Negative selection—a new performance measure for automated order execution

Abstract Automated Order Execution is the dominant way of trading at stock markets. Performance numerous execution algorithms measured through slippage from some benchmark. But measuring true in algorithmic a difficult task since as well benchmarks are function market activity. In this paper, we propose new performance measure for algorithms. The measure, named Negative Selection, takes posterior look window and allows us to determine what would have been optimal order placement if knew advance, before actual trading, complete information during window. We define difference between hypothetical position execution. This calculated taking into account all prices traded quantities within considered time Thus, capturing impact caused by our own cost that affects trades. Properties which make it defined objective discussed. Some empirical results on real trade data presented.