We ♥ this note from Bank of America Merrill Lynch’s Ruslan Bikbov and Priya Misra.
It’s on a subject dear to our own hearts here on FT Alphaville — the curious case of persistently low volatility and the idea that it might be masking systemic risk. It also weaves together a plethora of other themes — massive short volatility positions, search for yield, correlation, LTCM – we’ve touched on.
So sit back, think of stubbornly low volatility measures like the Vix, and read on:
The Chart of the Day shows that our measure of average cross-asset correlation has notably increased. This measure is a weighted-average correlation of the 10y US swap rate with the S&P 500, oil and DXY dollar index. The rise in correlations is evident in almost all individual cross-asset pairs. For example, the three-month correlation between 10y and S&P increased to 64% (1.7 standard deviation above its historical mean). In the Chart of the Day, a positive value corresponds to a risk-on/ risk-off trading environment. Days when rates increase tend to coincide with the rallies in risky assets and weaker US dollar, while days when rates decline tend to go along with sell-offs in risky assets and dollar strengthening.
It is very unusual, however, to see high levels of cross-asset correlation together with declining volatility. This is because cross-asset correlations tend to rise during times of market stress, and these times normally experience high volatility. We have found only two historical episodes where high correlations coincided with declining volatilities: May-August 1998 and June-July 2010. In both cases, however, volatility subsequently jumped higher: in the first episode it was the collapse of Long Term Capital Management; and in the later episode it was capitulation of growth expectations and onset of QE2. Although this time may be different, we believe high cross-asset correlations are signaling two important systemic risks which might bring higher market volatility:
- * It’s all one trade. When cross-asset correlations are high, it means that many investors are essentially in one trade, even though they may not be aware of it. As a result, there may be more crowded trades than most investors realize. If investors exit at the same time, market movements could be chaotic.* Focus on one risk factor only. High risk-on/risk-off correlation implies that investors are increasingly focused on a single risk factor that affects all asset classes. Although our correlation analysis is silent about the nature of what this factor may be, we suspect it is related to global growth risks. The focus on a single factor, however, suggests that investors may underestimate other risks. If another major risk factor suddenly arises, for example something related to the US debt ceiling, we could see a great increase in market volatilities.
What explains high cross-asset correlation? As we just mentioned, one potential explanation is that investors are ignoring other risk factors. Another explanation is that the environment of very low interest rates created by highly accommodative Fed policy has prompted investors to search for higher returns in more risky assets. This is known as “portfolio rebalancing” channel of monetary policy, as defined by Fed officials. Logically, this should lead to growing capital and leverage allocated to riskier assets and trading strategies.
In particular, the chase for yield could be one of the main reasons behind an impressive growth of the hedge fund industry. The 2010 net inflows in macro hedge funds of US$17bn were even larger than in 2006 and 2007 … Note that financial stress largely calmed down by mid-2009, so a reversal of the flight to- quality sentiment is a less likely explanation of the growth of macro hedge fund community in 2010. As of the end of 2010, macro hedge funds had assets under management of an all-time record US$380bn. Although hedge-fund leverage is not readily measurable, because it can be achieved through the use of derivatives, we believe that the search for yield might have forced hedge funds to increase leverage since the financial crisis.
The chase for yield has made selling volatility a popular trade, and this has been a factor in low implied and realized vols. At the same time, growing macro (crossasset) portfolios likely contributed to the rise of cross-asset correlations. Hence, an extremely accommodative monetary policy may have simultaneously resulted in low volatility and high cross-asset correlations. Note that although QE2 is coming to an end, monetary policy remains extremely accommodative and we expect the chase for returns to continue until the Fed signals an exit.
Given that the assets under management of macro hedge funds are 30% higher than in 2007 and leverage has likely increased since the peak of the crisis, crossasset portfolios could be a potential for a contagion risk, which can be amplified further by the net short volatility base of the hedge fund community. The collapse of LTCM in September 1998 is a case in point. At that time a seemingly small shock in the EM (Russian default) resulted in severe global market volatility due to fire sales of an over-leveraged hedge fund community. The calm period preceding the events of 1998 was characterized by a similar pattern of rising cross-asset correlations and declining volatilities.
Heady stuff, eh?
Related links:
The Great Vega Short - Artemis Capital
More on the literal Bernanke put – FT Alphaville
Volatility as the new Black-Scholes – FT Alphaville
CDS options market multiplies alongside questions - FT Alphaville