faj.v61.n1.pdf (EC3400)

faj.v61.n1.pdf (EC3400) - Financial Analysts Journal Volume...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
January/February 2005 www.cfa pubs .org 45 Financial Analysts Journal Volume 61 Number 1 ©2005, CFA Institute Practical Issues in Forecasting Volatility Ser-Huang Poon and Clive Granger A comparison is presented of 93 studies that conducted tests of volatility-forecasting methods on a wide range of financial asset returns. The survey found that option-implied volatility provides more accurate forecasts than time-series models. Among the time-series models, no model is a clear winner, although a possible ranking is as follows: historical volatility, generalized autoregressive conditional heteroscedasticity, and stochastic volatility. The survey produced some practical suggestions for volatility forecasting. olatility forecasting plays an important role in investment, option pricing, and risk management. We conducted an extensive review of the volatility-forecasting research in the last 20 years (Poon and Granger 2003) and provide here a summary and update of our findings. The definition of volatility we used is the standard deviation of returns. The assets studied in the 93 articles surveyed included stock indexes, stocks, exchange rates, and interest rates from both developed and emerging financial markets. The forecast horizon ranged from one hour to one year (a few exceptions extended the forecast horizon to 30 months and to five years). We review three main categories of time-series model—namely, historical volatility, models in the autoregressive conditional heteroscedasticity (ARCH) class, and stochastic volatility models—as well as forecasting based on implied volatility derived from option prices. We present here a description of these models, a summary of our survey results, and a discussion of the characteris- tics of market volatility that affect the choice of model, common objectives of volatility forecasting, and the impact of outliers. Finally, we provide some practical advice on volatility forecasting. Types of Volatility Models The four types of volatility-forecasting methods we surveyed are historical volatility (HISVOL), ARCH models, stochastic volatility, and option- implied volatility. The HISVOL model is (1) where = expected standard deviation at time t φ = the weight parameter σ = historical standard deviation for periods indicated by the subscripts This group includes random walk, historical aver- ages, autoregressive (fractionally integrated) mov- ing average, and various forms of exponential smoothing that depend on the values of φ , the weight parameter. The second group is the ARCH model and its various extensions, including the nonlinear ones: r t = µ + ε t , (2a) where r t = return of the asset at time t µ = average return ε t = residual returns, defined as (2b) where z t is standardized residual returns and h t is conditional variance, defined as (2c) in which ω = a constant term p = number of autoregressive terms j = order of the autoregressive term β = autoregressive parameter q = number of moving-average terms k =order o f the moving-average term α = moving-average parameter
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 02/28/2010 for the course ECO 211 taught by Professor Gilo during the Spring '10 term at Young Harris.

Page1 / 12

faj.v61.n1.pdf (EC3400) - Financial Analysts Journal Volume...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online