基于遗传算法优化跟踪误差的股指期货套利分析

Arbitrage analysis of stock index futures based on the genetic algorithm for optimizing the tracking error

运用遗传算法对现货组合的跟踪误差进行优化,使得跟踪误差减小到0.01%以下,ETF现货组合对沪深300指数的跟踪效果较好,运用遗传算法求得的最小跟踪误差之下的现货组合权重来进行套利实证分析,并根据不完美市场条件下的期货定价模型求出无套利区间的上下界,准确地把握了最优套利机会,得到了较好的套利结果.

This thesis uses the genetic algorithm in the process of a ETFs portfolio construction experiment to opti- mize tracing errors ; if the error can be downgraded under 0. 01%, the tracking performance for the CSI300 Index will be much better. This thesis takes the weight in the portfolio that is constructed by the genetic algorithm which obtains the minimum tracing error for an empirical analysis of arbitrage. And then, it gives the upper and lower bounds of the non-arbitrage interval based on the cost of the ownership model in the imperfect market condition.Thus, it helps seize the optimal arbitrage opportunity which leads to a better arbitrage result.