Dto. Business Administration

EPSI de Gijón

University of Oviedo

(34) 985 18 21 47

david@uniovi.es alejandro.garrido@telefonica

ABSTRACT

The application of Artificial Intelligence described in this article is

intended to resolve the issue of speculation on the stock market.

Genetic Algorithms is the technique that is used, with the article

agomez@epsig.uniovi.es

Categories and Subject Descriptors

I.2.8 [Artificial Intelligence]: Problem Solving, Control Methods,

and Search – Heuristic methods.

J.4 [Social and behaviour sciences]: Economics.

General Terms: Algorithms, Economics.

Keywords: Stock exchange speculation, Genetic

Algorithms, Technical Analysis, Chartism.

1. INTRODUCTION

There are two trends or types of analysis used by stock market

traders. Exponents of Fundamental Analysis start from the

hypothesis that the market mirrors a company’s value based on its

growth potential. They therefore anchor their forecasts on analysis

of company accounts and trading figure projections. In this way,

they can deduce whether a company is overvalued or undervalued.

Technical Analysis is the second of the two trends. Its advocates do

not concern themselves with ‘fundamental’ values such as sales,

regulations or the working environment, but instead base their ideas

on the hypothesis that any factor that truly influences the market

will immediately show up in a company’s share price and its

negotiated volume. This technique therefore only studies indexes

(digital filters for share prices and negotiated volume) and the charts

that describe their movements. Chart analysis is a part of Technical

Analysis called Chartism.

The Stock Market has attracted considerable academic attention,

given the enormous sums of capital that are moved around it, and

the wide range of techniques applied to forecasting stock prices

range from those that border on the philosophic, such as the Golden

Proportions of Elliot Waves, to more elaborate techniques such as

Fuzzy Logic and Chaos Theory. All of them pursue the same end: to

find some structure in a seemingly random signal. One of the classic

techniques is the Tendency Line or minimum squares in

Fundamental Analysis. Another that is frequently used for the Stock

A forecast of flat movements would entail biding one’s time, as

constantly joining and leaving the market entails considerable

The Relative Strength Index (RSI): This index ranges

Moving Average Convergence Divergence (MACD); this

index tries to predict market tendency changes before they

The Stochastic Index attempts to forecast tendency changes,

Page 2

The evaluation function will have a chromosome as input and will

give us a higher value the more the chromosome fits our criteria. For

example, data that are not found will be penalised and will placed

last when the population is organised. In contrast, if we find a

chromosome that has occurred repeatedly, it should be scored

positively and the number of times the chromosome has occurred

should be evaluated and averaged [3]. It should be remembered that,

given the discrete nature of the indicators that were quantified and

the ongoing variation of the signals, neither of them will fit a single

Figure 1: Zone where the chromosome maintains its value.

Point P1 is the first point where the market conditions are met. From

this point an imaginary rectangle would be drawn defined by a

percentage gain %g and a percentage loss %p. These two values

define the upper and lower extremes of the rectangle and represent

the maximum and minimum gains and losses. Their value will

depend on the profitability sought and the risk that the investor is

prepared to accept.

The evaluation function will evaluate situations providing profit

positively, and situations leading to losses negatively. Another key

factor is the time required to acquire profits or losses. Shorter