Forecasting volatility via conditional autoregressive value at risk model based on support vector quantile regression

다국어 초록 (Multilingual Abstract)

The conditional autoregressive value at risk (CAViaR) model is useful for risk management, which does not require the assumption that the conditional distribution does not vary over time but the volatility does. But it does not provide volatility forecasts, which are needed for several important applications such as option pricing and portfolio management. For a variety of probability distributions, it is known that there is a constant relationship between the standard deviation and the distance between symmetric quantiles in the tails of the distribution. This inspires us to use a support vector quantile regression (SVQR) for volatility forecasts with the distance between CAViaR forecasts of symmetric quantiles. Simulated example and real example are provided to indicate the usefulness of proposed forecasting method for volatility.

참고문헌 (Reference)

1 Vapnik, V, "Statistical learning theory" Wiley 1998

2 심주용, "Restricted support vector quantile regression without crossing" 한국데이터정보과학회 21 (21): 1319-1325, 2010

3 Koenker, R, "Regression quantile" 46 : 33-50, 1978

4 Li, Y., "Quantile regression in reproducing kernel Hilbert spaces" 102 : 255-268, 2007

5 황창하, "Mixed Effects Kernel Binomial Regression" 한국데이터정보과학회 19 (19): 1327-1334, 2008

6 황창하, "Kernel Machine for Poisson Regression" 한국데이터정보과학회 18 (18): 767-772, 2007

7 Yuan, M, "GACV for quantile smoothing splines" 50 : 813-829, 2006

8 Mercer, J, "Functions of positive and negative type and their connection with the theory of integral equations" 415-446, 1909

9 Perez-Cruz, F., "Estimating GARCH models using support vector machines" 3 : 1-10, 2003

10 김말숙, "Claims Reserving via Kernel Machine" 한국데이터정보과학회 19 (19): 1419-1427, 2008

11 Engle, R. F, "CAViaR: Conditional autoregressive value at risk by regression quantiles" 22 : 367-381, 2004

12 Pearson, E. S, "Approximate means and standard deviations based on distances between percentage points of frequency curves" 52 : 533-546, 1965