Market Volatility Patterns: How Investors Can Filter Out NoiseVW Staff
Market Volatility Patterns: How Investors Can Filter Out Noise by Frances Hudson, Standard Life Investments
Volatility in markets seems to follow patterns. With careful analysis, investors can filter out noise, providing useful insights over a variety of time horizons
After a hectic start to 2016 in a range of financial markets, investors could be forgiven for being bemused. Global equity markets, oil (Brent) and wheat performed round trips of 24%, 16% and 14% respectively, only to end the first quarter little changed. This article considers proximate causes and behavioral aspects of why markets are more or less volatile in different time periods. Of particular concern is the unwelcome return of ‘risk on-risk off’ in explaining recent market movements. In examining the evidence, it would be helpful to establish if there is something more systematic going on; for example, developing trends or seasonal patterns. We can also try to identify any unifying factors, whether causal or in terms of behavioral responses, say, in reaction to central bank signals.
A new seasonality
For six of the last 10 years equity markets have wobbled in the first quarter and, while the variety of ex-post explanations suggest otherwise, it is tempting to herald a new seasonal pattern. Investor shorthand could take the form of ‘New Year jitters’ superseding the old ‘January effect’ (see Chart 1) and providing a contrarian check on the ‘Santa Claus rally’. The January effect, characterized by markets rallying at the start of the year, has actually been less common since 2002. As far as causes are concerned, could it be that market seasonality is echoing economic or even meteorological seasonality?
Where fundamental news on companies is concerned, the reporting season kicks off in mid-January. Ahead of earnings announcements, news-flow about companies tends to be meager and event driven. Logically, economic data which predominantly concerns prior periods should be a secondary rather than a primary driver for forward-looking markets. Forecasts and surveys, many of which have a spotty track record and provenance, are frequently accorded undue attention by media and commentators. We do use survey data in our research process but generally find that it is more useful as a contrarian indicator. As a side issue, seasonality in economic data has itself become a contentious subject, with researchers at Federal Reserve banks in San Francisco and New York supporting different interpretations of the extent to which seasonal and weather-related factors impact growth numbers.
Signals and noise
Is there a technical explanation for risky assets selling off at the start of the year? Liquidity, trading volumes and information flows are generally lower around the end of calendar years, providing sparse inputs for price discovery. Some market participants may be absent. Others, including algorithmic traders, may be waiting on the sidelines pending an improvement in signal quality. Strong price discovery is associated with a variety of motivations, time-frames, valuation approaches and opinions being represented in the markets. Against a quiet backdrop, new news may be accorded disproportionate weight. In addition, despite the lack of fundamental developments, investors are being bombarded by information. In this case, more does not equate to better quality. Delivery methods, such as 24-hour news and social media, further amplify noise and distort signals.
An intuitive explanation would be that market reversals and exaggerated moves are more likely when low volumes detract from confirming or undermining recent trends, meaning that investor conviction is low. At the same time, on a behavioral level, the availability heuristic tells us that we are prone to use immediate examples to support investment decisions rather than expend effort in carrying out a more thorough analysis or applying appropriate filters to the noise. Holidays in many markets may provide investors with an opportunity to think and reassess their portfolios and positions at the very time when the quality of the readily available information is at low ebb. This is not dissimilar to the rationale for ‘sell in May and go away’, one seasonal pattern that is still evident in some markets.
Popular explanations for early calendar market volatility latterly have included a host of ‘known unknowns’ in geopolitics. In 2014 and 2015, unusually severe US winter weather disrupted trading and affected market sentiment, while in the mild winter of 2016 another episode of China economic weakness seemed to spark things off. The common ground shared by these elements is that they are tail risks.
China has been a prime tail risk as far as markets and economies are concerned for some time. The IMF’s recent Global Financial Stability Report included a chapter on financial spillovers from emerging markets, which concludes, unsurprisingly, that as these countries become more integrated into the global economy and financial system, spillovers to equities and exchange rates in developed and other emerging markets have risen. Indeed, the IMF estimates that spillovers from shocks in emerging markets now explain more than a third of the variation in asset returns and, as amplification channels including fund flows and corporate borrowing build, can be a source of systemic risk. It is natural then, that news about China’s growth will have a global impact, and explains how China’s policy response to volatility in its own equity market or currency, which might seem to be primarily a local concern, impacted global markets in August 2015 and January 2016.
Central bank and economic drivers
Volatility tends to regimes. In general, periods on economic growth, low inflation and low interest rates are associated with subdued market volatility, albeit with sudden jumps. The size of central bank balance sheets may also be a factor. Over longer-term demographic horizons, that environment seems to be appropriately priced in. The volatility spikes that disrupt progress towards the ‘grey horizon’ may be linked to specific events or regulation.
In the shorter term, sensitivity analysis provides clues on what is driving markets, with the caveats that it may not be possible to attribute how much of a move is driven by a particular factor in a given instance, and when discussing markets it is not always possible to determine which market is the tail and which is the dog. Here, a sense check on the impact of policy announcements provides a degree of insight. If markets are back to ‘risk on – risk off’ mode, then central bank policies – everyone following the same signals – are also a likely unifying candidate. Whether it is by inflating asset prices through QE or spurring lending through dabbling in negative interest rates or just providing gnomic guidance, central bank pronouncements have clearly influenced market outcomes.
When considering investor positions it is important to remember that markets are discounting mechanisms. Hence, in the run up to widely anticipated policy changes, such as the Federal Reserve raising interest rates in December 2015, speculators (non-commercial traders) in foreign exchange markets are active as they reposition according to what they think is priced in. At the time of the announcement, and during the following days, positioning is more neutral and less subject to change, so markets move less.
Tools and coping mechanisms
Big data approaches that use news analysis proffer a means of taking advantage of the incessant information flows, but signals here are likely to be fairly short-lived. Another way to filter out some of the noise is to pay attention to what investors are doing rather than what they may be saying or how commentators are interpreting markets. Besides basic price, volume and fund flow data, the most widely recognized measures of volatility in markets include VIX, VDAX and MOVE indices (see Chart 2). These estimate the future volatility of US and German equity indices, and US Treasuries respectively, based on weighted averages of implied option volatilities. However, as this type of volatility index has become more tradable and traded, the signals become less useful.
Elsewhere, the Commodities Futures Trading Commission (CFTC) collects information on virtually all futures contracts traded on US exchanges. These can be analyzed via a host of associated technical measures, such as option skews, open interest, futures premia and commitment of traders (see Chart 3). Familiarity with the dataset and the quality of the signals in specific underlying assets means that by looking at extreme speculative positions, we can gauge vulnerability to market setbacks and unwinding of positions. Similarly, signs of futures premia can provide an indication of the extent to which equity market futures have deviated from fair value; sustained deviation would lead us to expect a reversal. We incorporate these measures in our behavioral analysis of markets.
In the medium term, the investment process can pay more attention to valuations or trends in corporate earnings. An obvious coping strategy for longer-term investors, when faced with market volatility that appears unconnected with fundamental changes, is to extend their time horizons and focus on better-quality signals. These could include measures of sustainability and attention to governance of companies or countries.