Research On Stock Index Futures Arbitrage With Controllable Risk In China

As China’s stock index future market is gradually mature since its launch over3yearsago, No-Risk Arbitrage opportunities are rather few and the profit is fairly low. In this context,this paper proposes a new trading strategy called Index-Futures Arbitrage with ControllableRisks and rather completely establish a quantitative model for stock index futures arbitrage.An empirical analysis is done strictly to test the feasibility of this new model, which provedthat this new model can provide arbitrageurs with a preliminary theoretical framework toapply in the mature market.The paper first analyzes index-futures arbitrage using the data of one-minute spot monthprices and prices of the underlying index which contains32future contracts since the marketwas launched. The result shows that in2010there are numerous opportunities for No-RiskArbitrage and the profit is huge. However, in the next two years, with dramatically feweropportunities and rather lower profits, No-Risk Arbitrage is not a good choice for investment.In this case, the framework of Index-Futures Arbitrage with Controllable Risks isproposed. The first key point is that arbitrageurs are allowed to enter a position at the pointthat they may suffer losses. Arbitrageurs should optimize strategies for closing a position toachieve the maximum Sharpe Ratio. Secondly, for different entering points, the maximumtrading losses differ. But the losses are controllable, available to be calculated before trading.Each entering point corresponds with a unique maximum Sharpe Ratio and a uniquemaximum trading loss. These unique ratio and loss form a trading strategy of the new model.Finally, by combing all trading strategies, portfolio strategies and efficient frontier areobtained，which provide evidence for arbitrageurs with different risk preference to invest.After that, the mathematical foundation of the new model is presented based on financialtime series theories. To be specific, white noise series, unit root series, ARMA models, ARCHmodels and Stochastic Volatility models are respectively employed to mathematically analyzeand calculate the probability of trading signals, maximum trading loss and Sharpe Ratio andefficient frontier under the given entry strategies.Finally, by using the historical data, we find that the return rate in this new model issignificantly higher than risk-free arbitrage’s even though the maximum trading loss is rather low. Consequently, the new model is quite valuable and gives arbitrageurs a new choice in themature market.