As we pointed out on Friday, implied correlation closed at an all time high. While this phenomenon is likely the most concerning indicator of a wholesale market meltdown (not to mention that market neutral funds continue on their rapid progression to oblivion: for reference check HFRXEMN and HSKAX), more so than even the Hindenburg Omen (which however does make for a cool soundbite) as there are no endogenous safe-haven sectors within stocks (the safe haven is simply known as bonds, and stunningly high yield also applies even though HY, especially B2/B and lower, and stocks probably differ in some way, but we still haven't quite figured out what), the threat level will only be obvious in retrospect. The very curious topic has incentivized some in the sellside realm to present their own cautionary tales about the trade off and the cost-benefit analysis of a record high cross asset correlation. One among these is BNY's Nicholas Colas, who points out that due to the subliminal perception of record low stock dispersion (or liminal if such people read sites like Zero Hedge), investors have decided to diversify on their own not within stocks, but outside of stocks, the result being record inflows into bond funds (and outflows out of equities). His summary is very concerning: "Investors, even if they have not learned it formally, understand that diversification means lower correlations. As long as stocks, bonds, precious metals, and other assets all move in lock step, retail investors will most likely favor less risky assets." This statement captures the problem better than most: stocks, and their liquid, synthetic, nouveau-CDO cousins, the ETFs, continue to trade higher on ever greater vapors, as the underlying asset base is increasingly devoid of cash. And when the margin call-based liquidations set in, and they will, stocks will collapse more violently, and with far greater amplitude than they did on May 6 (incidentally a date which is an anniversary of the real Hindenburg disaster). Only in retrospect will the current record correlation levels be perceived for the loud alarm bell they truly are. But, courtesy of our idiot regulators, this will certainly not occur sooner, or before it is too late.
More from Colas on this striking phenomenon.
We are, for a variety of reasons, living in some alternative investment universe (call it “Bizarro Investment World,” if you are a Seinfeld fan) where, for the moment, correlations are very high for a variety of asset classes. We noted this phenomenon is last month’s analysis, and in this month’s update we find little diminution of the trend. In some areas it is actually accelerating. The following charts and tables have the data that support the following analysis:
- Among U.S. stock market sectors, correlations declined very modestly, but remain at near-record highs. Five of the ten industries we track still have correlations to the S&P 500 of at least 95%, and the lowest correlations are around 83% (Utilities and Consumer Staples). Four months ago the lowest correlations were 65-66% (Health Care and Utilities.
- Rising markets and lower expected volatilities (as measured by the CBOE Volatility Index, or VIX) do tend to push down the correlations among industry sectors. However, even last month’s strong rally in U.S. stocks did little to reduce how much sectors move in tandem with the market overall.
- High yield bonds reached a new high for correlation with U.S. stocks. In contrast, high grade corporate bonds remained one of the least correlated asset classes - a real bright spot in the analysis.
- Precious metals – we track gold and silver – both saw rising correlations to U.S. stocks in July. That is a reversal of a trend that started early in 2010, and price movements in precious metals have become more tied to equity priced over the past 60 days.
- Currencies bucked the trend to higher correlations, at least among the Euro, Aussie dollar and Yen. In fact, the much maligned Euro is at 8 month lows in terms of its relationship with U.S. stocks.
There is a pretty steady diet of explanations for increasingly correlated markets. Here are two of the more popular ones:
- One school paints the sector correlations as a function of High Frequency Trading (HFT). By most accounts this type of money management dominates the day-to-day trading of U.S. equities. One strategy – the arbitrage of Exchange Traded Funds (ETFs) to underlying stocks – could partially explain why stocks and sectors move closer to indices like the S&P 500. This is especially true, we suspect, in a rising market where the SPY ETF gets a significant amount of inflows because investors want an easy, low cost, way to play the rally.
- Another explanation is that macroeconomic drivers such as
- Federal Reserve monetary policy
- A rolling set of concerns about sovereign debt risk
- push money into and out of assets without much fine-tuning on considerations such as industry sectors or credit quality. If it is a “risk on” day, everything works. If not, then nothing goes up.
We are sure there are other explanations – these two just come up the most.
I would only point out that there is a real price to pay for such high correlations. This isn’t science fiction any more, it is science fact. The most dramatic development of the last 12 months is that retail investors have pulled cash out of U.S. equity mutual funds even though stocks have done well. Some has gone into stock ETFs, yes. Far more, however, has gone into bonds. Investors, even if they have not learned it formally, understand that diversification means lower correlations. As long as stocks, bonds, precious metals, and other assets all move in lock step, retail investors will most likely favor less risky assets.