Comparison of Chinese 50 ETF put option pricing based on four algorithms

Abstract

The article starts with the traditional Black-Scholes(B-S) option pricing models. Three more models: Long-term and short-term memory networks(LSTM), support vector machine (SVM) and random forest(RF) are introduced to be compared to the B-S model and to each other on 50 ETF put option pricing. It is showed that each model has its advantages when used in different position. The neural network pricing result is better than that of the B-S model From the four evaluation indicators of MD, MSD, MAD and MPD, the absolute values of the four errors of the prediction results of the neural network are all smaller than the absolute values of the corresponding errors of the prediction results of the B-S model.