Find method in madness

Last Updated: Sun, May 06, 2012 07:51 hrs

In 'The Drunkard's Walk', a book examining randomness, physicist Leonard Mlodinow referred to a couple of instances of amazing success in market predictions. In end-January 1978, Leonard Koppett publicly claimed that he had a way to predict the Dow's annual direction (up or down) and released the first of 19 annual predictions. Koppett's system got the direction correct every year from 1978-1989. He made a mistake in 1990. Then he had another winning streak between 1991-98. A strike rate of 20/21 is amazing by any standards.

The second instance Mlodinow mentioned was the performance of the Legg Mason Value Trust Fund, which was solely managed by Bill Miller. Miller beat his benchmark, the S&P 500 Index, for 15 years running from 1991-2005. Somebody crunched the relevant data for all US funds over a 40-year period (1965-2005) and calculated the chances of Miller doing this was one in 3.2 lakh. Mlodinow asserted that it was actually closer to 3 out of 4!

Koppett's system was simple. He was a sports journalist playing a joke. The Super Bowl, the biggest event in US football, is a final match between the winners of two leagues (now 'conferences' after a merger). It's held in late January.

Koppett assigned 'up' to one team and 'down' to the other and his annual prediction was based on which team won. Most people, including Koppett, would agree that he was simply very lucky. It is worth noting that he could equally, have been very unlucky, if he had reversed assignations.

Unlike Koppett, Miller had obvious stock-picking expertise. He used value investing methods to pick his stocks.

There were hundreds, if not thousands of others, using similar methods. What made Miller's results so outstanding? Mlodinow explains how luck played a large part in Miller's excessive success.

Over that 40-year span, some 6,000-odd fund managers worked in the US market for spans of 15 years or more. The odds on a specific manager beating the market 15 years running were indeed in the range of 1 in 3.2 lakh. However, the chance that at least one random manager in that large pool would beat the market for 15 years running was as high as 75 per cent, according to Mlodinow's calculations. Between 10 -20 per cent of fund managers beat the market in a given year.

Say, about 1,000 managers beat the market in year one. And, 500 of those first year winners also beat the market in year two. In year three, some 250 of them manage a hat-trick. Meanwhile, another 500 have beaten the market two years running, and so on. Sooner or later, given the timespan and the numbers, the chances were very good somebody would hit a 15-year streak. Miller was the lucky guy, with all due respect to his stock-picking skills.

Statistical logic can be an argument in favour of passive investing. If you guess market direction, you may be very lucky, or very unlucky. If you pick stocks, you may beat the market, or not. Most fund managers and stock-pickers don't beat the market consistently. Even those who do, may just be luckier, rather than more skillful.

Warren Buffett offered a counter-argument.

The world's most successful investor agreed luck could be a factor in extraordinary success. But he said, also examine the methods of people who are consistently lucky. If you find something common to their methods, it may not be all luck.

Buffett himself has a extraordinary record of beating the market over 60 years. As he pointed out, several other chelas of his Guru, Benjamin Graham had also been successful over long periods. It may be noted that Buffett and other Graham-trained investors have also had bad streaks.

Another guru from an entirely different school, Ed Seykota, has similar extraordinary streaks of beating the market over the last 40 years. Seykota is a trend-following technical analyst. He's uninterested in the fundamentals of anything he trades. If his computer programs suggest a trade, he takes it and holds it, until he gets an exit signal. Seykota and other trend-following technicians also have their bad streaks.

What does one make of this? The investor can derive a common sense strategy. If you want the market return without any trouble, stick to passive index investing. If you want to beat the market, you need either extraordinary luck, or extraordinary skill coupled to at least some luck.

If you know how to generate extraordinary luck, don't waste your time reading this. Developing extraordinary skills is possible but it's a full-time job. If you can't spare the time or lack the commitment, stick to passive investing.

More from Sify: