Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Rethinking the Idiosyncratic Volatility Puzzle: Long-term and Short-term

Wang, Zixiang LU (2013) NEKP02 20121
Department of Economics
Abstract (Swedish)
The empirical evidence on the cross-sectional relation between stock returns and monthly idiosyncratic volatility is mixed. Ang, Hodrick et al. (2006) document that stocks with high past idiosyncratic volatility underperformed those with low idiosyncratic volatility. In this thesis, I show that the pricing effect of idiosyncratic volatility disappears when a sample of stocks with high market value of equity is used. There is no significant relation between monthly idiosyncratic volatilities and future returns. I further decompose the monthly idiosyncratic volatilities into long-run and short-run parts by using the Wavelet Analysis and the Hodrick–Prescott filter. When the two components are added into the cross-sectional regressions with... (More)
The empirical evidence on the cross-sectional relation between stock returns and monthly idiosyncratic volatility is mixed. Ang, Hodrick et al. (2006) document that stocks with high past idiosyncratic volatility underperformed those with low idiosyncratic volatility. In this thesis, I show that the pricing effect of idiosyncratic volatility disappears when a sample of stocks with high market value of equity is used. There is no significant relation between monthly idiosyncratic volatilities and future returns. I further decompose the monthly idiosyncratic volatilities into long-run and short-run parts by using the Wavelet Analysis and the Hodrick–Prescott filter. When the two components are added into the cross-sectional regressions with control variables, I find that the long-term component of idiosyncratic volatility is negatively related with future return, while the short-term component is positively related with future return. Both of the two relations are highly significant. The results indicate that the net pricing effect of idiosyncratic volatilities may depend on the relative dominance of the two components. This finding might explain some of the mixed results from previous cross-sectional studies. (Less)
Please use this url to cite or link to this publication:
author
Wang, Zixiang LU
supervisor
organization
course
NEKP02 20121
year
type
H2 - Master's Degree (Two Years)
subject
keywords
idiosyncratic volatility, cross-sectional returns, decomposition, wavelet analysis
language
English
id
3411477
date added to LUP
2013-02-15 10:39:28
date last changed
2013-02-15 10:39:28
@misc{3411477,
  abstract     = {The empirical evidence on the cross-sectional relation between stock returns and monthly idiosyncratic volatility is mixed. Ang, Hodrick et al. (2006) document that stocks with high past idiosyncratic volatility underperformed those with low idiosyncratic volatility. In this thesis, I show that the pricing effect of idiosyncratic volatility disappears when a sample of stocks with high market value of equity is used. There is no significant relation between monthly idiosyncratic volatilities and future returns. I further decompose the monthly idiosyncratic volatilities into long-run and short-run parts by using the Wavelet Analysis and the Hodrick–Prescott filter. When the two components are added into the cross-sectional regressions with control variables, I find that the long-term component of idiosyncratic volatility is negatively related with future return, while the short-term component is positively related with future return. Both of the two relations are highly significant. The results indicate that the net pricing effect of idiosyncratic volatilities may depend on the relative dominance of the two components. This finding might explain some of the mixed results from previous cross-sectional studies.},
  author       = {Wang, Zixiang},
  keyword      = {idiosyncratic volatility,cross-sectional returns,decomposition,wavelet analysis},
  language     = {eng},
  note         = {Student Paper},
  title        = {Rethinking the Idiosyncratic Volatility Puzzle: Long-term and Short-term},
  year         = {2013},
}