VIX/VIX3M Inverse Volatility Timing

The VIX/VIX3M ratio (formerly VIX/VXV) compares the Cboe's 30-day implied volatility index (VIX) to its 3-month implied volatility index (VIX3M). Developed by investment analyst Vance Harwood in 2012 and published on SixFigureInvesting.com, this simple but powerful ratio is used to time inverse volatility exchange-traded products.

The Problem with Inverse Volatility

Inverse volatility strategies profit from contango — the natural upward slope of the VIX futures term structure where longer-dated futures trade at a premium to spot. This structure causes long volatility products to lose value over time due to the roll cost, while inverse products gain. Contango is present roughly 75–80% of the time.

The challenge is the remaining 20–25% when volatility spikes. During these events (often brief but violent), inverse volatility products can lose 20%, 50%, or even 96% in a single day. The Feb 5, 2018 "Volmageddon" event — where XIV lost 96% in a single session — is the extreme example. The VIX/VIX3M ratio strategy aims to capture the contango profits while sidestepping these catastrophic spikes.

The Ratio

The core insight is simple: compare the 1-month (VIX) to 3-month (VIX3M, formerly VXV) implied volatility of SPX options using a ratio. This strips away absolute levels and only cares about the shape of the volatility term structure:

VIX / VIX3M = Ratio

RatioStructureMeaning
< 0.85Strong ContangoOptimal for short vol — steep roll yield
0.85 – 0.917ContangoFavorable for short vol — normal conditions
0.917 – 1.0Elevated / WatchTerm structure flattening — exit zone per Harwood
≥ 1.0BackwardationInversion — inverse vol positions at severe risk

The Strategy

Core Rules

  • Exit / stay in cash: When VIX/VIX3M ratio ≥ 0.917
  • Enter / stay invested: When VIX/VIX3M ratio < 0.917 (some practitioners use 0.9 for entry to reduce flip-flopping)
  • Position: All-in / all-out on the signal

Refinements

  • Two-day confirmation: Wait 2 consecutive days above 0.917 before exiting to reduce noise-driven churn
  • High-ratio caution: If the ratio was in the 0.95–0.97 range the prior day, wait an extra day before re-entering after it drops below threshold
  • Mid-term contango monitor: Normal medium-term contango runs ~2.5–3.5% per month. If it drops below this, the structure is flattening and more susceptible to volatility upticks

Original Instrument: ZIV

The strategy was developed using ZIV (VelocityShares Medium-Term Inverse Volatility Fund), which tracked the inverse (-1×) of the SPVXMTR medium-term rolling index — holding inverse positions in 4-to-7-month VIX futures. ZIV was inherently less volatile than short-term products like XIV or SVXY.

ZIV was liquidated in 2018 after the XIV crash. Modern proxies include SVXY (-1× short-term VIX futures, but more volatile) or direct VIX futures / options strategies.

Performance History

PeriodPerformanceContext
2004–2010SidewaysLow medium-term contango
2010 – Aug 2013ExcellentSteepening contango (70% ann., 11% DD, 9 Sharpe)
Sep 2013 – Jun 2016SidewaysChoppy vol, mean-reverting VIX
2017StrongAll short vol strategies worked
Feb 5, 2018IN CASHRatio ≥ 0.917 on Feb 1. XIV -96%, ZIV -23%

Why It Works

The VIX/VIX3M ratio captures the shape of the volatility term structure without caring about absolute levels. When the ratio is low, the curve is steeply contango — ideal for short vol. As the ratio rises toward 1.0, the curve flattens, and the market is pricing in higher near-term volatility relative to longer-term expectations — a warning sign.

The threshold of 0.917 was identified empirically through backtesting, and notably, Barclays' XVZ dynamic volatility fund independently uses 0.9 for a similar purpose — shifting from its largest short volatility position to a less short portfolio.

Limitations

  • ZIV is defunct — modern implementation requires SVXY or custom options strategies (both more volatile)
  • The optimal threshold (0.917) was identified during a specific contango regime and may not hold in all market environments
  • The strategy goes sideways when medium-term contango is absent (2004–2010, 2013–2016)
  • The 70%/11% backtest results were during a unique period of steep contango from 2010–2012
  • Not suitable for large capital due to liquidity constraints in the underlying products

References