Recently many investors and market participants have been perplexed as the VIX and volatility in general have decreased – even given the backdrop of rising political uncertainty and geopolitical risk. Here we analyse the factors currently keeping volatility low and use history to explore why the VIX may not be the best measure of risk and uncertainty.
Measures of market volatility, particularly the VIX Index, which measures the implied volatility of S&P500 index options, have increasingly been viewed as measures of broad market risk. Indeed the VIX has been widely referred to by the broader market as the “fear index”. Recently many investors and market participants have been perplexed as the VIX and volatility in general have decreased – even given the backdrop of rising political uncertainty and geopolitical risk. Indeed despite major unforeseen political events (Brexit, Trump election) and rising geopolitical tensions (North Korea, Russia, Syria, etc.) volatility has remained remarkably low.
Will it last? Here we analyse the factors currently keeping volatility low and use history to explore why the VIX may not be the best measure of risk and uncertainty.
The “fear index” and volatility, what is it?
To begin, it is important to distinguish between the distinct (but entwined) concepts of implied and realised volatility.
- Realised volatility is a backwards looking measure of the variability of a price path.
- Implied volatility, as measured by the popular VIX index, is derived from traded option derivatives on the Chicago Board Options Exchange, and is itself a traded commodity through futures derivatives.
- Buying or selling VIX futures allows investors to take a view on the future realised volatility; as such it is a ‘forward looking’ measure of volatility as it is a measure of the options market’s expectations.
So despite being commonly referred to as the fear gauge or index, the VIX is not a measure of market fear, but simply a measure of expected volatility over the next 30 days, as implied by index options.
Realised volatility and implied volatility are different and distinct concepts; nonetheless the chart below shows a very apparent link between the two.
Chart 1: S&P 500 30 day annualised volatility and VIX Index.
Source: Bloomberg and CFSGAM. Data to 17 July 2017.
The best guess for volatility tomorrow is volatility today
The times in which implied volatility spikes often occur after realised volatilities have moved. As the VIX only measures expected volatility over the next 30 days, it does not necessarily reflect long-term risks and indeed longer term risks can steadily rise, but not be reflected in the VIX at all. Generally events must be chaotic for the VIX to spike – Asia Crisis, LTCM Crisis, Worldcom Crisis, GFC, Greece etc. In all these examples, prior to the events unfolding, implied volatilities were at low levels compared to history. It was not until after the market started reacting (that is, realised volatilities were increasing) that implied volatilities likewise increased.
Two other related measures and drivers of volatility include:
- Cross sectional volatility – represents the dispersion of individual stock returns within the index, generally this will rise during times of volatility (if the market falls we would expect defensive stocks to fall significantly less than growth stocks increasing the dispersion of returns).
- Correlations – usually measured as average pairwise correlations, measure the average correlation between any two stocks in the index. Higher correlations suggests that returns are being driven by a common factor and usually coincide with periods of higher volatility.
Both of these measures have fallen to near historic lows (as shown in charts 2 and 3) and much of the recent decline in aggregate market volatility seems to have been due to declining correlations. Again, this doesn’t really speak to the longer term risk, correlations can (and will) rise sharply as macro influences become more significant or a correction occurs, so perhaps this is contributing to the artificially low short-term risk outlook.
Charts 2 and 3: Correlations and cross sectional volatility near historical lows
Source: Factset and CFSGAM calculations. Data to 14 July 2017.
Knowns and Unknowns
When considering the wedge between high uncertainty and low implied volatility it is also important to recognise the fundamental differences between risk, uncertainty and volatility.
- Risks are known unknowns, they have an unknown outcome but generally a known underlying outcome distribution.
- Uncertainty represents unknown unknowns, an unknown outcome from an unknown underlying distribution.
- Volatility reflects expectations around financial outcomes, or specifically dispersion of returns for a given security or index.
The nature of political uncertainty means that investors in aggregate will struggle to find a price for such complex non-linear scenarios given the number of variables and difficultly in forecasting such events. It is natural that they will not be priced into implied volatility until after the fact, given the unknown distribution from which they are drawn.
As the chart below illustrates, the historical link between uncertainty and volatility appears to have broken down in the last 12-24 months. This likely reflects the difficulty in not only predicting the uncertain events but also their outcomes; how do investors put a price on Brexit or what policies Trump may (or may not) implement?
Chart 4: Policy uncertainty has increased while volatility has fallen to decade lows
Source: Bloomberg and CFSGAM calculation. Data to 18 July 2017.
As mentioned above, just because volatility (implied or realised) is low does not mean risk is. Indeed, to quote Nassim Taleb (author of Black Swan and scholar on the subject of tail risk) “don’t confuse a lack of volatility with stability, ever.” Low volatility can create an illusion of current stability, but unless it is actually predictive of future stability, it is merely an illusion.
Why is volatility so low?
Many factors have combined to create the current low volatility environment, such as low correlations, stability of earnings, macro factors such as central bank policy and QE programs, structural factors such as the rise of passive investing and increasing capital requirements.
As mentioned earlier, low pairwise correlations are likely one of the drivers of low volatility. Generally low correlations mean that returns are being driven by stock specific factors, however this does not appear to be the case today. Currently, correlations are low due to large sector and style rotations driven by quant flows, monetary policy and political developments (e.g. growth-value, low volatility-high volatility, ‘Trump trade’ and its unwind).
Earnings stability has also played a role, with earnings risks at a minimum currently. We have seen reporting seasons worldwide deliver positive surprises and very few negative surprises, when placed in a historic context, dampening stock specific volatility. As shown in the chart below not only has EPS been relatively stable but volatility in EPS is also at historically low levels. It is important to note that the improvement and stability of earnings is at least in part due to central bank policy, particularly low and stable interest rates, which have reduced the corporate debt burden.
Chart 7: S&P 500 EPS has been relatively stable compared to history
Source: Bloomberg and CFSGAM calculations. As at 13 July 2017.
Further supporting the low volatility environment has been the reduction of fundamental risks, with company balance sheets still conservative, particularly at the top of the index (the heavyweights), which make up a significant proportion of the VIX. Again, central bank policy has been a significant driver of balance sheet improvement not only lowering interest rates but also increasing demand for higher yielding corporate debt. One way of looking at this is the relatively low level of negative ratings actions compared to history. As shown below the number of corporates put on negative watch by the major credit rating agencies is at the lowest level in 10 years, supporting lower volatility and reduced credit surprises.
Chart 8: Number of credit watch downgrades by Moody’s and S&P (US only)
Source: Bloomberg and CFSGAM calculations. 13 July 2017.
On the macro side, the lower standard deviation of economic forecasts and monetary policy expectations have contributed to lower volatility environment. Central banks have offered unprecedented levels of policy guidance to reduce potential surprises, while economic growth has generally been relatively stable (but subdued) for the last few years. Not only has growth been stable but, as shown below, the dispersion of forecasts has also been relatively low.
Chart 9: Actual and forecast global growth has been remarkably stable
Source: Bloomberg and CFSGAM calculations. As at 13 July 2017.
The persistence of low interest rates across the yield curve in most developed markets has also acted to lower expected returns and drive asset allocation increasingly towards higher risk asset classes - such as equities and HY debt. The shift in allocation has lowered the realised volatility of these riskier asset classes. Further, lower interest rates have, to a large degree, reduced fundamental balance sheet risks as the burden of corporate debt repayments has decreased.
The massive growth in central bank balance sheets since the GFC has also played a role. Successive rounds of QE from major central banks have provided an abundance of capital flow into markets, dampening risk premiums (including term premiums in fixed income markets – see our recent blog post), expanding market multiples and reducing market volatility. While It appears that these flows have now peaked they should still remain supportive in the medium term despite expected US Federal Reserve’s balance sheet roll off with any unwind of QE from the European Central Bank likely to be gradual and no sign of the Bank of Japan stopping their purchases anytime soon.
Chart 10: Central bank net purchases
Source: Bloomberg, CFSGAM Research
These flows have also led to a persistently high level of excess cash held by non-bank investors globally, providing an effective backstop for both equities and bonds, reducing both downside volatility and its persistence. In addition to the move up the risk spectrum, many investors are also holding large cash balances and are looking for opportunities to enter both equity and fixed income markets at the right levels, but hard data on this is relatively scarce. The chart below illustrates one way of looking at the trend, through balances in cash accounts at the New York Stock Exchange (NYSE), which are near all-time highs.
Chart 11: NYSE Free Credit Balances in Cash Accounts near all-time highs
Source: Bloomberg, data as at May 2017.
Structural changes in markets have also played a role in lower volatility since the GFC. The reduction in bank leverage driven by increasing capital requirements and tighter regulation combined with the ban on proprietary trading has led to a reduction in liquidity and volatility, particularly in equity and bond markets.
Further, the continued growth of passive investing, with a doubling in US quantitative and passive investing strategies to 40% of equity assets over the last 10 years, has contributed to less active stock trading. Recent estimates from JP Morgan suggest that only about ~10% of trading volumes now originate from fundamental discretionary traders.
Chart 12: Share of passive investing (including ETFs) in US equity universe continues to increase
Source: JP Morgan Flows & Liquidity, ICI.
Given all these factors, particularly central bank policy, the current low volatility environment is hardly surprising. However, while some of the factors identified are likely to persist, such as the structural growth in passive investing, other factors are likely to change in the medium term, such as central bank flows and low interest rates.
Will it last?
While the current low level of volatility appears to be justified given the reasons outlined above, the important questions for investors are; will this environment last and if not when will it change? If history is any guide, periods of low volatility can persist for some time, but will eventually come to an end - this period is likely no different.
With the hawkish turn in global central bank commentary and the unwind of the US Federal Reserve’s US$4.5trn balance sheet on the horizon, along with a tapering of the European Central Bank and Bank of England’s QE program likely early next year, it seems that at least some of the factors that have driven this low volatility environment will be fading in the next year.
The most important thing to remember is that low volatility does not mean low risk and is not necessarily indicative of stability. The concerns that have driven political uncertainty higher largely represent severe tail risks – war with North Korea, a confrontation with Russia, a big Trump mistake. These tail risks are low probability, high impact events, which aren’t actually well captured by the VIX and market expectations of 30 day volatility.