Stochastic Volatility Models: A Survey with Applications to Option Pricing and Value at Risk

Abstract

This chapter presents an introduction to the current literature on stochastic volatility models. For these models the volatility depends on some unobserved components or a latent structure.Given the time-varying volatility exhibited by most financial data, in the last two decades there has been a growing interest in time series models of changing variance and the literature on stochastic volatility models has expanded greatly. Clearly, this chapter cannot be exhaustive, however we discuss some of the most important ideas, focusing on the simplest forms of the techniques and models used in the literature.The chapter is organised as follows. Section 1 considers some motivations for stochastic volatility models: empirical stylised facts, pricing of contingent assets and risk evaluation. While section 2 presents models of changing volatility, section 3 focuses on stochastic volatility models and distinguishes between models with continuous and discrete volatility, the latter depending on a hidden Markov chain. Section 4 is devoted to the estimation problem which is still an open question, then a wide range of possibility is given. Sections 5 and 6 introduce some extensions and multivariate models. Finally, in section 7 an estimation program is presented and some possible applications to option pricing and risk evaluation are discussed.Readers interested in the practical utilisation of stochastic volatility models and in the applications can skip section 4.3 without hindering comprehension.

Volatility forecasting in emerging markets with application of stochastic volatility model

“…Z t and u t are error terms for the asset return and log-normal volatility, respectively. Prior researches had argued that SVM models provide a better description of volatility than GARCH models, because the former imitate the volatility behaviours observed in financial markets (Andersen and Sørensen, 1996;Billio and Sartore, 2003). Many financial assets present clustering, time-varying, diffusion and jumping behaviours or they are generated with a noise term that depends on current returns.…”

Eficiencia de las técnicas de medición del riesgo de mercado ante situaciones de crisis

“…En este sentido, emplear modelos más fl exibles y generales, que permiten recoger efectos asimétricos del precio sobre la volatilidad como son los FIEGARCH, FIAPARCH, SV (volatilidad estocástica) y otros similares, tal y como hace entre otros, Billio y Sartore (2003), Eberlein et al (2003), Sadorsky (2005), Chan et al (2007) y Bali y Theodossiou (2007), limitaría las posibles soluciones al problema de scaling tal y como se analiza más adelante.…”

Measuring time dependent volatility and cross-sectional correlation in Australian equity returns

“…By adopting this approach we are using the Kalman filter and its likelihood function as an optimisation method to obtain estimates for both the state variable h(t) and the state-space model parameters. Even though ξ(t) is non-Gaussian, this procedure will provide the best linear unbiased estimator of h(t) given r (t) [8,23]. We note that this approach for estimating volatility does have several limitations.…”