when trading with small banks that quote non-conventional spreads in order to accommodate the induced inconvenience and increased trade frequency involving smaller trades. Consequently, large multinational bank dealers in major financial centers will quote wider spreads in response to the request for quotations by small (Taiwanese) bank dealers who widen spread quotes. This inference agrees with Goodhart and Figliuoli's (1991) finding that some of the larger spreads emanate from Far East Asian dealers during the European trading time. Cheung et al. (2004) reported that 16% of dealers in London quote non-conventional spreads when trading with ill-informed and small banks. The number is even higher in Taiwan, for about 32% of dealers widen spreads when trading with small banks. Consequently, our arguments reconcile the dealers' widening spread behavior from time to time in a large number of empirical studies with the dealers' intentions of maintaining the conventional spreads even in turbulent periods in survey studies such as Cheung and Wong (2000), Cheung and Chinn (2001), and Cheung et al. (2004). Cheung and Chinn (2001) have pointed out that “it is difficult to gather market-wide data on foreign exchange trading.” Hence, the conventional research approach of theoretical modeling and estimation using either quoted or traded exchange rate data from various sources (such as Reuters, Bloomberg, Bridge, and EBS) can yield limited information on individual traders' beliefs. The advantage of using a survey approach is that it can generate information on the views, behavior, and practices with regard to the exchange rate quotations of heterogeneous market participants. The trader's view can be compared with the results from conventional empirical studies to strengthen our understanding of the foreign exchange market microstructure. Therefore, I conducted a questionnaire survey in March 2001 to more deeply explore the foreign exchange dealers' quoting behavior in regard to when, who, and why dealers widen bid–ask spreads. Due to accessibility issues, the questionnaire was circulated only among dealers in the Taipei inter-bank foreign exchange market. Therefore, I cannot explore and compare the dealers' quotation behavior in Taiwan with that of various markets around the world. I can only study and draw inferences from the dealers' quotation behavior in Taiwan and cannot derive more general implications. This is a limitation of my study. People may, then, question why it is necessary to study the behavior of dealers in Taipei. Why is it important to study dealers' behavior in Taipei? Based on the findings of Goodhart and Figliuoli (1991), the conventional quoted spreads for a currency vary across different markets. Their study pointed out that some larger spreads were brought about by Far East Asian dealers during the European trading time. Taiwan, one of the Far East Asian countries, had 53 local banks and 39 local branches of foreign banks at the end of March 2001.2 All of these banks engaged in foreign exchange-related business. The foreign banks in Taiwan are mostly large global banks and conduct a broad range of foreign exchange business. The local branches of foreign banks are the local units that are part of a global foreign exchange trading presence. Approximately half of the foreign bank dealers' currencies traded involve non-local currencies, while roughly 50% of their trades are with dealers in other markets. Dealers in local branches of foreign banks are subjected to their respective banks' global trading policies and guidance. Consequently, they may behave similarly to their colleagues in other financial centers. A survey on foreign bank dealers in Taiwan allowed me to explore their behavior in bid–ask spread quotations, to make comparisons with Cheung and Wong's (2000) and others' findings, and study the when, who, and why questions related to dealers' widening spreads. The remainder of this paper is organized as follows. Section 2 describes the foreign exchange transaction status in Taiwan. Section 3 discusses the design of the questionnaire and its validity and reliability analysis. Section 4 details the method used to weight answers in cases where respondents were allowed to select more than one answer. Section 5 tests the homogeneity of behavior between local and foreign bank dealers. Section 6 analyzes the dealers' behavior in regard to widening bid–ask spreads. The concluding remarks and a brief discussion are presented in the final section.

A survey on dealers in Taiwan allowed me to explore their behavior in relation to bid–ask spread quotations, to compare their behavior with dealers in other financial centers, and to try to answer the questions as to who initiates non-conventional spreads and when these non-conventional spreads are initiated. I found that some small local bank dealers and dealers with local branches of global banks respond to possible losses resulting from unexpected news or exchange rate fluctuations by initiating wider spreads to accommodate risk. Studies by Bollerslev and Domowitz (1993), Huang and Masulis (1999), and Cheung and Wong (2000) show that multinational bank dealers tend to quote conventional spreads in a competitive market. However, when dealing with small banks, dealers in major Asian financial markets quote non-conventional spreads as Cheung and Wong (2000) and others found. As a result, when dealers of large multinational banks in major financial centers respond to small regional bank dealers' requests for quotations, they will quote wider spreads to accommodate induced inconvenience and increased trade frequency with smaller trades. In this way, our findings reconcile the behavior of dealers that typically maintain conventional spreads even in a turbulent market in survey studies with that of dealers who widen spreads from time to time as is the case in many empirical studies. When do dealers widen spreads? Local bank and foreign bank dealers were found to have some dissimilar considerations in terms of setting bid–ask spread quotes. Foreign bank dealers widen their spreads mainly following a counterpart's move, while local bank dealers widen spreads primarily to avoid loss of position from exchange rate fluctuations and to cover costs. The timing of decisions to widen bid–ask spreads also differs significantly between local bank and foreign bank dealers. Local bank dealers widen spreads mostly when price volatility increases, when important news is disclosed, and when markets open, while foreign bank dealers widen spreads primarily when price volatility increases, when trading with small banks, and as a reaction to a counterpart's actions to widen spreads.