Testing standard technical analysis parameters' efficiency, a metaheuristic approach

Abstract

The thesis proposed is aimed to exploit the profitability of Technical Analysis (TA) identifying among the wide set of indicators a sub-set of them, widely used both in literature and in practise, to optimize with metaheuristic algorithms and to test using a trading system. Since its introduction, there has been a wide discussion about the profitability of TA and mostly of the time it has been compared with the fundamental and Buy-And-Hold strategy. To achieve superior performance a trading system should be run using TA indicators that promptly signal when it is time to buy and sell. However, to find such TA indicators' values it requires the solution of a complex problem that can be expressed as an optimization one.The thesis proposed wants to compare three different metaheuristic algorithms, Particle Swarm Optimization (PSO), Differential Evolution (DE) and Fireworks (FWA), searching among them which one perform better.