Abstract
Closed-End Fund (CEF) discounts have intrigued researchers for decades. Of the many explanations offered, the behavioural framework of Lee et al. (Citation1991), which posits noise traders subject to sentiment, is the most discussed. In this article, we contribute some novel evidence to the evaluation of this theory by examining the role of implied market volatility (VIX, i.e., the ‘fear index’) in fund discounts using a Dynamic Conditional Correlation (DCC) approach. We find that VIX has almost no role in determining discounts except during periods of extreme market turbulence, providing strong but indirect evidence for the sentiment story.
Notes
1 Normalization is required because the principal component analysis is not scale-invariant.
2 The DCC model is a generalization of the Constant Conditional Correlation (CCC) estimator of Bollerslev (Citation1990). We skip technical details for the Dynamic Conditional Correlation (DCC) model to save space.
3 Detailed information on individual funds is available upon request.
4 Engle's (2002) test for the CCC is accepted for the tranquil (11 May 2004 to 13 July 2007) and turbulent periods (16 July 2007 to 17 February 2011) with 0.7959 and 0.9184 p-values, respectively. All results are available upon request.