There problem with the VIX: it no longer seems able to properly gauge the overall level of risk aversion. In particular, I showed here that the VIX does not drive the portfolios flows towards EM markets anymore. Nor does it explain or even track the returns of high yield bonds or emerging assets.
In a year characterized by the widening of local/idiosyncratic risk, it is not really a surprise. But if the VIX does not drive a wide range of risky assets anymore, what are the factors that influence the VIX?
By construction, the VIX index is supposed to focus on US risk (implied volatility of US stock market). The charts below show two things:
i. The VIX (and the curve of the VIX futures strip) is correlated with the volatility of the economic cycle. In particular, any rise in economic uncertainty (I use the 6-month rolling standard deviation of the monthly changes of the ISM Manufacturing as a proxy) is associated with a higher VIX
or a flattening, or even inversion, of the curve.
ii. There are periods when the VIX shuns rises in the economic volatility. This was the case between 2004 and 2007. This has also been the case since early 2012. The growing economic volatility of the last few quarters has been ignored by the VIX. There is an interesting disconnect: even though the level of the VIX did not adjust, the slope of the VIX curve did.
It is worth noting that the periods of disconnect between the VIX and the volatility of the economic cycle are generally those when there is a strong consensus among analysts on the future path of the economy. As can be seen below, 2004/2007 and today are both characterized by a very low dispersion of the views gathered by the consensus forecast.
In such periods, a temporary rise in the economic volatility might not be able to affect the consensus view, which would explain why investors did not panic excessively in the wake of a bad release of the ISM Manufacturing.
It also highlights the fact that only a very strong shock on the prospects for growth in the US, deflation in the Eurozone, failure of Abenomics, or below-7% growth in China might hamper the ongoing positive view on global growth for 2014. So far, the lack of spillover from the EM crisis is clearly reflective of the view that most of the struggling countries have to deal with disequilibrium that they created or did not manage well enough on their own.
It also suggests that some radical switches in the nature of the risks have to be observed before the consensual view evolves in a direction that would spur a much higher level for VIX. Only then would this indicator be a signal of a genuine risk-off regime.
This explanation seems more convincing than the lack of external uncertainty. The resilience of the VIX has been particularly strong recently, as it also remained particularly insensitive to a spike in political risk since early 2013 (see chart below).
Bottom Line: The VIX is a poor indicator of global risk aversion. This may be temporary, but it is the case for the time being. In a global economy characterized by growing idiosyncratic risks, this is not really a surprise. Yet, the ongoing disconnect of the VIX with the economic volatility has to be explained. To me, it mainly reflects the historically low dispersion of economic forecasts. The consensus for 2014 remains rather strong in spite of the recent EM crisis and US data-related challenges. January's NFP report, by the duality of its message (lower than expected job creation but a fall in the unemployment ratio, in spite of a higher participation rate), will not change the picture.
What does it mean for volatility trading? It is always necessary to repeat that for retail investors, a volatility trade is not a buy and hold strategy as the returns of VIX-related ETFs such as VXX or VXZ are caught in a "contango trap": the curve of the VIX futures curve is structurally positive, which means that there is a negative carry for any VIX-related ETF holders. Even if VIX-related ETFs are negatively correlated to traditional risky assets, their negative carry (hence poor return performance) make them unattractive on a buy-and-hold basis.
Episodes of inversion of the VIX curves happen concurrently with spikes in the VIX. This requires building leading indicators of forthcoming VIX spikes to manage the timing of VXX investing. From the analysis above, the recent volatility in the economic news flow is not enough, neither are the ongoing troubles in EM markets. It would require a genuine regime switch (deflation in Europe, sharp decline in US CapEx spending, below 7% growth in China) to make VIX rise much higher.
If you don't really believe in that scenario, a wait-and-see attitude might be more rewarding than a buy-and-hold strategy on volatility-related products. For those willing to start to buy protection as soon as possible, I would recommend VIX ETFs based on medium dated futures, where the steepness (hence the roll) of the future curve is the smallest.