A Novel Window Analysis and Its Application to Evaluating High-Frequency Trading Strategies

In order to examine the efficiency of decision-making units (DMUs) over time, the Window Data Envelopment Analysis (WDEA) is frequently applied. Since the WDEA considers a DMU at a period to be a distinct DMU, this method has several limitations, including a high computational cost and a lack of knowledge regarding the consistency of a DMU during a window. This study proposes a novel window analysis in which the information of a DMU during the window is linked, utilizing the new notions of “linked DMU” and “linked variable”. Consequently, an inconsistent DMU in a window would not be eligible for a high efficiency score, despite the fact that it might be the most efficient at certain times. To approximate the globally optimal result, the Whale Optimization Algorithm, one of the state-of-the-art meta-heuristics, is used. This ensures that the optimal solution is not trapped in local extremes, as is the case with linear programming methods. Furthermore, the proposed method is applied to evaluating the effectiveness of foreign exchange investment strategies as well as the effectiveness of companies in the utility industry, listed in the Ho Chi Minh City Stock Exchange. To our knowledge, this is the first time a WDEA-based method has been utilized in those fields. The results show that the new window analysis can identify effective and stable trading strategies/companies over time.