RDP 8804: Pricing Behaviour in Australian Financial Futures Markets 4. Review of Empirical Evidence

Empirical evidence on the relationship between financial futures and spot markets has focussed on two main questions: first, are the futures markets efficient in the sense of fully reflecting information about expected future spot prices; and secondly, does futures trading tend to stabilise or to destabilise spot prices.

4.1 Efficiency

On the question of efficiency, interest rate futures markets in the United States have been tested fairly extensively for evidence of arbitrage profits or pricing biases, based on deviations between actual futures prices and the expected spot prices implied in the term structure. These are really tests of the joint efficiency of spot and futures markets. Results have been somewhat ambiguous, with studies on ten year bond futures (for example, Kolb, Gay and Jordan (1982) and Resnick and Henniger (1983)) generally offering no evidence against the efficient markets hypothesis, while comparable studies on ninety-day futures (Rendleman and Carabini (1979), Elton, Gruber and Rentaler (1984)) have found departures from jointly efficient pricing. These departures are however, claimed to be small compared with spot transaction costs.

As might be expected, much larger departures from joint efficiency have been found with respect to the share market, where transactions costs are higher. Cornell and French (1983) for example found the premium on Chicago share index futures could not be explained by an efficient arbitrage model, and a similar finding for Sydney SPI futures is reported by Bowers and Twite (1985). These results are consistent with the fact that arbitrage between spot and futures is much more costly in share indices than in interest rate contracts, reflecting the cost of assembling and trading a portfolio which is reasonably representative of the index. It should be noted however that the interpretation of these findings is not clear cut: a rejection of joint efficiency of the spot and futures markets, still leaves open the question as to where any inefficiency is located.

The main empirical evidence on the Australian interest rate futures markets is in papers by Sharpe (1984), Kearney, MacDonald and Hillier (1987) and Juttner, Tuckwell and Luedecke (1985). These studies use monthly data to perform standard tests for biases in bank bill futures prices as predictors of future spot prices. All three studies offer support for the hypothesis of no bias, implying at least a weak form of efficiency, although in some cases stronger forms of efficiency are rejected. A major criticism of these studies however is their use of data sets containing a high proportion of quoted prices at which no trades were actually made. This is a consequence of using monthly data points on contracts maturing three months ahead, when in practice only end-quarter maturing contracts have significant trading volumes. These data problems must cast doubt on the robustness of the reported results.

4.2 Volatility

The second main empirical question on futures markets concerns the stabilising or destabilising consequences of futures trading. Here the usual approach has been to test for significant changes in the variability of spot prices after the commencement of trading in a new futures contract, or to investigate correlations between spot price volatility and activity on futures markets. Early work on commodity futures showed, if anything, that futures trading tended to reduce variability of spot prices; much of this work is summarised by Power (1970). More recently, similar work has been carried out in financial futures markets (see for example Froewiss (1978), Simpson and Ireland (1982, 1985) and Rutledge (1986)); little or no relationship between futures trading and volatility has been detected, even with data sampled as frequently as daily. On a related issue, Kawaller, Koch and Koch (1987) investigated the lead-lag relationship between Chicago share index futures prices and the underlying index values. Their results suggested that the futures market slightly leads the spot on intra-day price movements. There appears to be no comparable Australian evidence on this point.