Revisiting Agent-Based Models of Algorithmic Trading Strategies

Algorithmic trading (AT) strategies aim at executing large orders discretely, in order to minimize the order’s impact, whilst also hiding the traders’ intentions. The contribution of this paper is twofold. First we presented a method for identifying the most suitable market simulation type, based on the specific market model to be investigated. Then we proposed an extended model of the Bayesian execution strategy. We implemented and assessed this model using our tool AlTraSimBa (ALgorithmic TRAding SIMulation BAcktesting) against the standard Bayesian execution strategy and naïve execution strategies, for momentum, random and noise markets, as well as against historical data. Our results suggest that: (i) momentum market is the most suitable model for testing AT strategies, since it quickly fills the Limit Order book and produces results comparable to those of a liquid stock; (ii) the priors estimation method proposed in this paper \(-\) within the Bayesian adaptive agent model \(-\) can be advantageous in relatively stable markets, when trading patterns in consecutive days are strongly correlated, and (iii) there exists a trade-off between the frequency of decision making and more complex decision criteria, on one side, and the negative outcome of lost trading on the agents’ side due to them not participating actively in the market for some of the execution steps.