High frequency information based portfolio decision for downside risk-return tradeoff using differential evolution algorithm

A performance measure of intraday investment is constructed using high frequency trading data, which takes the trade-off between return and downside risk into account. The principle of the measure consists in partitioning the return on each intraday interval into loss and gain according to a user-specified target. Both the excess return and profit-making opportunity above the target constitute the return term of the new measure, while the intraday price movement below the target is assumed as the proxy of risk. The paper proceeds to demonstrate that the new measure has implication in portfolio decision where the weights could be estimated using the differential evolution algorithm and the results on a market test are promising. 1548-7741/