I came across this interesting paper (which can be found here) the other day, regarding uncorrelated returns. The basic premise of the paper was that there is no such thing as uncorrelated assets. The author conveniently cherry picks the last 36 months to prove his point. Of course the last 36 months can easily be described as unique if not an outlier. Many have been quick to come to the conclusion that the last 36 months not only disprove the efficient market hypothesis, but also disprove the theory of uncorrelated assets. This is highly flawed in my opinion.
Let me begin to dissect this issue from the beginning (without getting bogged down in too much mundane theory). Anyone who is a regular reader has likely taken the time to read the “about us” section on the site. If so, you know that my investment theories aren’t just some cookie cutter “fill the Morningstar box” approach. I believe the efficient market hypothesis is one of the greatest tricks ever played on the investment community. Any market is nothing more than the summation of the decisions of its participants. Markets, by definition are highly complex dynamic systems that are susceptible to chaos. To assume that the summation of these decisions is somehow efficient would mean that the decision makers as a whole are efficient. While this might be true to some extent, human beings (and even the algorithms written by humans) are guaranteed to be inefficient decision makers in a chaotic system.
The investment world is the civilized version of natural selection. It cuts to the core of every emotion imaginable. When Joe Schmo goes to work for 25 years straight in an attempt to create a better life for his family and suddenly sees his life’s savings going down the tube because Lehman Bros went bankrupt you can’t possibly expect him to react rationally in such an environment. This is no different than the man whose family is attacked in the middle of the night. Do you expect that man to react rationally when everything he lives for is suddenly in harms way? Do human beings make rational and efficient decisions in chaotic scenarios? Even more important, will 1 million humans working in tandem make efficient decisions all within the same system? No, the majority of them will make highly inefficient decisions. “Mistakes” as we like to call them. We all make them.
If we have learned anything over the course of the greatest mean reversion in stock market history over the last 24 months it is that markets are HIGHLY inefficient. Why? Because the humans that write the algorithms are using flawed theories and the emotions upon which these trades are placed are not psychologically efficient. But does this mean the market is so inefficient that there are no negative correlations? This delves into a much deeper and more important question for this is where the holy grail of investing lies – the all important land of high risk adjusted returns.
David Swensen and the other endowment fund managers were famous for creating high risk adjusted returns using a portfolio of highly diversified uncorrelated assets. The theories and strategies appeared to work for years and then – poof – it just stopped working in 2008. What happened? Was it due to inefficient markets or are there simply no uncorrelated assets? The answer to this question is far more complex than I can answer in the short space here, but is likely simpler than some might assume. It is not that there are no uncorrelated assets, but simply that the correlation between assets are constantly changing.
Too many investors like to think that there is one holy grail approach to investing. Some believe it is buy & hold, others believe it is long only. The truth is, the investment environment is ever changing. Buy & hold will work great in some environments and will fail miserably in others. At the March bottom when I was telling people to buy stocks for the first time in 2009 everyone was saying buy & hold is dead. I said:
“Unlike the legions of investors who have proclaimed buy and hold dead, I actually believe it is more viable than ever right now”
Although I had been against buy and hold for several years the macro environment had changed so drastically that the period of early 2009 actually favored such a strategy.
The investment world is in many ways similar to a battlefield. No battle is ever the exact same. No investment environment is ever the exact same. A cookie cutter approach is guaranteed to get you killed. This is why the investment world requires a great amount of flexibility and unbiased thinking. This is one of the reasons why the comparisons to 1930 are ludicrous. This environment is its own unique environment. It is about as similar to 1930 as H1N1 is to the common flu. Fine for simple comparisons, but upon closer inspection – ENTIRELY different. Investors who believe they can apply one approach to all market cycles and expect to deliver above market returns are fooling themselves. Like a strain of the flu, you must attack that particular flu virus with the correct anti-virus. Likewise, each market requires a different battle plan.
Traditionally, the investment community has believed that real estate, private equity and hedge funds are uncorrelated to equity markets. Why they thought this would be true ad infinitum is beyond me. Markets are non-linear dynamical systems. They will never be the exact same in two different instances. Therefore, all assets are destined to operate differently in different instances.
This is the point in the argument where Nassim Taleb would say – “aha! black swans can’t be avoided”. But to assume that the black swan is unavoidable is incorrect in my opinion for there are always and will always be uncorrelated asset classes in any investment environment. Good risk managers know how to consistently maintain these high levels of non-correlation in all markets (I should add – good risk managers are a rare breed). Of course, the true holy grail is pinpointing those uncorrelated assets at the correct time, but that doesn’t mean they aren’t there.
As a risk manager I utilize two uncorrelated tools that are too often overlooked - cash and currencies. See, the investment world doesn’t like high cash levels because that means they don’t make any fees. In fact, they are paying YOU to hold cash. It is no different than the blackjack player who sits out most hands. Casinos hate that. Why? Because there is no money in a player that doesn’t play. Your broker thinks no differently of you. Hence, the reason why 95% of all investment banks and advisors maintained at least a high level of buy and hold ratings all last year. They want to keep you in the game. See the chart below and notice the consistent 5% sell ratings by analysts. Do you mean to tell me that during one of the greatest collapses in economic history the level of stocks rated “sell” never moved above 5%? That is an utter embarrassment for the entire Wall Street community. In my own business I rarely have more than 5% of the entire stock market on my BUY list.
The beauty of cash and currencies is that it always has a low level correlation to all assets. And in the zero-sum currency world, by definition there is ALWAYS something uncorrelated. I did not outperform the market by 60% last year because I am a genius. I did so because I moved into cash and uncorrelated currencies when the risks appeared skewed to the downside. In essence, I mucked my cards after an incredible series of face cards ran the deck and the odds changed.
In summary, there are always uncorrelated assets, but they are unlikely to maintain the same correlation throughout all market cycles due to the fact that all market cycles are different. Finding them is of course the key, but assuming they aren’t there is just as silly as assuming that the Wall Street banks are looking out for you by keeping you in the game all the time.
(Related topics – Money Management)