On February 25, Izabella Kaminska in FT Alphaville blasts the forecasting ability of the Vix futures, a CBOE futures contract that measures the volatility of the S&P 500 stock market index. Since 2009, the Vix futures has remained in contango–meaning that the far out prices exceed the near delivery future prices-and Vix futures repeatedly exaggerated their predictions of the actual volatility.
The first graph below shows the term structure of the Vix futures contracts for several trading dates in 2010 and 2011. The graph indicates that the futures curve has remained in contango. A closer examination indicates a relationship between the level of volatility and the steepness of the curve: A higher (smaller) level of the spot/near term volatility seems associated with a less (more) pronounced contango. The second graph shows the same relationship after normalizing the near term futures price to 100. On July 20, 2010, the near volatility was 28.45 and the far out volatility was 33.2, a 16.7 percent higher. By January 24, 2011, when the near volatility was a much lower 16.15, the far out contract was 26.6, a whopping 64.7 percent difference.
The graph also shows that between January 24, 2011 and February 23, 2011 the near term volatility increased from 16.15 to 22.25. Between these two dates the futures curve remained in contango, but its slope was considerably reduced from 64.7 percent to a mere 10 percent, the lowest level in months. This shift indicates that after a big reversal in volatility, contango appears to flatten.
These two findings seem to imply that volatility shows some mean reversion. This should not be surprising, for periods of relative calm are succeeded by periods of turbulence when volatility spikes. This is what the next graph reveals.
But if volatility indeed varies through time, then investors watching closely the risk aversion gauge must factor this volatility risk into the Vix futures prices. Vix futures prices are then not just long term expectations of volatility. By incorporating a premium for volatility risk, futures prices systematically deviate from actual volatility. What Izabella Kaminska sees as a forecasting ineptitude of the Vix futures contract is simply the result of the fact that risk contains risk.
According to Bloomberg, the S&P 500’s ride upward in recent months has been accompanied by a 70 percent drop in the options market volatility indicator, an indicator that measures the 30-day realized volatility. The indicator fell to 10 on Feb 15, 2011 from last year’s peak of 32.53 on June 10. And in November 2008, the indicator soared to 82.21 and remained at extraordinary high levels for the most part of 2009. The fact that volatility in recent months has been at very low levels does not mean that it cannot bounce back surprisingly quickly, as it did at least four times since the October 1987 stock market crash, or just last week when the Vix shot up 30 per cent.
There is a second reason as to why the contango was so steep on Vix futures and futures had such a hideous forecasting record during 2010. The first three quarters of 2010 was a period of unusually high background uncertainty, as Fed Chairman Ben Bernanke remarked. Analysts and policymakers’ forecasts showed a striking dispersion. Mohamed El-Erian, CIO of Pimco and an experienced observer of financial markets, characterized the situation in the following way:
“It is the shape of such dispersion that strikes us as particular important. It seems that, wherever we look, the snapshot for ‘consensus expectations’ has shifted: from traditional bell shaped curves – with a high likelihood mean and thin tails – to a much flatter distribution of outcomes with fatter tails.”
When opinions are so deeply divided and expectations vary considerably, volatility tends to widen with maturity, and futures prices become more contangoed.
Anyone who has ever attempted to dance the tango, knows that it is a risky effort, with rapid and sharp gyrations. The danger of a faux-pas is always eminent until the music stops. So it seems to be in the business of betting on volatility. Anyone that trades the Vix knows well how fast it can swing. A hideous forecasting error? Don’t cry for the Vix futures, Argentina!