Modeling Financial Data Research Articles

The Multivariate SupOU Stochastic Volatility Model

Using positive semidefinite supOU (superposition of Ornstein–Uhlenbeck type) processes to describe the volatility, we introduce a multivariate stochastic volatility model for financial data which is capable of modeling long range dependence effects. The finiteness of moments and the second‐order structure of the volatility, the log‐ returns, as well as their “squares” are discussed in detail. Moreover, we give several examples in which long memory effects occur and study how the model as well as the simple Ornstein–Uhlenbeck type stochastic volatility model behave under linear transformations. In particular, the models are shown to be preserved under invertible linear transformations. Finally, we discuss how (sup)OU stochastic volatility models can be combined with a factor modeling approach.

Coupled continuous-time random walk approach to the Rachev–Rüschendorf model for financial data

In this paper we expand the Rachev–Rüschendorf asset-pricing model introducing a coupled continuous-time-random-walk-(CTRW)-like form of the random number of price changes. Such a form results from the concept of the random clustering procedure (that resembles the coarse-graining methods of statistical physics) and, on the other hand, indicates applicability of the CTRW idea, widely used in physics to model anomalous diffusion, for describing financial markets. In the framework of the proposed model we derive the limiting distributions of log-returns and the corresponding pricing formulas for European call option. In order to illustrate the obtained theoretical results we present their fitting with several sets of financial data.

A vector integer-valued moving average model for high frequency financial count data

A vector integer-valued moving average (VINMA) model is introduced. The CLS and FGLS estimators are discussed. Empirically, it is found that the spillover effect from Ericsson B to AstraZeneca is larger than that from AstraZeneca to Ericsson B.

Econometric Analysis of Discrete-Valued Irregularly-Spaced Financial Transactions Data Using a New Autoregressive Conditional Multinomial Model

This paper proposes a new approach to modeling financial transactions data. A new model for discrete valued time series is proposed in the context of generalized linear models. Since the model is specified conditional on both the previous state, as well as the historic distribution, we call the model the Autoregressive Conditional Multinomial (ACM) model. When the data are viewed as a marked point process, the ACD model proposed in Engle and Russell (1998) allows for joint modeling of the price transition probabilities and the arrival times of the transactions. In this marked point process context, the transition probabilities vary continuously through time and are therefore duration dependent. Finally, variations of the model allow for volume and spreads to impact the conditional distribution of price changes. Impulse response studies show the long run price impact of a transaction can be very sensitive to volume but is less sensitive to the spread and transaction rate.

A simple non-linear model with fractional integration for financial time series data

This paper provides several examples of simple non-linear time series models with fractionally integrated disturbances. Both types of models (non-linear and fractional integration) have been widely used in recent years when modeling financial data. We use a testing procedure that permits us to test the order of integration in raw time series in the context of non-linear models. The tests are applied to several financial time series, the results showing that when the non-linear sign structure is taken into account, the order of integration of the series is much higher than one, finding thus conclusive evidence against mean reversion in their behavior.

Forecasting interest rates: a comparative assessment of some second-generation nonlinear models

Modeling and forecasting of interest rates has traditionally proceeded in the framework of linear stationary methods such as ARMA and VAR, but only with moderate success. We examine here three methods, which account for several specific features of the real world asset prices such as nonstationarity and nonlinearity. Our three candidate methods are based, respectively, on a combined wavelet artificial neural network (WANN) analysis, a mixed spectrum (MS) analysis and nonlinear ARMA models with Fourier coefficients (FNLARMA). These models are applied to weekly data on interest rates in India and their forecasting performance is evaluated vis-à-vis three GARCH models [GARCH (1,1), GARCH-M (1,1) and EGARCH (1,1)] as well as the random walk model. Both the WANN and MS methods show marked improvement over other benchmark models, and may thus hold out several potentials for real world modeling and forecasting of financial data.

The Volatility of Realized Volatility

In recent years, with the availability of high-frequency financial market data modeling realized volatility has become a new and innovative research direction. The construction of “observable” or realized volatility series from intra-day transaction data and the use of standard time-series techniques has lead to promising strategies for modeling and predicting (daily) volatility. In this article, we show that the residuals of commonly used time-series models for realized volatility and logarithmic realized variance exhibit non-Gaussianity and volatility clustering. We propose extensions to explicitly account for these properties and assess their relevance for modeling and forecasting realized volatility. In an empirical application for S&P 500 index futures we show that allowing for time-varying volatility of realized volatility and logarithmic realized variance substantially improves the fit as well as predictive performance. Furthermore, the distributional assumption for residuals plays a crucial role in density forecasting.

A Gamma Activity Time Process with Noninteger Parameter and Self-Similar Limit

We construct a process with gamma increments, which has a given convex autocorrelation function and asymptotically a self-similar limit. This construction validates the use of long-range dependent t and variance-gamma subordinator models for actual financial data as advocated in Heyde and Leonenko (2005) and Finlay and Seneta (2006), in that it allows for noninteger-valued model parameters to occur as found empirically by data fitting.

A Gamma Activity Time Process with Noninteger Parameter and Self-Similar Limit

We construct a process with gamma increments, which has a given convex autocorrelation function and asymptotically a self-similar limit. This construction validates the use of long-range dependent t and variance-gamma subordinator models for actual financial data as advocated in Heyde and Leonenko (2005) and Finlay and Seneta (2006), in that it allows for noninteger-valued model parameters to occur as found empirically by data fitting.

Tukey-Type Distributions in the Context of Financial Data

Using the Gaussian distribution as probabilistic model for (leptokurtic) financial data is widespread, especially in practice. However, departure from normality seems to be more the rule than the exception. The H-distributions, introduced by Tukey (19601977), are generated by a single transformation (H-transformation) of the standard normal distribution or, more generally, of a symmetric “parent” distribution Z and allow for leptokurtosis represented by the (elongation) parameter H > 0. Alternatively, the J-distributions of Fischer and Klein (2004) or the K-distributions of Haynes et al. (1997) may be applied. In order to additionally introduce skewness, some have these distribution families have been generalized subsequently. Within this work we “complete” the class of so-called Tukey-type distributions by introducing KQ- and JQ-distributions, on the one side, and KK-, JJ-, and -distributions, on the other side. Moreover, we investigate the goodness-of-fit of such Tukey-type distributions for different parent distributions Z in the context of financial return data. In particular, the interplay between Z and different transformations is focussed. Finally, our results are compared to those of popular multi-parametric distribution models.

Focus on software

Focus on software

Multivariate Stochastic Volatility via Wishart Processes

Financial models for asset and derivatives pricing, risk management, portfolio optimization, and asset allocation rely on volatility forecasts. Time-varying volatility models, such as generalized autoregressive conditional heteroscedasticity and stochastic volatility (SVOL), have been successful in improving forecasts over constant volatility models. We develop a new multivariate SVOL framework for modeling financial data that assumes covariance matrices stochastically varying through a Wishart process. In our formulation, scalar variances naturally extend to covariance matrices rather than to vectors of variances as in traditional SVOL models. Model fitting is performed using Markov chain Monte Carlo simulation from the posterior distribution. Because of the model's complexity, an efficiently designed Gibbs sampler is described that produces inferences with a manageable amount of computation. Our approach is illustrated on a multivariate time series of monthly industry portfolio returns. A test of the economic value of our model found that minimum-variance portfolios based on our SVOL covariance forecasts outperformed out-of-sample portfolios based on alternative covariance models, such as dynamic conditional correlations and factor-based covariances.

Bivariate Time Series Modeling of Financial Count Data

A bivariate integer-valued moving average (BINMA) model is proposed. The BINMA model allows for both positive and nagative correlation between the counts. This model can be seen as an inverse of the conditional duration model in the sense that short durations in a time interval correspond to a large count and vice versa. The conditional mean, variance, and covariance of the BINMA model are given. Model extensions to include explanatory variables are suggested. Using the BINMA model for AstraZeneca and Ericsson B, it is found that there is positive correlation between the stock transactions series. Empirically, we find support for the use of long-lag bivariate moving average models for the two series.

TESTING GOODNESS OF FIT BASED ON DENSITIES OF GARCH INNOVATIONS

Testing goodness (or lack) of fit for distributions of observable and nonobservable random variables is one of the main topics in statistics. When they exist, the corresponding density functions and their shapes allow researchers to easily recognize the underlying distribution functions. The present paper is concerned with the densities of (unobservable) generalized autoregressive conditional heteroskedasticity (GARCH) innovations and also with developing goodness-of-fit tests for the densities. Specifically, we construct and investigate large-sample properties of a kernel-type density estimator for GARCH innovations based on (observable) residuals. The authors sincerely thank the Co-Editor Oliver Linton and three anonymous referees for constructive criticism and suggestions that helped us to prepare a much revised version of the original manuscript. The feedback by participants of the Conference on Statistical Models for Financial Data at the University of Graz in May 2004 is also greatly appreciated. The research of the first author was partially supported by NSF grant INT-0223262 and NATO grant PST.EAP.CLG 980599. The research of the second author was partially supported by a Discovery Research Grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada.

Multivariate Stochastic Volatility Via Wishart Random Processes

Financial models for asset and derivatives pricing, risk management, portfolio optimization, and asset allocation rely on volatility forecasts. Time-varying volatility models, such as GARCH and Stochastic Volatility (SVOL), have been successful in improving forecasts over constant volatility models. We develop a new multivariate SVOL framework for modeling financial data that assumes covariance matrices stochastically varying through a Wishart process. In our formulation, scalar variances naturally extend to covariance matrices rather than vectors of variances as in traditional SVOL models. Model fitting is performed using Markov chain Monte Carlo simulation from the posterior distribution. Due to the complexity of the model, an efficiently designed Gibbs sampler is described that produces inferences with a manageable amount of computation. Our approach is illustrated on a multivariate time series of monthly industry portfolio returns. In a test of the economic value of our model, minimum-variance portfolios based on our SVOL covariance forecasts outperform out-of-sample portfolios based on alternative covariance models such as Dynamic Conditional Correlations and factor-based covariances.

ESTIMATING THE SKEWNESS IN DISCRETELY OBSERVED LÉVY PROCESSES

We consider models for financial data by Levy processes, including hyperbolic, normal inverse Gaussian, and Carr, Geman, Madan, and Yor (CGMY) processes. They are given by their Levy triplet (μ(θ),σ2,eθxg(x)ν(dx)), where μ denotes the drift, σ2 the diffusion, and eθxg(x)ν(dx) the Levy measure, and the unknown parameter θ models the skewness of the process. We provide local asymptotic normality results and construct efficient estimators for the skewness parameter θ taking into account different discrete sampling schemes.I thank Prof. Dr. L. Ruschendorf for his steady encouragement, the referees for helpful comments, and the German National Scholarship Foundation for financial support.

Mixed Normal Conditional Heteroskedasticity

Both unconditional mixed-normal distributions and GARCH models with fat-tailed conditional distributions have been employed for modeling financial return data. We consider a mixed -normal distribution coupled with a GARCH -type structure which allows for conditional variance in each of the components as well as dynamic feedback between the components. Special cases and relationships with previously proposed specifications are discussed and stationarity conditions are derived. An empirical application to NASDAQ index data indicates the appropriateness of the model class and illustrates that the approach can generate a plausible disaggregation of the conditional variance process, in which the components’ volatility dynamics have a clearly distinct behavior that is, for example, compatible with the well-known leverage effect.

Fitting non-gaussian Models to Financial data: An Empirical Study

In this paper are presented some experiences about the modeling of financial data by three classes of models as alternative to Gaussian Linear models. Dynamic Volatility, Stable Lévy and Diffusion with Jumps models are considered. The techniques are illustrated with some examples of financial series on currency, futures and indexes.

Financial data modeling by using asynchronous parallel evolutionary algorithms

In this paper, the high-level knowledge of financial data modeled by ordinary differential equations (ODEs) is discovered in dynamic data by using an asynchronous parallel evolutionary modeling algorithm (APHEMA). A numerical example of Nasdaq index analysis is used to demonstrate the potential of APHEMA. The results show that the dynamic models automatically discovered in dynamic data by computer can be used to predict the financial trends.

Conditional Heteroskedasticity in some Common Count Data Models for Financial Time Series Data

Conditional heteroskedasticity properties are derived for some common count data regression and time series models. New extensions are suggested and discussed.