بسته شدن بازار سهام و نوسانات روزانه بازار آتی شاخص سهام معاملات: "سرایت"، تعصب بر پیشنهاد یا سوال و یا هر دو؟
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14340||2001||14 صفحه PDF||سفارش دهید||5245 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Pacific-Basin Finance Journal, Volume 9, Issue 3, June 2001, Pages 219–232
Chang et al. wJournal of Business 68 Ž1. Ž1995. 61x examine the impact of the closure of the New York Stock Exchange ŽNYSE. on S&P500 stock index futures traded on the Chicago Mercantile Exchange. They document a decline in futures market volatility immediately after the close of the NYSE, and an increase 15 minutes later when the futures market closes. They attribute this to contagion–i.e. a decline in information transfer from equities to futures markets following the closure of the underlying market. This paper examines the impact of the extension of trading hours in Hang Seng Index futures traded on the Hong Kong Futures Exchange on the 20 November, 1998 to 15 minutes after the close of the underlying market Žthe Stock Exchange of Hong Kong.. Using the unique natural experiment provided by this change, a pattern similar to US markets is documented for the Hang Seng Index Futures following the change in trading hours. This provides strong evidence that the intraday pattern in volatility is caused by market closure. Unlike US futures exchanges, price reporters on the floor of the Hong Kong Futures Exchange collect quote data in addition to trade data. This data facilitates a test of another plausible microstructure explanation for the observed behaviour–bid–ask bounce associated with trading activity. This paper provides evidence that bid–ask bounce also explains part of the observed intraday behaviour in price volatility. q2001 Elsevier Science B.V. All rights reserved.
Chang et al. (1995) examine the behaviour of stock index futures markets when underlying markets close. They argue that such research provides evidence on the impact of underlying market closure for institutional or regulatory reasons, such as 24-hour trading of derivatives and trading halts in stock markets. In motivating their research they note the relative dearth of literature examining the impact of underlying market closure on futures market behaviour [p. 66]: Most of the extant research on volatility presents results of analyses while both the underlying asset market and the futures market are open, but essentially no evidence seems to exist on the effect of closing of the underlying asset market on the futures market. Apart from the research by Chang et al. (1995), this motivation is still relevant today. This study aims to provide further evidence on this issue by examining the intraday trading behaviour of the Hang Seng Index futures contract traded on the Hong Kong Futures Exchange (HKFE). Chang et al. (1995) examine price volatility around the last 15 minutes of trading in S&P500 futures, which trade for 15 minutes longer than the New York Stock Exchange. On 20 November, 1998 the HKFE extended the trading hours of Hang Seng Index futures to commence trading 15 minutes prior to the opening, and continue for 15 minutes after the close of the underlying market (the Stock Exchange of Hong Kong). This change provides a natural laboratory experiment for extending the analysis of Chang et al. (1995). The change in HKFE trading hours also provides an opportunity to examine trading behaviour in stock index futures contracts when the underlying market opens. Chang et al. (1995) use two theoretical models to develop predictions on the behaviour of futures markets when the underlying market closes. The contagion model developed by King and Wadhwani (1990) implies that traders in one market draw information from observed price movements in another, hence price movements in one market affect price behaviour in other related markets. The model predicts a decline in futures market volatility when the underlying market closes. The strategic trading models developed by Admati and Pfleiderer (1988) and Foster and Viswanathan (1990) imply that informed traders seek to transact during periods when liquidity (uninformed) traders trade in order to minimise execution costs, such as when a market opens or closes. Consistent with these models, Chang et al. (1995) demonstrate that price volatility in S&P500 index futures drops immediately when the underlying market closes, and subsequently increases in the closing minutes of trading. The theories relied on by Chang et al. (1995) imply that a similar mini U-shaped pattern in price volatility is expected at the close, as well as the open of trading on the HKFE following the change in trading hours. However, no such pattern is expected prior to the change in trading hours. This provides a natural control for testing the impact of cash market closure on futures price volatility. Another plausible explanation exists for the pattern in stock index futures market volatility after the close of the underlying market. It is well documented that the levels of trading activity in equity derivatives and equities markets are positively related (e.g. Anthony, 1988). This implies that the level of futures trading activity is likely to increase in line with underlying market trading activity, as the close of the underlying market approaches (see Chan et al., 1995). However, it is also likely that the level of trading activity in stock index futures following the close of the equities market will decline, which can influence price volatility through bid–ask bounce. and As will be demonstrated in this paper, the probability of bid–ask bounce is likely to decrease following underlying market closure, thereby reducing the upward bias in transaction price based volatility measures used by Chang et al. (1995) (see Venkatesh, 1992). Smith and Whaley (1994) identify that US exchanges do not collect bid and ask quotes posted by traders. Unlike US futures exchanges, price reporters on the floor of the HKFE collect quote and trade frequency information. This provides another unique opportunity to extend the analysis in Chang et al. (1995). This study uses this data to test a bid–ask bounce explanation for the decline in price volatility following underlying market closure. In summary, this paper extends the work in Chang et al. (1995) by analysing patterns in intraday price volatility in a new market setting–Hong Kong. The new setting allows us to extend the work in the earlier paper in four significant ways. First, the new setting provides a recent natural controlled laboratory experiment which allows us to re-test the implications of the contagion hypothesis on trading in futures markets following the close of the underlying market. Second, the market setting also facilitates a test of the implications of the contagion hypothesis at the opening of the futures market, as the underlying market in the new setting opens after the futures market. Third, and most importantly, the availability of quote data allows us to test a new explanation for observed patterns in futures market volatility–bid–ask bounce. Finally, the market setting provides a rich data-set which permits an examination of a number of other variables including trade frequency and bid–ask spreads. The remainder of this paper is structured as follows. The next section describes the data and institutional detail. This is followed by Section 3, which re-examines price volatility patterns following market closure (and opening) for Hang Seng Index futures. Section 4 examines the extent to which bid–ask bounce explains observed patterns in intraday price volatility, while the final section concludes and provides some suggestions for future research.
نتیجه گیری انگلیسی
This paper corroborates the findings in Chang et al. Ž1995. and provides evidence of a decline in futures market volatility following the close of the cash market and increase at the close of trade using transaction price based measures of volatility for the HKFE. Evidence of an elevation in price volatility at the open of trade, and increase in price volatility following the opening of the cash market is also provided. Evidence is also provided that part of the observed behaviour in the transaction price based measures of volatility is caused by systematic changes in bid–ask bias associated with changes in trading activity. However, bid–ask bounce explains only part of the observed behaviour in price volatility. Consistent with Chang et al. Ž1995., it is concluded that the change in futures market volatility in response to spot market opening and closure is also influenced by AcontagionB effects. There are a number of possible future research directions. First, this study has demonstrated that patterns in trading frequency can lead to similar patterns in price volatility through the effects of bid–ask bounce, in the absence of any other fundamental changes in market behaviour. Hence, the methodology applied in this paper can be used to re-examine the intraday patterns in price volatility documented in previous research based on other markets. Such research could provide further insight into the causes of price volatility in securities markets. Second, the findings documented in this paper suggest that trading behaviour in futures markets when the underlying market is closed Že.g. during overnight trading. may be shaped by fundamentally different factors to when the underlying market is open. A useful future research direction is to examine the trading behaviour on overnight futures markets such as SYCOM on the Sydney Futures Exchange, GLOBEX on the Chicago Mercantile Exchange or Project A on the Chicago Board of Trade. Such research can provide insight into the desirability of closing underlying markets andror 24-hour trading.