Most traders use the VIX to time the S&P 500. Rob Hanna tested whether that actually works. His findings, published in a paper that won the 2024 NAMA Founders Award, suggest the conventional approach has it backwards. The S&P times itself better than the VIX does. And when you flip the relationship, SPX readings become a more reliable signal for trading VIX-based instruments.
Rob runs Quantifiable Edges and has been trading full-time since 2001. This episode covers the research in detail: how VIX instruments work, what the data shows about long-term versus short-term VIX indicators, the specific finding that contradicts conventional wisdom, and practical strategies for trading UVXY, VXX, and VIX futures.
Watch the full episode below, then read on for the complete breakdown.
What the VIX actually measures
The VIX is an estimate of implied volatility in S&P 500 options, looking out 23 to 37 days (the range that approximates 30 calendar days). It is not a direct measure of fear. It is a measure of how much the market is paying for protection, and because that cost rises during sell-offs, the VIX and the S&P tend to move inversely.
Rob uses the Rule of 16 to translate VIX readings into expected daily moves. A VIX of 16 implies roughly 1% daily moves in the S&P. A VIX of 32 implies ~2%. VIX of 48 implies ~3%. The practical implication: sustaining a high VIX requires sustained large daily moves. When selling slows down or the market stabilizes, realized volatility drops and implied volatility tends to follow it lower.
This also explains a key observation from the research. You do not necessarily get a high VIX in every bear market. The 2022 bear market was driven by rising interest rates, not the kind of systemic panic that drove 2008 or COVID. The VIX topped out around 35 to 40 in 2022 versus 80 in 2020. Same 20%+ drawdown in the S&P, very different fear environment.
Conventional wisdom: use VIX to time the S&P
The prevailing approach treats VIX readings as signals for S&P positioning. High VIX (overbought, hitting 10-day or 20-day highs) is interpreted as a bounce setup for the S&P. Low VIX (oversold, complacent) is treated as a warning that the S&P may be vulnerable.
This logic is not wrong. Rob’s data confirms it does work. An overbought VIX does predict above-average S&P performance over the next day. The issue is whether the VIX is the best signal available, or just one that looks good because it correlates with something better.
The key finding: S&P times itself better than VIX does
Rob tested both approaches using AmibroKer, bucketing every instance into quintiles based on indicator readings (0-20, 20-40, 40-60, 60-80, 80-100 for RSI, and high/low extremes for multi-day ranges). He used RSI periods of 2, 3, and 4, and multi-day high/low lookbacks of 5, 10, 15, and 20 days.
The finding was consistent across all variations: a 10-day low in the S&P is more predictive of a next-day bounce than a 10-day high in the VIX is. A 20-day low in the S&P outperforms a 20-day high in the VIX. The most oversold quintile of S&P RSI2 predicted better next-day S&P performance than the most overbought quintile of VIX RSI2, even when the number of instances in each bucket was nearly identical.
The flip side was even more notable. An oversold VIX carries almost no useful signal for a decline. VIX can stay in oversold territory far longer than the S&P can stay overbought, so the signal degrades quickly. An overbought S&P, on the other hand, does predict a measurable chance of a down day.
Rob’s summary: traders use the VIX because they think they are looking at something different from the S&P. But that different thing does not work as well as just looking at the S&P directly.
Why long-term VIX indicators fail
Rob also tested whether long-term VIX filters (200-day moving average, golden cross/death cross) could time the S&P. They could not. The same indicators applied directly to the S&P worked well for avoiding bear markets. Applied to the VIX, performance deteriorated versus simple buy-and-hold.
The reason is structural. The S&P drifts upward over time. Moving averages work on instruments with long-term trends. The VIX is mean-reverting. It will always be “high” around 40 and “low” around 10 regardless of the year. Applying a moving average to a mean-reverting instrument with no directional drift just produces choppy, low-quality signals.
VIX trading instruments: futures, ETFs, and the contango headwind
Because the VIX itself cannot be traded directly, there are three main ways to trade volatility:
- VIX futures: Expire monthly, typically on a Wednesday morning. Available out to eight or nine months. The front-month futures price will converge to the VIX at expiration. When VIX is at 14 and the two-month futures are at 16, the market is pricing in a 14% rise in volatility over two months, reflecting the risk premium buyers pay for uncertainty.
- VIX options: Settle at the same time as the front-month futures, not the spot VIX. This means they effectively price off the futures, not the index. Spreads, strangles, straddles, and butterflies are all available.
- Volatility ETFs/ETNs (VXX, UVXY, SVXY): The most commonly traded. These hold a constant-maturity mix of front-month and second-month VIX futures. The key mechanic: VIX futures are in contango roughly 86% of the time historically, meaning the second month is priced higher than the first. As the ETF rolls from front month to second month each day, it sells lower-priced futures and buys higher-priced ones. That is a structural headwind. UVXY (which traded at 2x leverage until the volmageddon event of 2018 forced a reduction to 1.5x) declined from a split-adjusted $22 billion to around $8 over its history, largely because of this contango drag. Short volatility strategies benefit from this; long volatility holders pay it.
Using SPX signals to trade VIX instruments
The practical application of the core finding is to use S&P readings as the signal for VIX trades rather than VIX readings themselves. An oversold S&P, whether measured by RSI2 hitting extreme lows or a 10-20 day price low, predicts VIX increases more reliably than a high VIX does.
This is the approach Rob details in his paper and in a course on VIX trading. Specific strategies use RSI readings on the SPX to time entries and exits in VXX, UVXY, or VIX futures. The paper is publicly available on SSRN (abstract ID 4808230) and provides the full data tables backing the claims.
Assessing strategy robustness in volatility trading
Rob applies the same robustness checks to volatility strategies that he applies to any other system. He looks at whether performance holds across different lookback periods, whether it degrades gracefully as parameters shift, and whether the logic makes intuitive sense given how VIX instruments actually work.
One practical point: the VIX recovery speed after a spike is itself a useful signal. Faster-than-average recovery from an elevated VIX level has historically correlated with shorter drawdown periods in the S&P. Tracking whether VIX is recovering “on schedule” relative to historical patterns gives an additional filter for timing re-entry into equity positions after corrections.
Related episodes
- A new approach to trading volatility with Rob Hanna
- Forecasting a volatility tsunami with Andrew Thrasher
- How to use volatility in trading strategies
Get the show notes & transcript
Want to understand how volatility signals actually work in systematic trading?
Subscribe to the Better System Trader podcast for weekly interviews with the world’s top systematic traders and quantitative researchers.
