Volatility Library

Article by: Peter Christoffersen, Vihang Errunza, Kris Jacobs, Hugues Langloi
Published by: 16th Annual Global Investment Conference
Date: 16 Mar 2010

“Since understanding and quantifying the evolution of security co-movements is critical for asset pricing and portfolio allocation, we investigate patterns and trends in correlations over time using weekly returns for large systems of developed markets (DMs) and emerging markets (EMs) during the period 1973-2009. We use the DECO, DCC, and BEKK correlation models, and develop a novel dynamic t-copula which generalizes the normal copula, to allow for dynamic tail dependence. We demonstrate that it is possible to overcome the well known dimensionality problems and compute correlation and tail dependence in international markets using large samples, without relying on factor models. Our results suggest that correlations have been significantly trending upward for both the DMs and EMs. Further, the evidence clearly contradicts the decoupling hypothesis. Although the tail dependence is increasing through time for both EMs and DMs, the level of the tail dependence is still very low at the end of our sample period for EMs as compared to DMs. Therefore, while the correlation analysis suggests that the diversification potential of EMs has largely disappeared, this is contradicted by our findings on tail dependence. Thus, even though diversification benefits might have lessened in the case of DMs, the case for EMs remains intact.”

Full article (PDF): Link

Article by: Robert F. Whitelaw
Published by: NYU, Stern School of Business
Date: 19 Nov 1997

“This paper documents predictable time-variation in stock market Sharpe ratios. Predetermined
nancial variables are used to estimate both the conditional mean and volatility of equity returns,
and these moments are combined to estimate the conditional Sharpe ratio. In sample, estimated
conditional Sharpe ratios show substantial time-variation that coincides with the variation in ex
post Sharpe ratios and with the phases of the business cycle. Generally, Sharpe ratios are low
at the peak of the cycle and high at the trough. In out-of-sample analysis, using 10-year rolling
regressions, we can identify periods in which the ex post Sharpe ratio is approximately three times
larger than its full-sample value. Moreover, relatively naive market-timing strategies that exploit
this predictability can generate Sharpe ratios more than 70% larger than a buy-and-hold strategy.”

Full article (PDF): Link