The impact of the introduction of futures contracts on the spot market volatility: the case of Kuala Lumpur Stock Exchange

Abstract

In investigating the impact of futures trading on spot market volatility, it is not obvious to what extent the results obtained using data from well developed and highly liquid markets are applicable to emerging markets. This paper provides evidence on the impact of the introduction of futures trading on spot market volatility using data from both the underlying and non-underlying stocks in the emerging Malaysian stock market. Results show that the onset of futures trading increases spot market volatility and the flow of information to the spot market. It is found that the underlying stocks respond more to recent news, while the non-underlying stocks respond more to old news. The lead–lag and causal relationship between futures trading activity and spot market volatility is also examined. VAR results show that the impact of the previous day's futures trading activity on volatility is positive but short (only a day). This is further confirmed by Granger's causality test.

Notes

¹ See Edwards (1988a, 1988b), Harris (1989), Antoniou and Holmes (1995), Kan (1997), Lee and Tong (1998), Dennis and Sim (1999).

² See Figlewski (1981) who offers reasons for the stabilizing as well as destabilizing effects of futures trading on the spot market volatility. Danthine (1978) also offers reasons for reduction in the spot market volatility due to introduction of futures trading.

³ See Gulen and Mayhew (2000) who examine the time series properties of stock indexes in 25 countries including New Zealand, Sweden, Finland, Denmark, Hungary, Israel and Venezuela and Chile.

⁴ See Chong et al. (1999), De Santis and Imrohoroglu (1997) and Pan et al. (1999).

⁵ Harris (1989) and Galloway and Miller (1997) use this approach. However, their investigations are based on data from the US market.

⁶ Gulen and Mayhew's (2000) study includes data until 30 December 1997, which covers the first half of the crisis. However, they ignore the effect of the Asian Financial Crisis.

⁷ Until June 2001, Malaysia had two futures exchanges: the Kuala Lumpur Options and Financial Futures Exchange (KLOFFE) and the Commodity and Monetary Exchange of Malaysia (COMMEX Malaysia). KLOFFE provides a market for only futures and options contracts based on Kuala Lumpur Stock Exchange's Composite Index (KLSE CI), while COMMEX Malaysia provides financial futures contract based on 3 month Kuala Lumpur Interbank Offer Rate (KLIBOR) together with commodity futures contract, the Crude Palm Oil (CPO) futures contract. On 11 June 2001, both exchanges were merged and renamed the Malaysia Derivatives Exchange Berhad (MDEX).

⁸ Extracted from the speech given by the former Chairman of Securities Commission of Malaysia, Dr Mohd. Munir Abdul Majid on 28 February 1997.

⁹ Extracted from the Bank Negara Annual Report 1999.

¹⁰ KLSE's Annual Report 2000, p. 8.

¹¹ See Edwards (1988a; b), Baldauf and Santoni (1991), Antoniou and Foster (1992), Lee and Ohk (1992), Antoniou and Holmes (1995), Galloway and Miller (1997), Pericli and Koutmos (1997), Dennis and Sim (1999), Hernandez-Trillo (1999), Gulen and Mayhew (2000) and Rahman (2001).

¹² Bollerslev et al. (1992) review the empirical evidence on the ARCH modelling in finance and suggest that most financial series follow a GARCH(1,1) process.

¹³ Antoniou and Foster (1992) and Antoniou and Holmes (1995) use such a proxy variable (for which there is no futures trading) to filter off the general market volatility effect in addition to the dummy variable that captures the impact of introduction of futures.

¹⁴ We agree with the rationale suggested by Garcia et al. (1986) and use V _t /OI _t rather than V _t as a proxy for futures trading activity because V _t /OI _t reflect speculative activity and the problem of daily volume which may only represent trading activity and/or time to expiration can be addressed through this definition.

¹⁵ Pre-futures period, 25/10/1993–14/12/1995; Post-futures before crisis period, 15/12/1995–31/07/1997; Post-futures crisis period, 1/08/1997–14/09/1998; Post-futures after capital controls, 15/09/1998–31/07/2001.

¹⁶ See Bollerslev et al. (1992).

¹⁷ This part of the result differs from those of Antoniou and Holmes (1995) and Antoniou and Foster (1992) where their findings showed that the pre-futures sample was integrated (or permanent) and the post-futures sample was stationary.