Multi-scale capability: A better approach to performance measurement for algorithmic trading

Abstract

This paper develops a new performance measurement methodology for algorithmic trading. By adapting capability from the quality control literature, we present new criteria for assessing control, expected tail loss and risk-adjusted performance in a single framework. The multi-scale capability measure we present is more descriptive and more appropriate for algorithmic trading than the traditional measure used in finance. It is robust to non-normality and the multiple time horizon decision processes inherent in algorithmic trading. We also argue that an algorithmic trading strategy, indeed any investment strategy, which satisfies the criteria to be multi-scale capable also satisfies any definition of prudence. It will be unlikely to harm the investor or external market participants in the event of its failure, while providing a high likelihood of satisfactory risk-adjusted performance.

High-Frequency Trading and Conflict in the Financial Markets

“…They seek to compete by operating strategies that consistently make profits. Recent papers (Kumiega et al, 2014;Cooper et al, 2015) establish the criterion for HFT firms to remain viable and called this ''multi-scale capability.'' In essence, a trading firm that is multi-scale capable can ''almost always'' cover its average costs with average profits, including infrastructure costs, over an acceptable period.…”

“…Nevertheless, safety is a desirable goal. Cooper et al (2015) argue that operating in control is part of being prudent in HFT. In the trading context, the first steps toward control mean building expectations about the relevant control variables defining a strategy.…”

The Mysterious Ethics of High-Frequency Trading

“…Cooper, Ong, and Van Vliet (2015) have proposed a framework that can serve as a proxy for prudence in algorithmic trading. That framework borrows concepts from the literature of quality control.…”

“…Since “the greatest concerns about HFT are the risks it poses for both firms and the markets” (Boatright 2015), this framework is an explicit attempt to incorporate two principles of Nielsen (2010, 309)—avoiding or preventing harm to others, and leverage-proportionality and prudence (see also Ryan, Buchholtz, and Kolb 2010). Cooper, Ong, and Van Vliet (2015, 64) argue that this framework is a sufficient standard, one that, if applied by reasonable people or at least certified by a firm of reasonable people, must satisfy any reasonable definition of prudence. Furthermore, we argue now that a marketplace of prudent strategies is essential for an effective market.…”

“…10.Cooper, Ong, and Van Vliet (2015, 59) quantifies this measure of capability as in equation (1).Once control is established, if the distribution of average payoffs μ has a location and shape sufficient to cover its costs c (i.e. μ n – c > 0) over a sample period n with some specified level of confidence (i.e.…”