Pricing and trading European options by combining artificial neural networks and parametric models with implied parameters

Abstract

Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. Pricing and Trading European Options by Combining Artificial Neural Networks and Parametric Models with Implied Parameters AbstractWe compare the ability of the parametric Black and Scholes, Corrado and Su models, and Artificial Neural Networks to price European call options on the S&P 500 using daily data for the period January 1998 to August 2001. We use several historical and implied parameter measures. Beyond the standard neural networks, in our analysis we include hybrid networks that incorporate information from the parametric models. Our results are significant and differ from previous literature. We show that the Black and Scholes based hybrid artificial neural network models outperform the standard neural networks and the parametric ones. We also investigate the economic significance of the best models using trading strategies (extended with the Chen and Johnson modified hedging approach). We find that there exist profitable opportunities even in the presence of transaction costs.2

Generalized parameter functions for option pricing

“…In addition, we require at least four data-points per maturity to ensure 21 that every maturity is satisfactorily represented. The final dataset has a total of 37202 observations and compares favourably with previous studies that test nonparametric methods (e.g., Hutchinson et al, 1994, Aït-Sahalia and Lo, 1988, Andreou et al, 2008. Sample characteristics for the dataset can be found in Table 1, where the average (contract-specific)…”

“…An artificial neural network is a collection of interconnected simple processing elements structured in successive layers and can be depicted as a network of links (synapses) and nodes (neurons) between layers (see also Andreou et al, 2008). A typical feedforward neural network has an input layer, one or more hidden layers and an output layer.…”

“…Hutchinson et al (1994) conclude that, although they do not constitute a substitute for the more traditional arbitrage-based pricing formulas, such methods may be more accurate and computationally more efficient alternatives than parametric models when the underlying asset's price dynamics are unknown, and when the pricing equation cannot be solved analytically. Evidence concerning the out-of-sample pricing performance of standard artificial neural networks is sometimes mixed (see discussions in Andreou et al, 2008), and their application for pricing options also has its limitations. First of all, standard neural networks are commonly interpreted as "black boxes"…”

Forecasting foreign exchange rates with adaptive neural networks using radial-basis functions and Particle Swarm Optimization

“…(2006) apply Higher Order Neural Networks in forecasting the AUD/USD exchange rate with a 90% accuracy. Panda and Narasimhan (2007) use a single hidden layer feedforward NN to produce statistical accurate forecasts of the INR/USD exchange rate having several linear autoregressive models as benchmarks while Andreou et. al.…”

Multiple kernel support vector regression for pricing nifty option

“…The reason being that all these methods are capable of predicting risky financial market as they can be used for nonlinear function approximation without any assumptions on the option pricing data [2][3][4][5][6][7]. Support Vector Machine (SVM) was proposed by Vapnik [8] based on the statistical learning theory.…”