Overview
The Volatility Risk Premium (VRP) is the persistent tendency for Implied Volatility to exceed subsequent Realized Volatility, averaging 4-5 annualized percentage points on the S&P 500 since 1990. It is fair compensation for bearing undiversifiable, unhedgeable risks — not a market inefficiency — though it turns sharply negative during genuine crashes, which is exactly when short-volatility sellers get hurt most.
Key Concepts
- Why the Premium Persists (Rational) — Jump Risk (overnight gaps where delta-hedging is impossible), Correlation Risk (diversification fails exactly during crises, when correlations go to 1), and Vega Convexity (option sellers are “short convexity”: losses accelerate non-linearly as turmoil increases).
- Risk-Neutral (ℚ) vs. Physical (ℙ) Measures — the VRP exists because the ℚ measure (used for pricing, e.g. the VIX) assigns a higher probability to crash events than the ℙ measure (real-world historical probability); that gap between assigned probabilities is the VRP.
- Limits to Arbitrage — pensions/endowments are structurally net-long and natural protection buyers (permanent demand imbalance), while shorting volatility is capital-intensive and subject to margin calls during crises that force liquidation at the worst possible time.
Measuring the Premium
- Practitioner's Spread (Ex-Post) —
VRP_t = IV_t − RV_(t,t+30), comparing today's VIX against the subsequent 30-day realized volatility (annualized std. dev. of daily log returns, scaled by √(252/N)). - Academic Variance Risk Premium — uses Variance rather than Volatility, since variance swaps can be perfectly statically replicated (model-free), making it the “purest” measure:
VRP_t = E^ℚ[∫σ²ds] − E^ℙ[∫σ²ds].
Harvesting Strategies
| Strategy | Yield | Primary Risk | Complexity |
|---|---|---|---|
| Short Put (ATM) | High | High (equity beta ~0.6) | Low |
| Short Straddle | Very High | Extreme (gamma risk) | Medium |
| Iron Condor | Medium | Defined/capped | Medium |
| Variance Swaps | Purest | Convex (Vega²) | High (institutional) |
Systematic Put-Writing — selling 10-20% OTM index puts, holding collateral for max loss. Key Greeks: short Vega (profits as IV falls), short Gamma (delta becomes more negative as the market falls, forcing sales into weakness — the source of negative skew), long Theta (daily premium decay works in the seller's favor). The Cboe PutWrite Index (PUT) has historically delivered equity-like returns at only 50-70% of market beta, with drawdowns concentrated in sharp down-moves.
Case Study: Volmageddon (Feb 2018)
On Feb 5, 2018, the S&P 500 dropped ~4%, triggering forced end-of-day rebalancing by leveraged short-vol VIX ETPs (like XIV), which had to buy VIX futures into a liquidity panic — driving VIX from ~16 to ~34 in minutes. XIV lost ~96% of its value in one hour and was liquidated. The lesson: the theoretical VRP edge and the structural risk of leveraged short-vol products are two very different things.
Is VRP Always Positive?
- Positive VRP (85-90% of the time) — Implied > Realized; the standard premium option sellers collect.
- Negative VRP — occurs during genuine crashes (2008, March 2020) when realized volatility exceeds even the spiked implied volatility; this is when short-vol sellers suffer large losses.
- VRP Timing — a wide VRP spread signals low market complacency and often precedes a period of low realized volatility, making it a potentially attractive time to sell — the opposite of intuition.
Key Academic Research
- Carr & Wu (2009), “Variance Risk Premia” — established standard synthetic-variance-swap methods for measuring VRP; found VRP is strongly negative (investors pay to hedge) and, surprisingly, predicts future equity returns.
- Bollerslev, Tauchen, Zhou (2009) — linked VRP to macroeconomic uncertainty; a high VRP spread is one of the best short-term predictors of higher aggregate stock returns.
Key Takeaways
- VRP is compensation for real, unhedgeable risk (jump risk, correlation risk, vega convexity) — not free money or pure market inefficiency.
- The ℚ vs. ℙ measure gap is the theoretically clean explanation for why VRP must exist at all.
- Negative VRP episodes are rare (~10-15% of the time) but concentrated exactly during crashes — the worst possible time for a short-vol position.
- Structural products (leveraged short-vol ETPs) carry risks well beyond the underlying VRP edge itself, as Volmageddon demonstrated.
Related Reading
- Demystifying the Volatility Risk Premium: Theory, Measurement, and Trading Strategies — full article with the complete strategy comparison table and academic literature review.
- Full Research Paper