Since the global financial crisis, fundamental active fund managers have been complaining about the lack of volatility in financial markets, with the most popular measure of risk trading at historically low levels. This was likely to have been driven by the decision of the four main central banks to maintain their quantitative easing programmes.
However, volatility staged a return earlier this year, catching investors by surprise. On 5 February, a sudden bout of market volatility decimated XIV (an inverse volatility product), in a day so torrid that traders have since dubbed it “vol-mageddon”. The collapse of XIV and two other similar funds only exacerbated the turmoil. The US stock market suffered one of the swiftest 10 per cent corrections in history, and global equities lost $4.2trn (£3trn) over the course of the week.
As we all know, complex credit-derivative instruments provoked chaos in 2007, and earlier this year investors feared that history might repeat itself, with volatility ETFs replacing credit derivatives as the trigger of the next crisis.
Nevertheless, these instruments remain useful as a way of gauging the current levels of fear in the market rather than purely becoming a way for investors to speculate on future volatility.
Let’s take the example of the VIX, which is undoubtedly the most popular volatility index. VIX is calculated from the pricing of options on the S&P 500 index, so it is supposed to indicate how much volatility is being priced in for the coming three months.
When the index has climbed above 20 in the past, US equity market sell-offs have typically followed and this occurred again in March. Looking over the last five years, the weekly correlation between forecast and realised volatilities on the S&P 500 stands at 62.6 per cent, which indicates a strong positive relationship between implied and actual volatility.
Financial engineers can be very innovative. They have applied the methodology behind the VIX to currency and bond markets, creating CVIX and MOVE.
The CVIX is calculated based on the three-month implied volatilities of major currency pairs, and the MOVE index is calculated based on the three-month implied volatility of two, three and five-year Treasury yield options. Once again, the correlations between implied and realised volatility on those indexes stands at more than 60 per cent.
Nevertheless, we should not confuse correlation with causation as many investors still do.
US asset management company Tallus Capital Management has studied the relationship between the MOVE index and US Treasuries, focusing on the tapering tantrum of June 2013. They conclude that “while the MOVE index may well gauge current sentiment in the markets, its predictive power in the direction and severity of yield moves appears lacking.”
A common problem with volatility products is that investors only look at the product name without checking the methodology. Yes, there is a correlation between these instruments and realised volatility in markets, but this is purely due to their construction.
There is little evidence of a causal relationship, so they should not provide any real hedging power. After all, VIX, CVIX and MOVE indicate investors’ risk forecasts. As with economic forecasts, investors should therefore remain aware that their predictive power is limited, so should only be used to gauge current, and not future, market sentiment.
Charles Younes is a research manager at FE
