It was ten short years ago, in July 2008, that giant government mortgage agencies Fannie Mae and Freddie Mac failed. The US Treasury bailed them out, and their directors were rewarded with bonuses. Earlier that year, Treasury forced JP Morgan Chase to buy Bear Stearns, and later Washington Mutual. JP Morgan was also rewarded, with billions in fines. Later in the year, auction rate securities markets froze and money market funds were in jeopardy.
While the causes 2008’s financial meltdown are suppressed by state-run media, its effects were borne by everyone not in the good graces of government. The banking crisis that began in 2007 led to a collapse in stock prices, home values, and interest rates, crippling the retirement dreams of many millions of Americans. While almost everything was losing value, retirement plan investment managers claimed to be helpless to do anything about it. After all, the meltdown could not have been anticipated and its effects were not controllable, right?
As the dust cleared in late 2008, Black Swan Theory became a popular theme in investment policy meetings and newsletters. The mystical Black Swan is a metaphor for extreme outlier events that are unpredictable, have major consequences, and are rationalized after the fact. One rationale was that portfolio diversification did not work because everything dropped sharply at the same time. But was 2008’s perfect storm impossible to anticipate? Did diversification really fail?
Black Swan Hunting
In April 2007, professor and statistician Nassim Taleb published an influential book titled The Black Swan: The Impact Of The Highly Improbable. And his message is clear: authoritarian macroeconomics experts are not experts, and they don’t know it. Their tendency is to assign simplistic explanations for outlier events. Serious analysis of the problem is what Taleb described as “stemming from the use of degenerate metaprobability.” In plain English, he explains it as “What happens, happens.”
In 1999 another statistician and investment manager built a system that anticipates the possibility of Black Swan events – it does not predict them. And it calculates their possible effects. David Loeper’s Monte Carlo engine is based on the capital market assumptions (CMAs) of Modern Portfolio Theory, and he uses them to calculate the likelihood of positive and negative extreme events. The complex system known as the capital asset pricing model (CAPM) balances them with all other potential outcomes within a normal statistical distribution. His overriding premise is that the future is uncertain. As Loeper explains about the nature of capital market returns,
Being excessively conservative in assumptions does nothing other than needlessly sacrifice an investor’s life. The data exists to discover this if you examine it closely. If you have bad assumptions, the allocation will not behave as anticipated. Well reasoned assumptions should have modeled the extremes that result in such ‘asymmetric correlation’ markets.
In other words, the possibility of something like 2008 should have been anticipated, he did, and the personal sacrifices mitigated.
Asymmetric Correlation Ate My Homework
In plain English, asymmetric correlation means that your team didn’t play right. But in reality it was their own game plan. The diversification plays were based on flawed assumptions. The diversifying players such as hedge funds, real estate, and managed futures were expensive and unproven draft picks. Or as Loeper explains “they had insufficient data to draw valid conclusions.” The opposing team was the nature of capital markets, aka the reality of uncertain complex systems, coached by human history on planet earth.
In Modern Portfolio Theory, each asset class has two potential assumption errors – median return and risk. And not only were the assumptions for the diversifiers wrong, the addition of each one multiplies the potential for errors. This is because their correlation to each other must be considered as another capital market assumption. What happened in 2008 was that the diversifier asset classes became positively correlated with the risk assets they were intended to hedge, meaning they all lost value together. As a result, retirement plan investors were exposed to a great deal of needless sacrifice.
For example, in 2008 domestic stocks lost 37%. A portfolio consisting of just four asset classes (domestic stocks, foreign stocks, US government bonds, and cash), with 60% of that in stocks, lost about 18%. Yet many “more diversified portfolios,” also with 60% in stocks, were down more than twice as much. And besides the multitude of statistical assumption errors, there was another one – market strategists had an aversion to US government bonds.
Of Strategists and Optimizers
According to Loeper, “The purpose of fixed income in an asset allocation is to lower the overall risk of equity markets, particularly in shock environments.” Only US government bonds do this well, and there is a vast historical data set behind their CMAs to prove it. Newer bond asset classes like foreign, high yield, emerging market, and inflation protected have no such reliability. However many portfolio strategists do not use a suitable US government bond allocation to diversify risk assets. This is because of perceived inflation risk or other economic forecasts. And neither do many mathematical models known as mean variance optimizers (MVOs).
The similarity between strategists and optimizers is that they are engaged in predicting the future, which adds potential assumption errors. While strategists admit they couldn’t anticipate or manage the events of 2008, MVOs are a different story. To their analysts, all they needed to do was adjust the assumption inputs of the newer asset classes. Heck, those data sets were insufficient anyway, just change them to support their use as a tool for diversification. The problem is that the optimizer might select a portfolio dominated by these alternative asset classes. Ask Yale University’s endowment how that worked out.
“I Don’t Want My Money to Work Hard. I Want it to Relax.”
Jerry Seinfeld is on to something here. It took a lot of work to get things so wrong, and it was expensive. Those who religiously follow government policy couldn’t foresee the disastrous consequences of government policy. Those who place their faith in unreliable data sets built on self-fulfilling prophecies were also blindsided. Fortunately there is a proven alternative to the stress of beating the market – Moneyball. Billy Bean, GM of the Oakland A’s, showed the baseball world how to win the division while minimizing the cost per win. He replaced talent scouts with reliable data. For retirement plan investors, it’s about winning the future while minimizing risk, and replacing forecasters with reliable data.
The moral is to let the price mechanism of free markets do the work, and let the retirement nest egg relax. Minimizing errors includes integrating new information into existing data sets. The Center for Research in Security Prices did this in 2007. Yet it took 2008 for anyone not in government to learn the consequences of avoiding reality. The uncertainty of complex systems like capital markets and its price information must be reckoned with. After all, “the data exists to discover this if you examine it closely.”