Abstract:
|
This paper adopts a versatile multivariate conditional correlation model
to estimate daily seasonality in the returns, the volatility, and the correlations between stocks, bonds, gold and Bitcoin. Besides the well known
seasonality in stocks and bonds, the day-of-the-week effect is also present
in Bitcoin. Mondays are associated with higher Bitcoin returns, while
Wednesdays with higher Bitcoin volatility. As opposed to previous literature, our results indicate strong evidence of Bitcoin’s leverage effect.
Moreover, we show that daily correlations between Bitcoin and traditional
assets are higher at the beginning of the week, while the volatility of these
correlations decreases over the week. Our results offer interesting insights
in terms of investment and portfolio diversification, that can be applied
to the analysis of systematic risk asset allocation and hedging.
Keywords: Day-of-the-week effect; dynamic conditional correlation; Bitcoin; volatility seasonality.
JEL codes: G01; G10; G12; G22 |