Price estimation of a warrant using polynomial artificial neural networks

Titulos Opcionales (TO's) is the name that the Mexican Stock Market authorities have given to the financial instruments that have been internationally called Warrants. The model proposed by Black & Scholes is one of the most widely used models for valuing these financial instruments. The design of a model of Polynomial Artificial Neural Network (PANN) for estimating the closing price of a TO is shown in this paper. In order to demonstrate the potential estimate of the PANN, several regression and variance analyses are carried out. Firstly, the closing prices and the estimate calculated by the Black & Scholes model are compared, and afterwards the closing prices and the estimate calculated by PANN are compared.