1、 外文翻译 原文 The impact of stock index futures on the Korean stock market Material Source: http:/ Harris, 1989; Antoniou and Holmes, 1995). In particular, one of the primary concerns of previous studies has been the issue of whether futures trading destabilizes the underlying spot market. Although some
2、studies find increased volatility, the weight of the empirical evidence shows no increase in volatility following the introduction of trading in stock index futures. Among others, Freris (1990) examined the impact effect of Hang Seng Index Futures on the behaviors of the Hang Seng Index using data f
3、or the period from 1984 to 1987 and found that the introduction of stock index futures trading had no measurable effect on the volatility of the stock price index. Lee and Ohk (1992) examined the effect of introducing index futures trading on stock market volatility in Australia, Hong Kong, Japan, t
4、he UK and the USA using daily index data for periods of approximately four years spanning the start of trade in index futures. They found that for the three largest markets, return volatility increased significantly after the listing of stock index futures. However, for the Australian market there w
5、as nosignificant difference, and for Hong Kong stock return volatility actually decreased. Using international portfolios, they further found that although the creation of stock index futures generally exerted a volatility-increasing influence on the behaviors of cash market stock returns, it made t
6、he stock market relatively more efficient because volatility shocks were more quickly assimilated in that market. Kamara et al. (1992) investigated the effect of futures trading on the S that is, the futures market reflects new information before the spot market. If new market information disseminat
7、es in the futures market before the stock market, then the introduction of a futures market increases the amount of information reflected in the spot price. This might be explained by the fact that trading futures has the advantages of a highly liquid market, low transaction costs, easily available
8、short positions, low margins and rapid execution. Thus, informed traders may find they can act faster and at lower cost in the futures market than they can in the cash market, resulting in a lead-lag relationship between futures and spot prices. The lead-lag relation between movements of spot and fu
9、tures prices has been widely investigated with the methods used varying across studies. For example, Kawaller et al. (1987), Abhyankar (1998) and Tang et al. (1992) use modified/non-modified Granger causality tests. Whereas Wahab and Lashgari (1993), Fleming et al.(1996) and Pizzi et al. (1998) use
10、cointegration and error correction models. However, irrespective of methodology, the results can be summarized concisely: market information tends to disseminate in futures prices prior to, and at greater speed than, in stock prices. All of the research so far discussed focuses on developed markets.
11、 Very little work has investigated the impact of stock index futures trading in emerging markets. This article contributes to this sparse literature; it is the first to examine the impact on the Korean spot market of trading in futures. It focuses on three aspects. First, the impact of futures tradi
12、ng on price volatility in the spot market is examined. Second, it discusses long-run equilibrium and short-run adjustment through tests of cointegration and causality. Third, lead-lag relationships are analyzed. This article differs from previous studies that use closing prices for futures and spot
13、prices, by using data with matched closing times. This is desirable because in Korea trading in index futures and trading in stocks finish at different times. By using matched closing times, one avoids comparing non-synchronous closing prices of the spot index and futures contracts, which might lead
14、 to a significant source of error. The impact of futures trading on price volatility in the underlying spot market index is examined by adopting the generalized ARCH (GARCH) process in which the conditional variance of u at time t is dependent not only on past squared disturbances but also on past c
15、onditional variances. Empirical evidence, for example, Bollerslev et al. (1992), Huang et al. (1995) and Ryoo (2001), finds that returns in stock markets exhibit heteroscedasticity. Therefore, following Holmes (1996), a GARCH representation would seem to be an appropriate means by which to capture m
16、arket-wide price volatility. Consider the model: （ 1） （ 2） where Rs,t is spot price returns, the change in the logarithm of the spot price index in period t, Rp,t is returns on the market proxy variable (for which there is no associated futures index), ut is an error term representing unexplained pr
17、ice changes andt is the information set available at time t. Since the proxy variable covers market-wide influences on price changes, the error term captures the impact of factors specific to the market on which the futures contract is written and its 1| ( 0 , )t t tu N h , 0 1 ,s t p t tR a a R u v
18、ariance, ht, provides a measure of market-wide price volatility: （ 3） By estimating the model for periods pre- and post-futures trading and comparing the parameters of the variance equations, it is possible to determine how futures trading impacts on volatility. The coefficient 1 relates todays pric
19、e changes to yesterdays market-specific price changes and as these depend on the arrival of information yesterday, 1can be viewed as a news coefficient. If increases1 following the introduction of futures trading, this suggests that futures trading results in information being impounded into spot pr
20、ices more rapidly. Conversely, if l decreases in the post-futures period, this suggests that information is impounded into spot prices more slowly. Engle and Bollerslev (1986) and Engle and Mustafa (1992) showed for the GARCH (1, 1) model, that the persistence of volatility shocks depends primarily
21、on1+1. An increase (decrease) in 1+1 following the introduction of futures trading indicates increased (decreased) persistence of volatility shocks. The Korea Stock Price Index 200 (KOSPI 200) is the underlying stock index for traded futures and options contracts on the KSE. The KOSPI 200 is a capit
22、alization-weighted index that tracks the continuous price performance of 200 actively traded, large capitalization common stocks listed on the KSE. These shares account for approximately 7080% of domestic market capitalization so the index reflects overall market performance. In order to avoid unint
23、ended bias, the constituent stocks are rigorously revised over time. The base figure was set at 100.00 as of 3 January 1990. Trading of KOSPI 200 futures is implemented under an order-driven, continuous trading system. Since trading is executed through a computerized system, there is no physical tra
24、ding floor. KOSPI 200 futures expire four times a year, in March, June, September and December. The last trading day is the second Thursday of each contract month. One index point equals 500 000 Korean won and settlement of the contract is in cash. There are two trading sessions, morning and afterno
25、on. Until 5 December 1998, both stock and stock index futures contracts were traded on weekdays between 9:30 a.m. and 11:30a.m. in the morning session. In the afternoon session, stocks were traded from 1:00 p.m. until 3:00 p.m. and index futures were traded between 1:00 p.m. and 3:15 p.m. exception
26、the last trading day of each contract month, when 20 1 1 1 1t t th u h futures trading closed at 2:50 p.m. On Saturdays, both stock and index futures were traded from 9:30 a.m. until 11:30 a.m. and 11:45 a.m., respectively. Since 7 December 1998, there has been no Saturday trading and the morning se
27、ssion for weekdays has been extended from 9:00 a.m. until mid-day. With the introduction of stock index futures trading, there were daily price limits for futures contracts of 5% of the previous trading days closing price. On 2 March 1998 this was relaxed to 7% and on 7 December 1998 to 10%. There i
28、s also a system of circuit breakers. When the price of the previous trading days most active contract reaches 5% of that days closing price for one minute, the trading of all futures contracts is halted for the next five minutes. Also, when the KOSPI continues (for one minute) to lose 10% or more of
29、 its value compared to the previous days closing price, futures trading is halted for 20 minutes. 译文 股指期货对韩国股票市场的影响 资料来源 : http:/ Antoniou and Holmes, 1995） 。特别是，以往研究的首要问题之一是是否期货交易对现货市场起到了本质上的影响。尽管一些研究发现买卖股票指数期货增加了现货是市场的波动性，但是大多是经验证据表明股指期货推出对波动性没有影响。其中， Freris（ 1990） 对 1984 年至 1987 年之间的恒生指数进行研究，他发现恒
30、生指数期货的引进对恒生价格指数的波动没有起到明显的作用。 Lee 和 Ohk（ 1992） 利用跨越了四年的 每日股指期货数据，对在澳洲、香港、日本、 英国和美国引入股指期货后股票市场波动性影响进行研 究。他们发现对第三大市场来说，在股指期货推出之后，波动报酬率有了明显的增长。在 澳大利亚市场上，波动报酬率没有显着差异，香港股市场却有着明显的 下降。利用国际投资组合，他们进一步发现，虽然股指期货的产生对现货市 场 的股票收益率波动性的影响越来越大，这是因为波动变得越来越灵敏从而 使得 股市也随之变得更加有效。 Kamara （ 1992） 研究了以 S&P 500 指数为标的物的股指期货的稳定
31、性影响。他的研究结果表明，虽然每月回报的波幅维持不变，但是在 股指期货买卖后期的波动收益率明显高于股指期货买卖前期。 股指期货在期货市场上 最重要的作用是价格发现，也就 是 说，股指期货对新信息的反应比现货市场更加超前。如果新的市场信息传播在期货市场之前，那么期货市场引入增加信息量将在现货价格中反映出来。这可能解释为交易期货具有流动性高，交易成本低，低保证金和快速交易的市场优势。因此，知情交易者可能会发现，他们可以比现货市场上更快做出反应和以更低的成本在期货市场交易，这是由于期货和现货价格之间的超前滞后关系造成的。 各国学者们运用不同的方法，对现货价格和期货价格直接的超前滞后关系进行的调查研究
32、。如 Kawalle （ 1987 年） 和 Abhyankar（ 1998 年） 使用修正或者未修的 Granger 因果关系进行检验 。 Wahab 和 Lashgari（ 1993） ， Fleming （ 1996） 还有 Pizzi（ 1998） 等都使用协整和误差修正模型进行研究。然而，不论方法结果可以简要概括 ： 市场信息在期货市场上的传播往往早于在现货市场 。 到目前为止，所有关于股指期货的研究都集中于发达的市场，只有很少一部分关于股指期货交易在新兴市场的影响研究。这篇文章对新兴市场关于这方面的研究做出了贡献，因为它是第一个研究韩国股指期货买卖对现货市场的影响论文。它主要集中在
33、三个方面的研究 ：第一，期货交易对现货市场价格波动的影响的研究。其次，通过协整和因果关系检验讨论长期均衡和短期调整的研究。第三，超前滞后关系的分析。本文不同于过去使用期货收盘价格与现货价格数据，而是使用匹配的关门时间价格数据。这种研究方法是可取的，因为在韩国的股指期货和股票买卖交易在不同的时间完成。 通 过使用匹配的关门时间，它避免了因非同步收市而可能引起重要错误的 现货指数和期货合约之间的比较。 关于价格的相关现货市场交易期货指数的波动影响审查采用广义的 ARCH（ GARCH 模型 ） ， 在条件方差 u 不仅是过去时间 t 平方的干扰，也收到 过去的条件方差的影响。例如 ， Boller
34、slev（ 1992 年）等认为在股票市场收益中存在异方差。因此，继 Holmes（ 1996）之后， GARCH 模型一个获取整个市场价格波动性的 典型模型。模型如下： （ 1） （ 2） 当 RS,T 表示现货价格收益，现货价格指数对数的变化 , Rp,t 表示市场上收益的自变量 （ 与股指期货指数无关 ）， ut 表示代表价格变动的误差项， t 表示在时间段内，由于包括影响整市场价格变动的自变量，误差项是指字期货合和他的方差。 ht 表示整个市场价格的波动性的衡量因素： （ 3） 通过估计期货交易的期间前，后的模型和比较方差方程中的参 数，它有可能确定在期货交易如何影响波动性。 与今天价
35、格相关的参数 l 转变为昨天市场价格的变动，这些都参数都依赖于昨天的信息，它可以被看成是“信息”参数。如果按照期货交易引进的增加了系数 l就这表明期货买卖对信息的反应比现货价格的反应更加快速。相反地，如果系数 l 在股指期货买卖后期减少，这就意味着股指期货对信息的反应慢于现货市场价格的反应。 Engle 和 Bollerslev （ 1986） ， Engle 和 Mustafa（ 1992） 演示了 GARCH （ 1, 1） 模型，他们揭示了持久的波动取决于参数 1+1。当参数随 着股指期货交易的增加（降低 ） 而同向变动时，就会产生持续的波动。 韩国股票价格指数 200（ KOSPI20
36、0 指数）是在韩国证券交易所买卖期货及期权合约的股票指数。韩国综合指数 200 是指在韩交所上市的 200 个活跃普通股的市值加权指数。这些股票在国内资本市场上占 70%-80%，因此这个指数能够反映整体市场表现。为了避免偏差，这些成分股在一段时间内需要调整。KOSPI 200 指数的以 1990 年 1 月 3 日定为基数。 韩国 KOSPI 200 指数期货交易是一个持续的交易系统。由于它的交易是通过电脑系统完成，所以它们不存在实际的交易大厅。韩国 KOSPI200 股指期货到期一年有四次。分别在三月、六月、 九月及十二月。最后交易日为每个合约, 0 1 ,s t p t tR a a R
37、 u 1| ( 0 , )t t tu N h 20 1 1 1 1t t th u h 月份的第二个星期四。一个指数点等于 50 万韩元的合同并且以现金结算。 韩国的股指期货有两个交易时段，上午和下午。直到 1998 年 12 月 5 日，股票和股票指数期货合约交易都在每个工作日的上午 9:30 和上午 11： 30。 买卖股票从 13:00 至下午 3:00 而指数期货则在下午 1:00 及下午 3:15 之间，每个合约月份的最后交易日期货的交 易在下午 2:50 结束。在星期六，股票和指数期货的交易分别是从 9:30 至 11:30， 9:30 到 11 时 45 分。自 1998 年 12 月 7 日起，交易时间已经转变成平日 9:00 至中午和周六停止交易。 随着股指期货交易的推出，每日的价格限制由前一交易日期货合约收盘价的 5%， 到 1998 年 3 月 2 日放宽至 7 %， 到 1998 年 12 月 7 日则上升到了 10%。 股指期货具有熔断机制，当前一交易日最活跃的合约价格一分钟达到当天的收盘价 5%时候，所有的期货合约将停止交易 5 分钟。此外，当 KOSPI 指数继续（约一分钟）失去 10%或相对 于前一天的收盘价其价值时，期货交易停止工作 20 分钟。
