Correlations in equity markets have fallen dramatically since the beginning of 2017. We look at the drivers of this trend and the implications for investors.
5 minute read
For much of the past decade, markets frequently moved up or down in response to external factors such as political risk or shifts in monetary policy, a phenomenon that has come to be known as ‘risk-on, risk-off’. Investors moored in safe havens when the macroeconomic environment worsened, then paddled back towards riskier assets when the clouds cleared.
But in equity markets at least, this dynamic has begun to change. Since the beginning of 2017, the correlation between stocks within major indices has dramatically broken down. Equity performance is increasingly a reflection of corporate strengths and weaknesses rather than extrinsic circumstances, a development that has big implications for investors.
“Equity correlations were strong for a number of years, favouring strategies based on simple market exposure. But the decline in correlation raises questions over the merit of such approaches,” says Ahmed Behdenna, senior multi-asset strategist at Aviva Investors. “Fundamentals are becoming more important.”
As of October 11, the 10 major sectors in the S&P 500 were showing, on average, just a 41 per cent correlation with the wider index, down from 75 per cent a year ago, according to Datatrek, a research provider.1 The implied correlation across the S&P 500 fell almost 70 per cent between January and November 2 (see chart).
So what has caused the collapse in correlations? There is no definitive answer. Some analysts have cited the relatively benign macroeconomic environment. When global growth was anaemic, external factors such as political events had an outsized effect on investor sentiment; now that growth has picked up, fund managers have renewed their focus on the prospects of individual companies.
Central banks’ policy tightening may also be playing a role. The unprecedented monetary easing introduced to spur the economic recovery following the crisis drove interest rates to record lows and tended to push up stock prices regardless of fundamentals. Now that the Federal Reserve has started to withdraw stimulus and raise rates – albeit gradually – and other major central banks are preparing to follow suit, stocks are becoming more differentiated.
However, this monetary policy shift does not yet appear to have led to dispersion in credit – which remains fairly well correlated – suggesting there are other, more market-specific factors at work.
Ben Maynard, head of derivative strategy at Aviva Investors, points out that the fall in correlation across the S&P 500 accelerated during the so-called ‘Trump trade’ towards the end of last year. After Donald Trump’s victory in the US presidential election in November 2016, investors began to move into sectors they expected to benefit from tax reform, deregulation and greater infrastructure investment.
“Trump’s election caused a big move into inflation-dependent, pro-cyclical equity exposures,” says Maynard. “That in itself caused correlations to fall: when money flows from one group of stocks to another you get an increase in dispersion as a logical consequence.”
The president has so far failed to deliver on many of the policies that motivated the Trump trade, although the promise of tax reform – which could be passed later this year, pending legislative approval – continues to benefit sectors including technology and financials. Correlations briefly rose in early August, mostly due to heightened tensions on the Korean peninsula, but soon returned to multi-year lows.
Brexit and dispersion
Greater dispersion is also being observed in Europe and the UK. The decline in correlations in the FTSE 100 and FTSE 250 partly reflects the impact of Britain’s referendum on European Union membership, which led investors to take a more granular view of the market to try to sift the potential winners from the losers.
James Balfour, UK equity fund manager at Aviva Investors, points to a divergence in performance within sectors such as financials after the Brexit vote. “As with every stock there is no exact peer, but when you compare different financial institutions there has been a clear split since the decision.”
For example, as of November 6, HSBC shares were up 64 per cent since the referendum; by contrast Lloyds shares had risen only 19 per cent. This divergence is partly a function of the fall in the value of the pound against the dollar, which benefited companies with international revenues (such as HSBC) over domestically-focused firms (such as Lloyds), creating greater dispersion across both indices.
Accounting firm KPMG has split FTSE companies among two new indices to illustrate this growing dispersion. The KPMG UK50 index, comprised of companies that derive most of their revenues from the UK, such as banks, telecommunications and utilities firms, fell 5.5 per cent between the referendum and end-August 2017. The equivalent index of companies with mostly overseas revenues, notably energy and mining firms, rose more than 33 per cent over the same period.2
Active versus passive
As broad index correlations begin to break down, investors are putting greater scrutiny on corporate fundamentals, becoming more attentive to company balance sheets and more responsive to results announcements.
Balfour points to sharp falls in the share prices of British companies such as BT Group, which fell 22 per cent on January 24 following a profit warning, and technology company Imagination Technologies, a chip maker whose shares fell more than 60 per cent in April when it lost its contract as a supplier to US giant Apple. Research from Ernst & Young shows poorly-performing companies are now being punished more quickly and severely than was the case earlier in the year, a trend that is further contributing to the breakdown in correlations.3
As fundamentals become a bigger driver of performance, beta-driven strategies may prove less lucrative. The data supports this point: research from Morningstar finds a positive relationship between equity-market dispersion and benchmark-beating returns among active managers.4
Nevertheless, it would be somewhat simplistic to describe an uncorrelated market as necessarily a “stock-pickers’ nirvana”, as one commentator recently put it.5 After all, when markets are rising and falling in unison it is more likely that individual stocks are being mispriced. In theory, this should open up opportunities for active managers to notice discrepancies and take on contrarian bets, provided those fundamentals are reflected in the stock price further down the line, according to Giles Parkinson, global equities fund manager at Aviva Investors.
Parkinson cites the example of US high-street retail, a sector that remains fairly well correlated due to the impact of broad themes on investor sentiment. Fearful of the disruptive effect of e-commerce, many investors are shorting baskets of high-street retailers, giving active managers an opportunity to add to their exposure to individual firms that may be relatively insulated from the trend.
“Investors are shorting the US discretionary ETF [exchange traded fund], or bricks-and-mortar stores altogether, because they are worried about the impact of online retail. This means that the correlation between companies in this sector tends to increase. If you have conviction that a company is more resilient to the threat of online retail, this can present opportunities,” he says.
According to Parkinson, idiosyncratic factors are always likely to be the main drivers of stock performance over the long term. Over shorter time horizons, a less-correlated market should, on balance, be more likely to reward skilful stock pickers and punish those unable to distinguish the wheat from the chaff. And with lower market correlation, it should be easier to build more diversified – and therefore more resilient – equity portfolios, especially as intra-sector correlations have also begun to fall as well.
So how long is the current uncorrelated environment going to last? The last time stocks were this dispersed was during the run-up to the financial crisis in 2005-’07; in fact some figures show that correlation is now lower than at any time since 2001, during the Dot-com bubble. Recent academic research suggests a fall in correlations often portends market downturns. Does this mean a market correction is in the offing?
Maynard believes the current bull market has some way further to run, although he suggests investors keep monitoring broader macroeconomic signals as they refocus their attention on the minutiae of corporate reports. If signs of ebbing growth translate into lower forward-earnings estimates, volatility and stock correlations could quickly rise again.
“We expect correlations to stay low as long as volatility stays low, and that will remain the case until there is some kind of exogenous shock to the system – perhaps some kind of ‘Black Swan’ event – or expectations of global growth start to deteriorate, which would probably cause correlation and volatility to rise at the same time,” Maynard says.
CBOE S&P 500 Implied Correlation Index
1 'The moment stock pickers have been waiting for has finally arrived,’ CNBC, October 2017
3 Diverging Fortunes: Analysis of Profit Warnings, EY, Q3 2017
4 ‘The challenges active fund managers face,’ Morningstar, September 2017. The analysis focused on the relationship between active equity managers’ performance and cross-market dispersion since 2005, finding that dispersion accounted for 85 per cent of the variability in an individual manager’s success.
5 ‘Correlation crash clears the way for stockpickers,’ Financial Times, September 1, 2017
6 ‘Could the tide finally be turning for active vs passive?,’ Charles Schwab, August 28, 2017
7 Stock market dispersion, the business cycle and expected factor returns, EDHEC Business School, September 2015