Catálogo de publicaciones - libros

Over the last decade, numerous papers have investigated the use of Genetic Programming (GP) for creating financial trading strategies. Typically, in the literature, the results are inconclusive but the investigators always suggest the possibility of further improvements, leaving the conclusion regarding the effectiveness of GP undecided. In this paper, we discuss a series of pretests aimed at giving more clear-cut answers as to whether GP can be effective with the training data at hand. Precisely, pretesting allows us to distinguish between a failure due to the market being efficient or due to GP being inefficient. The basic idea here is to compare GP with several variants of random searches and random trading behaviors having well-defined characteristics. In particular, if the outcomes of the pretests reveal no statistical evidence that GP possesses a predictive ability superior to a random search or a random trading behavior, then this suggests to us that there is no point in investing further resources in GP. The analysis is illustrated with GP-evolved strategies for nine markets exhibiting various trends.

An investor will in general designate a goal in the investment process. The traditional constant proportion portfolio insurance (CPPI) strategy considers only the floor constraint but not the goal aspect. This paper proposes a goal-directed (GD) strategy to express an investor’s goal-directed trading behavior and combines it with the portfolio insurance perspective to form a piecewise linear goal-directed CPPI (GDCPPI) strategy. This piecewise linear GDCPPI strategy shows that there is a wealth position at the intersection of the linear GD strategy and linear CPPI strategy. This position guides investors to apply the CPPI strategy or GD strategy depending on whether the current wealth is less or greater than, respectively. In addition, we extend the piecewise linear GDCPPI strategy to a piecewise nonlinear GDCPPI strategy with a minimum function. These piecewise GDCPPI strategies when applying the minimum function can fully maintain the features of the CPPI strategy and the GD strategy without considering the explicit. This minimum function in fact can obtain the concept of the explicit , but it operates the implicitly. Furthermore, we argue that the piecewise nonlinear GDCPPI strategy owns a larger solution space and it can then outperform the piecewise linear GDCPPI strategy in terms of the return rate performance measure. This paper performs some experiments using the Brownian, GA and forest GP techniques to prove with statistical significance that the piecewise nonlinear GDCPPI strategy can outperform the piecewise linear GDCPPI strategy and that there are some data-driven techniques that can find better piecewise linear GDCPPI strategies than strategies based on the Brownian technique.

This paper describes a hybrid-agent organization model from a logical-heuristic point of view. The organization model is an ordered set composed of an environment (or domain), the hybrid agents as elements of the environmental structure, as well as some connecting functions. A hybrid-agent in an organization structure is composed of a heuristic-decision support system (HDSS) and the decision-maker (the user) is one of the units. By using this hybrid-agent organization modeling, we can provide instruments that will help with the making of decisions in a financial organization such as a commercial bank.