Volatility Risk Premium and Clustering: Intraday vs Overnight Dynamics

Key Takeaways

• Overnight variance risk premium is consistently negative across US, Europe, Asia
• Intraday variance risk premium often positive but statistically insignificant
• Overnight VRP predicts 6‑12‑month returns; intraday VRP predicts 1‑3‑month returns
• Volatility clustering is stronger in overnight returns than intraday
• Cross‑period clustering between overnight and intraday is weak globally

Pulse Analysis

The variance risk premium has long been viewed as a single, monolithic figure, but recent research shows that its composition varies dramatically between the non‑trading night and the active trading day. By employing a synthetic variance swap framework, scholars have isolated an overnight premium that is persistently negative, reflecting the compensation investors demand for bearing uncertainty when markets are closed. This component carries predictive weight for medium‑ to long‑term equity performance, suggesting that models which ignore the nocturnal slice may underestimate risk exposure and misprice volatility‑linked instruments.

Parallel investigations into volatility clustering reveal a complementary story. Across fifteen global equity markets, both developed and emerging, clustering—where high‑volatility periods tend to follow one another—appears in intraday and overnight returns alike. However, the magnitude is consistently larger overnight, indicating that price swings during the quiet hours are more persistent. The weak cross‑clustering between the two regimes implies that shocks in one window rarely transmit directly to the other, reinforcing the case for separate modeling of each period in risk‑management systems.

For practitioners, these insights translate into actionable adjustments. Portfolio managers can tilt exposure toward overnight volatility when seeking longer‑term hedges, while intraday signals may be leveraged for tactical, short‑horizon trades. Risk models that incorporate distinct clustering parameters for each window can better capture tail risk, improving VaR and stress‑test outcomes. As markets continue to extend trading hours and algorithmic participation grows, the relevance of dissecting risk across timeframes will only increase, making the overnight‑intraday split a critical lens for future strategy development.