back to news & insights


Sep 24, 2015

From the Archives: Optimizing the Chances of Long-Term Gains

Bill Miller

Over the course of his tenure as an investment professional, Bill Miller, CFA has written market commentaries, often providing insights into his unique perspective. Our “From the Archives” series highlights key concepts from these letters. This passage was excerpted from a commentary written April 26, 1991.

Adam Smith’s The Money Game was first published in 1967, at the height and end of the great bull market that began in the late 1940s.  It is an entertaining, sophisticated, and accurate look at the Wall Street environment at the peak of enthusiasm.  The Eudaemonic Pie, by Thomas Bass, which came out in 1986, is an account of a quirky group of physicists and computer scientists who tried to break the bank in Las Vegas by building into a shoe a computer that used statistical mechanics to predict the probable landing point of a roulette ball on the spinning wheel.

Early in The Money Game, Adam Smith quotes Keynes about the “game of professional investment” being boring unless one is possessed of “the gambling instinct; whilst he who has it must pay to this propensity the appropriate toll.”  It used to be that investment management consisted of buying and holding stocks and bonds of investment grade, as determined by some statistical rating service.  Performance was measured, if at all, over long periods of time, and little thought was given to predicting the vagaries of the market.

Investing was serious business, not a game.  “Playing the market” was used to describe those whose investment purposes were not entirely serious, were perhaps even a bit frivolous.  Dabbling in stocks was a pastime, not a profession for markets “players”.  The great stock market characters of the late nineteenth and early twentieth century – Jesse Livermore, “Bet a Million” Gates, Arthur Cutten, the Fisher Brothers – were the market’s equivalent of professional gamblers like Amarillo Slim and Johnny Moss.  There was a clear social and moral distinction between them and the stolid bankers who managed money.

That distinction became blurred, as Adam Smith recounts, in the late 1960s when the cult of performance first took hold.  Just as Nixon declared in 1971, “we are all Keynesians now”, portfolio managers are all performance driven now, since investable funds chase past performance records, and a portfolio manager without funds to manage is just expensive overhead.

As the players in the money game have shortened their time horizons due to the emphasis on short-term performance, their behavior has sometimes come to resemble that of the casino gambler more than that of a rational, long-term investor.  Getting the near-term direction of the market right, being in the stocks that are now outperforming and anticipating the next move are what investors seem to want from their advisors.  The Wall Street Journal recently reported that a major investment firm fired its strategist because he was too cautious in a rapidly rising market.  If the time horizon of strategy is now a few months, tactics must be a daily phenomenon.

The gamblers in Thomas Bass’ book were not strategists or money managers trying to predict the market, they were scientists trying to predict roulette.  Being quantitatively oriented, they knew that betting red or black had an expected loss of 5.3 cents for every dollar bet per spin of the wheel.  The longer one stays at the wheel, the greater the probability of total loss.  The optimal roulette strategy under normal odds for one who is actually trying to make money is to bet the entire roll on one spin.

The Eudaemons devised a methodology that optimized their chances of long-term gains.  Factoring in the tilt of the wheel, the speed of the ball, its angle of descent, and numerous other considerations, the computers concealed in their shoes were able to shift the odds from 5.3% against to 40% in their favor.

The success of their strategy was based on not betting on improbable outcomes.  The computer was able to eliminate about eight numbers per spin from the 36 a priori possibilities.  But the 40% advantage they had over the house only existed in the long run.  In the short run, statistical fluctuations could appear that would deplete or eliminate their capital.  Such fluctuations could also lead to large gains accruing to those with clearly sub-optimal or even irrational systems.  Success was virtually assured, but only over the long term, and only by making bets with a high probability of success.

The rewards of consistently beating the house at the casino or of regularly beating the market are substantial.  It is perhaps not surprising that some members of the Eudaemons have moved from the game of roulette to the money game.  Doyne Farmer, who led the group, now heads the theoretical division at Los Alamos National Laboratory.  He and a group of mathematicians, economists, and computer scientists have been trying to understand the market as a prelude to predicting it.  There is good news and bad news so far.  The good news is the market seems to exhibit certain patterns consistent with what is called deterministic chaos.  The bad news is the patterns are not predictable.

We have learned over the years that predicting the market is futile; understanding it is challenging enough.  Right now the players in the game are betting on an economic recovery that will begin sometime this summer or fall and carry into 1992.  Investors who were cautious or nervous last October are now increasingly bullish.  The percentage of investment advisors who are bearish, 23.1%, is the lowest level since September 1987.  At current prices, stocks appear to be discounting S&P 500 earnings of about $26.00, compared to 1991 estimates of $21.75.

If the recovery arrives on schedule and interest rates do not rise, the market would appear to be fairly priced at present levels.  Sustainable advances will likely depend on some combination of lower interest rates or better than expected earnings.  If the recovery is delayed, or rates move up, the market could give back some of the gains we saw this quarter.  Like the gambler who bets with the craps player because he’s on a roll, many investors are betting with the market because it’s on a roll.  Such a strategy is not likely to lead to any better results in the market than it does in the casino.

In any case, our long-term strategy will remain consistent.  We will not attempt to predict the market, except to note that the odds on its being higher in any 12 month period are about 66% since 1926.  In the past 50 years, the odds are better, nearly 75%.  The house advantage for investors in stocks has been 10.1% per year since 1926.  In order to maximize the probabilities of long-term success, we intend, like the Eudaemons, to avoid betting on improbable outcomes.  We will continue to look for sound companies whose shares can be purchased at discounts to what our analyses indicate they are worth, managed by people we can trust.

The views expressed are subject to change at any time. LMM disclaims any responsibility to update such views. The presentation should not be considered a recommendation to purchase or sell any security and should not be relied upon as investment advice. It should not be assumed that any purchase or sale decisions will be profitable or will equal the performance of any security mentioned. Past performance is no guarantee of future results, and there is no guarantee the views expressed will turn out to be correct.

©2015 LMM LLC. LMM LLC is owned by Bill Miller and Legg Mason, Inc.