Below is a chart of sector correlations versus the SPY going back to 2011. You can see how they rise during periods of S&P pullbacks/drawdowns. During this most recent pullback in which we are about 4.2% off the highs, correlations have started to rise. Since the start of 2013 when this V-bottom market of shallow corrections really kicked into gear, the average pullback in the S&P 500 is 4.5%. Considering we are 4.2% off the highs with rising correlations and a ton of oversold indicators, we give a higher probability to a potential bounce. So unless we get a bigger flush lower in prices, correlations are likely to stay in this range of .7 to .8 before correcting lower.
A recent note from the morning briefing by Nicholas Colas, the chief market strategist at Convergex, sums up the current environment:
"Markets don’t melt down because a few sectors or securities move lower; they drop like a stone because everything declines at once – essentially a price correlation of 100%. Bull markets, by contrast, tend to show declining correlations as varying industry and individual stock fundamentals c reate differentiated asset prices. On the plus side, that’s exactly what U.S. industry group equity price
correlations have done since 2011, when average S&P sector correlations to the index as a whole were +95%.
This month that average is 82%. Now, the bad news: the long run average since the end of Financial Crisis is 84%. Simply put, we haven’t really broken free of the macro “Risk on, risk off” gravitational pull of systematic worries (read “Greek Debt Crisis”) or central bank watching (read “Fed lift off” and/or “ECB QE”). Should capital market volatility continue to increase – the CBOE VIX Index is up from 12.1 to 19.7 in just 10 trading days – correlations will likely increase as well. That will make “Safe” parking lots for capital very hard to find."
He goes on to list a number of keen observations:
"1. Asset price
correlations for the 10 industries of the S&P 500 have been declining since
late 2011, when they reached +95%. That’s good news, as it allows
active investors to benefit more from stock and sector picking and gives
passive investors a smoother and less volatile ride.
2. The bad news is
that average sector correlations are up over the last month, from 77.6% to
82.4%. We’ve tracked these numbers for years – since the Financial
Crisis, when they skyrocketed – and the correlation between the CBOE VIX and
sector correlations is 67%, a remarkably high reading. Simply put, when
volatility spikes correlations tighten up very quickly.
3. The second bad
news “Foot” to fall is that average sector correlations since 2009 come in at
83.7%, or right on top of this month’s 82.4%. We have not, in other words,
made any real progress in pushing sectors as disparate as Health Care and
Financials to trade more on fundamentals than larger market-wide trends.
4. As market
volatility picks up – the CBOE VIX Index is up from 12.1 to 19.7 in just 10
trading days – stock correlations are already at the same elevated levels as
the market has exhibited for years. That’s a problem for two
reasons. First, it means it will be as hard to diversify an equity
portfolio from incremental volatility as it has been the last +5 years.
Which is to say “Very hard indeed”. Second, it also signals that equity
markets are still in the sway of macro drivers rather than industry and stock fundamentals.
Yes, given the headlines over everything from Greece to Federal Reserve policy,
perhaps that is understandable. Even as it is unwelcomed."
It will be important in coming weeks to observe how correlations react as the market carves its path. Will it be more of the same and we're reaching an inflection point in both correlations and volatility? Or will individual sectors begin to behave a bit more independently when volatility flares up?
It will be important in coming weeks to observe how correlations react as the market carves its path. Will it be more of the same and we're reaching an inflection point in both correlations and volatility? Or will individual sectors begin to behave a bit more independently when volatility flares up?
