Evolving Trading Rule-Based Policies

Abstract

Abstract. Trading-rule representation is an important factor to consider when designing a quantitative trading system. This study implements a trading strategy as a rule-based policy. The result is an intuitive human-readable format which allows for seamless integration of domain knowledge. The components of a policy are specified and represented as a set of rewrite rules in a context-free grammar. These rewrite rules define how the components can be legally assembled. Thus, strategies derived from the grammar are well-formed, domain-specific, solutions. A grammar-based Evolutionary Algorithm, Grammatical Evolution (GE), is then employed to automatically evolve intra-day trading strategies for the U.S. Stock Market. The GE methodology managed to discover profitable rules with realistic transaction costs included. The paper concludes with a number of suggestions for future work.

Co-evolving online high-frequency trading strategies using grammatical evolution

“…This compares to the performance of the zero-intelligence strategy used in [9]. From the equity curves, the hypothesis that the GE managed to find profitable trading strategies in the provided search space holds.…”

“…We include "domain" knowledge in the form of our if-statements, i.e. we know we need a ternary decision for the entry strategy (buy, sell or remain idle) and for the exit strategy (exitlong, exitshort and remain idle) [9]. The parameters for the grammatical evolution were set according to Table I.…”

“…In [9] and [10], the entry-and exit strategies were coevolved using GE. In [9], the fitness function incorporated risk and trading costs into its calculations, where the standard deviation of the total return was used as a proxy for risk.…”

“…In [9], the fitness function incorporated risk and trading costs into its calculations, where the standard deviation of the total return was used as a proxy for risk. In [10], four different fitness functions were benchmarked against each other, although trading costs were not included in the calculations.…”

Computing Trading Strategies Based on Financial Sentiment Data Using Evolutionary Optimization

“…i , [4], and [7]. Generating automatic trading rules has been a core topic in this domain, see especially [8], [5], [6], [13], and the references therein. One main technique in the field of meta-heuristics and technical trading is to let the optimizer generate optimal investment rules given a set of technical indicators.…”