Backgammon as a Model for Stock Investing


Libya and Japan dominate stock market returns in March

On February 18th David Edwards of our firm appeared on Bloomberg Radio's "The Hays Advantage" with Kathleen Hays. On that date, US stocks peaked at the highest levels of 2011, exactly doubling the 13 year low of March 9th, 2011. Edwards' lead off comment was that he thought "the stock market would fall 10% in the next 6 months, but he wouldn't change his strategy, because he anticipated a 20% rally on the other side." We live in an age of instant gratification, because US stocks fell 7.1% over the next 18 trading days. As of Friday's close, US stocks are down 4.8% from the high for the year, but are still up 2.1% for the year. Since specific news about corporations is light during March as companies enter the quiet period ahead of April earnings reports, investors are primarily reacting (and over-reacting) to civil war in Libya, which is disrupting the oil markets, and to the aftermath of the earthquake and tsunami in Japan.

The situation in Libya evolved out of an incident in Tunisia in December 17th during which a street vendor burned himself to death to protest police harassment of his livelihood of selling vegetables from a wheelbarrow. Somehow this act of suicide morphed into widespread civil disobedience that resulted in the overthrow of the Tunisian president within a month. Protest movements rapidly spread through the Arab world, with the president of Egypt as the next "domino" to fall, and the "leader of Libya" Muammar Gaddafi as the next likely victim. The Iranian government foments revolution through Shia proxies in Saudi Arabia, Bahrain, Lebanon and Yemen, while simultaneously suppressing a domestic revolution. Just as the fall of the Berlin Wall symbolized the "tipping point" beyond which the Soviet Union collapsed from its own internal contradictions, we may look back one day to that poor street vendor as the tipping point where 500 million people moved from the 19th century world to the 21st century world. But to predict that historic transition in December 2010 would be like predicting a hurricane from the flap of a butterfly's wings.

The situation in Japan is equally random. Of course it's well known that earthquakes and tsunamis occur often in that part of the world, but being able to forecast the timing of such geological events is decades in the future. The Japanese people build for earthquakes and drill for evacuating in the event of tsunamis. Compared to the 2004 tsunami in the Indian Ocean, which killed 200,000 people, mercifully the mortality from the Japanese quake will be in the 10,000 person range. We can extrapolate from the 1995 Kyoto quake, which killed 3,000 and cost $100 billion that this quake will cost $250-500 billion, which is not trivial. The 9/11 attacks in the United States killed 3,000 and cost $50 billion. Was there permanent damage to the global GDP as a result of either incident? No, because in a world of 6.5 billion people and $45 trillion in wealth (real estate plus stock markets) deaths in the thousands and costs in the billions are rounding errors in the world balance sheet.

Backgammon as a model for stock investing

So if we couldn't predict either the Libyan civil war or the Japan earthquake, how did we "know" that the two year rally in US stocks would falter? Quick answer is that we didn't "know" anything. But we do have a good sense of when stocks are fairly or over-valued, and therefore vulnerable to surprises.

Consider the game of backgammon, which comprises a board, 15 pieces per side, and dice. For a given configuration of the pieces and a given roll of the dice, there is only 1, or possibly 2 or 3 optimal moves that a player can make. Backgammon is a far simpler game than chess or even checkers. It is so simple that if two experienced players played 100 times, you would expect each to win about 50 games. So why play at all? The doubling cube makes all the difference. At any time either player can offer a "double" to his opponent, which is to say double the value of a win. A double is offered when one player perceives that he or she is more likely to win. Most times, it's obvious to the opponent, who concedes the point and a new game ensues. But occasionally the opponent disagrees with the evaluation of the board and accepts the double. A lucky streak may ensue, shifting advantage to the opponent, who may offer a double back. With each exchange, the value of the win doubles in the sequence 2, 4, 8, 16, 32, 64. What distinguishes a champion from amateur player is the ability to wield the doubling cube. A champion may offer a double if the probability of winning is only 5% higher than the probability of losing. Tournament play is decided by the first player to reach a certain number of points, which may take 5-20 games to achieve. The doubling cube can maximize the value of a win, and minimize the value of a loss. Amateurs, by the way, are easily seduced into accepting doubles they shouldn't take.

Stock investing has so many more dimensions than game playing (whether backgammon, poker, chess, or checkers.) No one can quantify outcomes in terms of decimal probabilities. We employ a graphical model where we assign "stoplight" colors to the inputs into stock market performance. For example, right now we color US Fed policy Green, inflation Green, world commodity prices Red, world geopolitics Red, US economic reports Yellow etc. If we see a preponderance of Green, then we invest more aggressively; more defensively if we see Yellows and Reds. In simplest form, our color map at present looks like:

So hard to be enthusiastic about stock investing, even though the picture shows less red than a month ago.

What can we do to protect ourselves?

One of our clients called this week and asked if we ever buy puts to protect our stock portfolios. No, because from long experience we find that puts are cheap when you don't need them and expensive when you do need them, leading to the experience we call "hedged against profit." We don't try to leap in and out of the markets either. We guarantee you that anyone who has tried to game the 1-2% daily swing in US stocks in the last three weeks has generally "bought high and sold low." Similarly, anyone who jumped out of the markets last June-August while the Eurozone debt crisis unfolded missed a 30% rally through year end.

Nor do we want to get in the habit of short term trading given that the majority of our clients are individual and families who benefit from us postponing the realization of long term capital gains, not generating short term profits. But as we have mentioned in recent months, we have added more dividend producing stocks such utilities, REITS and telecomms, and added healthcare and consumer staples that are less vulnerable to stock market swings. In some cases clients have added funds to their portfolios, which we will leave in cash until we feel more positive about the outlook. We'll wait until the board looks more favorable before we offer our opponent, the stock market, a double.


Already forgotten, the March US employment made incremental improvements for this first net year over year gains in jobs in about three years. Our forecast for US stocks remains at 8% for 2011. We felt no compulsion to raise our estimate in February when US stocks were already up 7% in 6 weeks, and we feel no compulsion to lower our estimate in light of geo-political developments. Starting the second week of April, we'll get another round of earnings reports which will probably confirm our thesis that US corporations are in good shape in the balance sheet and on the income statement.