One of the top performing companies for the past fifty-plus years has been Berkshire Hathaway (stock symbol BRKA), managed the entire time by Chairman and CEO Warren Buffett.  BRKA is a holding company with wholly owned stakes in operating companies (for example, See’s Candies) and stock investments in other public companies (for example, Wells Fargo and Apple).  The compounded annual growth rate of BRKA stock since 1966 as been 21%.  This means an $1 investment in 1966 (in Warren’s case $5 million) will have grown to $20,000 today.  By way of comparison, $1 invested in the S&P500 stock index would have increased to $140 over the same period.

Finance professors have had a hard time explaining the BRKA performance.  Under the Efficient Market Hypothesis (EMH), it is impossible to outperform the overall market over an extended period of time.  This is so because all available information is rapidly captured in stock prices.  This means that companies with excellent prospects are priced up so that a new investor can only expect a market return.  Similarly, downtrodden companies are priced down with the same result.  Unless an investor has non-public information, sustained out-performance is not possible.

So, how to explain fifty-plus years of out-performance?  For many years, the standard explanation was luck.  Suppose there were one million monkeys repeatedly flipping a coin.  By the laws of statistics, it is likely that at least one monkey would get 20 heads in a row.  Warren is that monkey.  Unfortunately for the argument, it was first made about 30 years ago and yet the monkey has continued to come up with heads (beating the overall market) year after year.

Recently, a new explanation has been proposed.  Through massive data mining, finance professors have found that Warren’s stocks had greater expected return than the overall market.  What?  Does that mean the academics now agree that Warren is a genius?  No.  The academics have found (through data mining) that certain classes of stocks have greater expected rates of return (and greater risk as well).  These classes include so-called “value” stocks and “quality” stocks.  Value stocks are stocks with high ratios of earnings, cash flow, dividends or book value to price.  Quality stocks have greater profitability than the average company, steadier growth and more conservative balance sheets.

All this time, Warren has been buying high quality stocks at low prices.  Combining this clever factor strategy with modest leverage (through his insurance companies, Warren is able to generate cash for investing at very low financing cost), finance professors have now been able to explain 50 years of 21% growth.  See, it was as easy as that.

Actually, I’m being a little facetious.  The academics didn’t begin with Warren’s portfolio.  Instead, the research design was, first, search for factors that predict historical returns and second, test to see if the predictions hold outside the original sample period.  It took decades of research to identify a set of factors that appear to be predictive (value, quality, etc.).  The recent step has been to comb through the BRKA portfolio and realize that Warren apparently independently discovered these factors several decades ago.

The academics acknowledge that it was clever of Warren to have identified and taken advantage of these factor opportunities long before they were discovered by the professors.  But they are excited that they have been able to show that BRKA’s performance does not violate the EMH.  Not surprisingly, there have been quite a few investment management companies started by finance MBAs and PhDs that attempt to achieve out-sized returns by utilizing factor investing.  Perhaps the most notable of these companies are Dimensional Funds Advisors (DFA) and AQR Capital Management (AQR), both founded by finance graduates of the University of Chicago.  These companies build highly diversified portfolios that capture one or more of the predictive factors.

Since there are now competing portfolios that mimic the BRKA portfolio, is there any reason to prefer BRKA?  One positive is that Warren is still in charge; he may still be capable of identifying investment themes well before empirical research catches on.  On the flip side, he is 88 years old today and may not have another fifty-year run.  Another positive is fees.  For a retail investor to enjoy the benefits of DFA funds, she must pay the DFA fees plus the fees of a personal financial advisor (DFA only distributes its funds through financial advisors).  The total of these fees is probably around 150 basis points or 1.5% of assets, per year.  Meanwhile, owning shares in BRKA means owning shares in a diversified portfolio and paying zero annual fees.  Yes, you must pay a commission to purchase your shares, but this represents a very small fraction of one basis point, and is paid only once.  This is an enormous advantage in favor of BRKA.  Finally, don’t forget about the low-cost leverage provided by BRKA’s insurance businesses.

All in all, it looks to me like BRKA is likely to continue to be a winner compared to the academics’ portfolios.  EMH supporters would say that these benefits are built into the price of the stock.  Perhaps they are correct.  The total market capitalization of BRKA is $500 billion and the book value of equity at the most recent reporting date was $375 billion.  You might conclude that BRKA is overvalued by 33%, but you’d probably be wrong because the equity number reflects the purchase price of wholly owned acquisitions, not the current market value.  While it is difficult to estimate the size of this difference, it does not appear that BRKA is significantly over-valued.