Stochastic volatility options pricing with wavelets and artificial neural networks

Abstract

The paper describes an alternative options pricing method which uses a binomial tree linked to an innovative stochastic volatility model. The volatility model is based on wavelets and artificial neural networks. Wavelets provide a convenient signal/noise decomposition of the volatility in the nonlinear feature space. Neural networks are used to infer future volatility from the wavelets feature space in an iterative manner. The bootstrap method provides the 95% confidence intervals for the options prices. Market options prices as quoted on the Chicago Board Options Exchange are used for performance comparison between the Black‐Scholes model and a new options pricing scheme. The proposed dynamic volatility model produces as good as and often better options prices than the conventional Black‐Scholes formulae.     

Predicting direction shifts on Canadian–US exchange rates with artificial neural networks

The paper presents a variety of neural network models applied to Canadian–US exchange rate data. Networks such as backpropagation, modular, radial basis functions, linear vector quantization, fuzzy ARTMAP, and genetic reinforcement learning are examined. The purpose is to compare the performance of these networks for predicting direction (sign change) shifts in daily returns. For this classification problem, the neural nets proved superior to the naïve model, and most of the neural nets were slightly superior to the logistic model. Using multiple previous days' returns as inputs to train and test the backpropagation and logistic models resulted in no increased classification accuracy. The models were not able to detect a systematic affect of previous days' returns up to fifteen days prior to the prediction day that would increase model performance. Copyright © 2001 John Wiley & Sons, Ltd.     

Anticipatory effects in the FTSE 100 index revisions

This paper examines the price impact of trading due to expected changes in the FTSE 100 index composition, which employs publicly-known objective criteria to determine membership. Hence, it provides a natural context to investigate anticipatory trading effects. We propose a panel-regression event study that backs out these anticipatory effects by looking at the price impact of the ex-ante probability of changing index membership status. Our findings reveal that anticipative trading explains about 40% and 23% of the cumulative abnormal returns of additions and deletions, respectively. The results are both statistically and economically significant.     

The Impact of Changes in the FTSE 100 Index

This paper investigates FTSE 100 index membership changes, which are determined quarterly by market capitalization and should have no information content. Return reversal around index additions and deletions suggests that buying (selling) pressure moves prices temporarily away from equilibrium, consistent with short‐term downward sloping demand curves. In contrast to widely reported results for the S&P 500, there is no evidence of permanent price effects. Further results suggest that investor awareness and monitoring due to index membership do not explain the price effects. There is statistically significant anticipatory trading in stocks that just fail to be promoted to the FTSE 100.     

Generative Bayesian neural network model for risk-neutral pricing of American index options

Financial models with stochastic volatility or jumps play a critical role as alternative option pricing models for the classical Black–Scholes model, which have the ability to fit different market volatility structures. Recently, machine learning models have elicited considerable attention from researchers because of their improved prediction accuracy in pricing financial derivatives. We propose a generative Bayesian learning model that incorporates a prior reflecting a risk-neutral pricing structure to provide fair prices for the deep ITM and the deep OTM options that are rarely traded. We conduct a comprehensive empirical study to compare classical financial option models with machine learning models in terms of model estimation and prediction using S&P 100 American put options from 2003 to 2012. Results indicate that machine learning models demonstrate better prediction performance than the classical financial option models. Especially, we observe that the generative Bayesian neural network model demonstrates the best overall prediction performance.     

Interpretable credit model development via artificial neural networks

Poor credit granting decisions are coming back to haunt providers of loan finance. Past poor credit granting decisions are in part due to the Equal Credit Opportunity Act (1975). This act requires lenders to explain the decision to grant or refuse credit. As a result, models such as artificial neural networks, which offer improved ability to identify poor credit risks but which do not offer easy explanations of why a loan applicant has scored badly, remain unused. This paper investigates whether these models can be interpreted so that explanations for credit application rejection can be provided. The results indicate that while the artificial neural networks can be used (with caution) to develop credit scoring models, the limitations imposed by the credit granting process make their use unlikely until interpretation techniques are developed that are more robust and that can interpret multiple hidden-layer artificial neural networks. Copyright © 2008 John Wiley & Sons, Ltd.     

Efficiency in the Pricing of the FTSE 100 Futures Contract

This paper studies the pricing efficiency in the FTSE 100 futures contract by linking the predictable movements in futures returns to the time-varying risk and risk premia associated with prespecified factors. The results indicate that the predictability of the FTSE 100 futures returns is consistent with a conditional multifactor model with time-varying moments. The dynamics of the factor risk premia, combined with the variation in the betas, capture most of the predictable variance of returns, leaving little variation to be explained in terms of market inefficiency. Hence the predictive power of the instruments does not justify a rejection of market efficiency.     

Enterprise resource planning enabling segmental information reporting practices of UK‐FTSE 100

The ERP impact on management accounting practices has been widely recognised as having a knowledge gap in relation to how it may facilitate Segmental Information Reporting following the International Financial Reporting Standard No. 8’s (IFRS‐8) management approach. This study contributes to filling this gap by investigating the joint effect of the ERP and IFRS‐8 Post‐Implementation Review (PIR) on dimensions: quality, quantity and the reporters’ identity of FTSE‐100 companies in the period 2013–2017. The study found that ERP is significantly and positively associated with the dimensions of segmental information reporting. The implications of this study extend research and the practices of segmental reporting on the importance of ERP in operationalising segmental reporting and in understanding variations.     

How trading activity scales with company size in the FTSE 100

This paper investigates the scaling dependencies between measures of ‘activity’ and of ‘size’ for companies included in the FTSE 100. The ‘size’ of companies is measured by the total market capitalization. The ‘activity’ is measured with several quantities related to trades (transaction value per trade, transaction value per hour, tick rate), to the order queue (total number of orders, total value), and to the price dynamic (spread, volatility). The outcome is that systematic scaling relations are observed: (1) the value exchanged by hour and value in the order queue have exponents of less than 1, respectively 0.90 and 0.75; (2) the tick rate and the value per transaction scale with the exponents 0.39 and 0.44; (3) the annualized volatility is independent of the size, and the tick-by-tick volatility decreases with the market capitalization with an exponent of ?0.23; (4) the spread increases with the volatility with an exponent of 0.94. A theoretical random walk argument is given that relates the volatility exponents to the exponents in points 1 and 2.     

GP algorithm versus hybrid and mixed neural networks

In the current paper, we present an integrated genetic programming (GP) environment called java GP modelling. The java GP modelling environment is an implementation of the steady-state GP algorithm. This algorithm evolves tree-based structures that represent models of inputs and outputs. The motivation of this paper is to compare the GP algorithm with neural network (NN) architectures when applied to the task of forecasting and trading the ASE 20 Greek Index (using autoregressive terms as inputs). This is done by benchmarking the forecasting performance of the GP algorithm and six different autoregressive moving average model (ARMA) NN combination designs representing a Hybrid, Mixed Higher Order Neural Network (HONN), a Hybrid, Mixed Recurrent Neural Network (RNN), a Hybrid, Mixed classic Multilayer Perceptron with some traditional techniques, either statistical such as a an ARMA or technical such as a moving average convergence/divergence model, and a naïve trading strategy. More specifically, the trading performance of all models is investigated in a forecast and trading simulation on ASE 20 time-series closing prices over the period 2001–2008, using the last one and a half years for out-of-sample testing. We use the ASE 20 daily series as many financial institutions are ready to trade at this level, and it is therefore possible to leave orders with a bank for business to be transacted on that basis. As it turns out, the GP model does remarkably well and outperforms all other models in a simple trading simulation exercise. This is also the case when more sophisticated trading strategies using confirmation filters and leverage are applied, as the GP model still produces better results and outperforms all other NN and traditional statistical models in terms of annualized return.     

Information costs and liquidity effects from changes in the FTSE 100 list

Abstract

In this paper we examine the stock price effect of changes in the composition of the FTSE 100 over the time period of 1984–2001. Like the S&P 500 listing studies, we find that the price and trading volume of newly listed firms increases. The evidence is consistent with the information cost/liquidity explanation. This is because investors hold stocks with more available information, implying that they have lower trading costs. This explains the increase in the stock price and trading volume of newly listed stocks to the FTSE 100 List. We find the reverse effect for the deletions from the FTSE 100.     

Analysing the determinants of narrative risk information in UK FTSE 100 annual reports

We investigate the relationship between the quantity of narrative risk information in corporate annual reports and ownership, governance, and US listing characteristics. We find that corporate risk reporting is negatively related to share ownership by long-term institutions, and thus the results of this study put forth that this important class of institutional investor has investment preferences for firms with a lower level of risk disclosure. Concerning governance, we find that different types of board director fulfil different functions, with both the number of executive and the number of independent directors positively related to the level of corporate risk reporting, but not the number of dependent non-executive directors. This supports a recent emphasis in the UK on the independent aspects of non-executive directors for good corporate governance. Separate investigation of business, financial, and internal control aspects of risk reporting that correspond to the three classes of risk-reporting guidance in the UK reveals that the pattern of risk information in the annual report may be dependent upon the form that reporting regulation takes.     

Competitive pricing strategies in social networks

We study pricing strategies of competing firms selling heterogeneous products to consumers. Goods are substitutes and there are network externalities between neighboring consumers. In equilibrium, firms price discriminate based on the network positions and charge lower prices to more central consumers. We also show that, under some conditions, firms' equilibrium profits decrease when either the network becomes denser or network effects increase. In contrast, consumers always benefit from being more connected to each other. We determine the optimal network structure and compare uniform pricing and discriminatory pricing from the perspectives of firms and consumers.     

Modelling spikes and pricing swing options in electricity markets

Most electricity markets exhibit high volatilities and occasional distinctive price spikes, which result in demand for derivative products which protect the holder against high prices. In this paper we examine a simple spot price model that is the exponential of the sum of an Ornstein–Uhlenbeck and an independent mean-reverting pure jump process. We derive the moment generating function as well as various approximations to the probability density function of the logarithm of the spot price process at maturity T. Hence we are able to calibrate the model to the observed forward curve and present semi-analytic formulae for premia of path-independent options as well as approximations to call and put options on forward contracts with and without a delivery period. In order to price path-dependent options with multiple exercise rights like swing contracts a grid method is utilized which in turn uses approximations to the conditional density of the spot process.     

Predicting sovereign debt crises using artificial neural networks: A comparative approach

Recent episodes of financial crisis have revived interest in developing models able to signal their occurrence in timely manner. The literature has developed both parametric and non-parametric models, the so-called Early Warning Systems, to predict these crises. Using data related to sovereign debt crises which occurred in developing countries from 1980 to 2004, this paper shows that further progress can be achieved by applying a less developed non-parametric method based on artificial neural networks (ANN). Thanks to the high flexibility of neural networks and their ability to approximate non-linear relationship, an ANN-based early warning system can, under certain conditions, outperform more consolidated methods.     

Deep neural network framework based on backward stochastic differential equations for pricing and hedging American options in high dimensions

We propose a deep neural network framework for computing prices and deltas of American options in high dimensions. The architecture of the framework is a sequence of neural networks, where each network learns the difference of the price functions between adjacent timesteps. We introduce the least squares residual of the associated backward stochastic differential equation as the loss function. Our proposed framework yields prices and deltas for the entire spacetime, not only at a given point (e.g. t?=?0). The computational cost of the proposed approach is quadratic in dimension, which addresses the curse of dimensionality issue that state-of-the-art approaches suffer. Our numerical simulations demonstrate these contributions, and show that the proposed neural network framework outperforms state-of-the-art approaches in high dimensions.     

Portfolio creation using artificial neural networks and classification probabilities: a Canadian study

Financial Markets and Portfolio Management - This study aims to verify whether using artificial neural networks (ANNs) to establish classification probabilities generates portfolios with higher...     

Feedforward versus recurrent neural networks for forecasting monthly japanese yen exchange rates

Neural networks are a relatively new computer artificial intelligence method which attempt to mimic the brain's problem solving process and can be used for predicting nonlinear economic time series. Neural networks are used to look for patterns in data, learn these patterns, and then classify new patterns and make forecasts. Feedforward neural networks pass the data forward from input to output, while recurrent networks have a feedback loop where data can be fed back into the input at some point before it is fed forward again for further processing and final output. Some have argued that since time series data may have autocorrelation or time dependence, the recurrent neural network models which take advantage of time dependence may be useful. Feedforward and recurrent neural networks are used for comparison in forecasting the Japanese yen/US dollar exchange rate. A traditional ARIMA model is used as a benchmark for comparison with the neural network models.Results for out of sample show that the feedforward model is relatively accurate in forecasting both price levels and price direction, despite being quite simple and easy to use. However, the recurrent network forecast performance was lower than that of the feedforward model. This may be because feed forward models must pass the data from back to forward as well as forward to back, and can sometimes become confused or unstable. Both the feedforward and recurrent models performed better than the ARIMA benchmark model.The author wish to thank the reviewers Drs. Kraft and Radford for their helpful comments.     

Order Imbalance in the FTSE Index Futures Market: Electronic versus Open Outcry Trading

Abstract:  This study examines trading activities before and after the transfer of the FTSE 100 index futures contract from open outcry to electronic trading. Daily order imbalance exhibits strong serial persistence in the electronic limit order market, but not in open-outcry trading. Both excess buying and selling reduce liquidity. In the electronic venue, prior market movements barely affect investors' buying or selling decisions. Excess buy orders do not generate any price impact, but sell orders do. Positive imbalances are more strongly autocorrelated than negative imbalances. No trading elements, such as order imbalance, volume, or open interest, are associated with volatility. Moreover, excess buying decreases volatility. Such evidence suggests that the development and growth of electronic trading has changed the dynamics of trading activities in many important ways.     

Stochastic covariance and dimension reduction in the pricing of basket options

This paper presents a tailor-made method for dimension reduction aimed at approximating the price of basket options in the context of stochastic volatility and stochastic correlation. The methodology is built on a modification to the Principal Component Stochastic Volatility (PCSV) model, a stochastic covariance model that accounts for most stylized facts in prices. The method to reduce dimension is first derived theoretically. Afterwards the results are applied to a multivariate lognormal context as a special case of the PCSV model. Finally empirical results for the application of the method to the general PCSV model are illustrated.