I use investors’ love affair with Warren Buffett to illustrate our need for someone like that to exist in my behavioral-finance talks.
The fawning is so extreme that it is almost heresy to question the “data.” But the data, plain before our eyes, don’t stack up to even cursory scrutiny. “Facts” are quoted about the returns of Berkshire Hathaway
that are simply not true. Performance has been poor, but most alarmingly, getting worse. What is causing this? The answer: He says one thing but does another. A bad case of style drift.
We remember his tirade about derivatives in 2002. This followed his more or less accidental involvement when Berkshire made the $22 billion purchase of General Reinsurance Corp. in 1998 (the largest U.S. property and casualty reinsurer at the time). The General Reinsurance purchase also included 82% of the stock of Cologne Reinsurance, the oldest reinsurer in the world. General Reinsurance Securities, a subsidiary of General Reinsurance initiated in 1990, was a derivatives dealer tied to global financial markets. Unfortunately, this relationship had unpredictable consequences. Buffett wanted to sell the subsidiary, but he couldn’t find an agreeable buyer. So he decided to close it, which was easier said than done, as this decision required him to unwind the subsidiary’s derivative positions.
Since then, Buffett referred to derivatives as “financial weapons of mass destruction” due to the long and expensive process of closing down the General Reinsurance Securities positions. But filings show that derivatives became a major source of Berkshire’s performance in the years that followed.
He has often mentioned while describing his folksy investment philosophy how he doesn’t invest in companies he doesn’t understand, particularly in technology companies. He prefers companies that make goods that can be seen and understood. This failure to participate in the bull-market sector of a lifetime has been a drag on his performance. He has simply sidestepped the most profitable investment of the past four decades, preferring to stick with old technology.
As if this weren’t bad enough, he again decided to not do what he says and enter the technology heavily at its peak. You can see in the chart, above, how Berkshire made a significant bet in the technology sector in mid-2011 after having avoided it for a very long time. From 2006 to 2011, Berkshire’s technology exposure was effectively zero. But starting in 2011 it went up very dramatically. This bet has gone poorly. In fact, this sector is the principal contributor to Berkshire’s recent so-called Alpha Return decay.
Alpha Beta Works does excellent work in separating systemic performance from factor performance: how much of a fund or fund manager’s performance can be attributed to the market itself and how much to their genuine skill input. In the chart, the black line represents Berkshire’s long equity portfolio total return. Within this, the gray line is performance due to factors, and the blue area reflects Berkshire’s positive returns from stock selection, or Alpha Return. Since “alpha” and security selection performance are widely and often inconsistently used, Alpha Beta Works defines a rigorous metric of security selection performance as Alpha Return, which is performance net of all factor effects, or the return a portfolio would have generated if markets were flat.
From 2006 to 2013 Berkshire’s return from stock selection alone was +29.9%. But in 2014 Berkshire’s Alpha Return started to decline (thinning blue area). Between 2014 and 2015, Berkshire’s return from stock selection was -16.3%.
As reported by Alpha Beta Works, the investment in technology from 2011 has gone poorly. In fact, this sector is the principal contributor to Berkshire’s recent Alpha Return decay. Berkshire’s long equity portfolio suffered over a 12% loss from technology stock selection, peak-to-trough.
To quote Alpha Beta Works directly: “Warren Buffett has long been very public about his avoidance of technology investments, citing (his) reluctance to allocate capital to a business he cannot understand. So why was a manager with no record in, and known skepticism towards, the technology industry making large bets in it? In this light, the above data indicates worrying style drift. Perhaps this large technology sector bet was made by one of his likely successors, which raises an equally important question about manager succession and the style drift risk therein: Does a new portfolio manager demonstrate security-selection skill (positive Alpha Return) in general, and in the areas of growing allocation like technology? And what other changes should investors expect of (the) new leadership?
“Whether Berkshire’s recent negative Alpha Return is Buffett’s own error or commission or a byproduct of portfolio-manager transition is beyond the scope of this piece. Our focus is identifying active return, its sources (factor timing, stock selection), and their predictive value. All of the above lead to improved performance. In Berkshire’s case, our approach has identified that an established star manager shifted focus to an area previously avoided (technology) and that this shift reduced stock selection performance. Whatever the reason, the results are troublesome. Their early warnings are critical to investors’ and managers’ performance.”
Such analysis of risk and performance using holdings data and a predictive equity risk model provides “indicators that enhance manager/fund selection and future performance. It can also highlight emerging issues, such as style drift.”
Trevor Neil is managing director of U.K.-based Behavioural Economics and Technical Analysis Group (BETA Group). This first appeared as a blog post called “Berkshire Hathaway Style Drift Taking Its Toll” on the company’s website.