Active management and the predictability of markets #1

That it is possible to add substantially more value to customers' portfolios than is subtracted from them in fees is the basis of the entire active management industry. The debate between those who believe that markets are efficient and thus unpredictable and those who don't will continue to rage on, and active managers should seek to be part of it.

The battle lines in this debate, however, are drawn in a somewhat puzzling way. To the passive world, all those seeking to 'beat the market' are fools. Yet to the active world, the passive world's endeavours are far from futile. The apparent anomaly lies in differences of opinion with respect to the nature of the 'game' being played. 

To the passive world, investing is a game of luck, so by definition theirs is the only worthwhile approach and accordingly they must disparage all others. To the active world, on the other hand, investing is a game of skill, in which there is no 'house', only other players.

If an active manager can be better than three quarters of its competition, it will almost certainly generate returns, net of fees, well in excess of the relevant index or passive equivalent, and thus over time respectable in absolute terms. The challenge is thus to be more skilful than other active managers, then to clearly articulate the approach adopted to customers (success for an active manager relies as much on the power of persuasion as the power of prediction.)

As mentioned, the passive world has something to offer because index funds do what they say they will - track an index. Furthermore, there will always be fund investors who lack the time, tools or inclination to identify skilful active managers and thus for whom an index fund is the sensible option.

It is somewhat ironic however that the constituents of index funds, namely companies, are all trying to do one thing: beat each other. Few would question that this competition is anything other than a game of skill and hard work. The better companies get to the top and tend to stay there, whether in toothpaste, passenger jets, active management or passive management. Let's be clear, Vanguard is not seeking, like its products, to be average but to beat iShares.

It would be nice if we lived in a world in which all companies win. Until then, active managers must seek to beat their competition and in so doing give customers value for money. 

This post looks at the work of recent Economics Nobel Laureate Professor Eugene Fama, the "father of modern finance". Although it was Fama who first popularized the idea in 1970 that markets are efficient, its origin goes back as far as French mathematician Louis Bachelier and his 1900 PhD thesis The Theory of Speculation. Bachelier is credited with being the first to model mathematically the random process known as Brownian Motion, and to associate it with stock prices.

60 or so years later, and following a number of other empirical studies, Fama set out to test systematically whether there were identifiable patterns (non-randomness) in stock prices. In 1965 he reported that shorter term returns were somewhat predictable from previous returns — they had a tendency to move in the same direction — but that the relationship was quite weak. Later, Fama stated that these patterns were so faint that attempts to exploit them would be wiped out by trading costs. This no-arbitrage model formed the basis of the Efficient Market Hypothesis.

In the decades since Fama's early work, there have been countless other empirical studies published that looked for patterns in stock prices and markets. One might think that if stock prices are essentially unpredictable over short timeframes then they are even more unpredictable over longer ones. However, this is not the case. Indeed it was Fama himself who in 1977 showed that the short-term interest rate could be used to forecast the return on the stock market.

If prices follow a random walk, then their variance (the square of the standard deviation or volatility) over two year periods should be twice the variance over one year and so forth. In fact, for both stocks and bonds, this is not the case. However, unlike short-term returns which Fama had shown exhibit slight momentum tendencies, longer-term returns are mean reverting (variance over two years is less than variance over one year). 

To put this into English, if you toss a coin ten times, there is a certain probability that you will end up with five heads and five tails. However, if the coin is mean reverting, meaning that following a head the coin is more likely to land tails (and vice versa), the probability of ending up with five heads and five tails is higher.

In recent years more and more pattern within stock prices and markets has been revealed, both in short term and longer-term movements. Fama is now a consultant to Dimensional Fund Advisors, a firm that structures funds that seek to take advantage of longer-term patterns such as low price-to-book companies outperforming high price-to-­book companies. 

Mastering the short term is Jim Simons and the firm he founded, Renaissance Technologies. On a per employee basis, Renaissance has the third largest super computer in the world, one which is able to delve far far deeper into price movements than the devices that Fama used in the 1960s. The annualised returns over a number of years of his Renaissance Medallion Fund, now closed to outsiders, were spectacular.

Intuitively, it makes sense to me that the Efficient Market Hypothesis is bunkum. Whether within markets or outside them, all events are a function of what came before. Club hits ball, ball hits tree. The key to successful active investing must therefore be to try to understand and exploit that function. That some are more able to do this than others seems completely natural to me. Like I said, investing is a game of skill not luck.