Search Now

Recommendations

Thursday, May 10, 2007

Race to Beat the Market - Sanjeev Pandiya


I had argued earlier how risk is defined wrongly and that Beta is a flawed measure and “outperforming stocks don't really (outperform)”. What I implied is that Classical Finance takes these high-sounding, laudable objectives and produces good-looking, 'logical' (therefore rational) models that are supposed to achieve these objectives.

But long after the Nobel Prizes have been given away and people have gone home, it often turns out that the model sold to an unsuspecting public was a lemon. Last time, we talked about CAPM, but other big models like Dividend Irrelevance or Efficient Markets Hypothesis do not stand up to scrutiny either. This time, we will talk about “market outperformance”, a common measure by which money managers (particularly the Mutual Fund industry) measures and sells itself. Since I got almost no mail last time, I presume that either everybody agrees with me or more likely, nobody understood what I said. Which is just as well, because the strategy I laid out last time is delivering returns that no mutual fund manager can even dream of.

“Market underperformance' is seen by ordinary investors as a measure of risk. The other side of the coin, “market outperformance” is not seen as risk, but is clubbed as return. No money manager will treat 'statistical variance' of his performance as part of risk/ uncertainty. Tracking error measures the variability of a fund manager's performance with that of the major stock Index, say, the Sensex.

Beta here has no meaning. The Sensex return is like the cost of capital in the investor's mind. Both investor and money manager treat Beta as a measure of risk, when actually, it is merely an indicator of variability. Like I mentioned last time, risk is not the same as variability, which in turn, is not the same as volatility. You cannot be a good trader unless you understand the nuances of the difference between these terms.

Yet, investors routinely input Beta in calculating their cost of capital, allocating a higher cost of capital to stocks/ portfolios with higher variability, with strange results. No wonder DCF models fail to predict stock prices. Like Warren Buffet pointed out sarcastically, a stock cannot be 'riskier' at $40 than it was at $80. Classical Finance, which measures historical volatility and then equates it with risk, would tend to classify a stock as riskier at $40, just when it is cheaper. Value investors know that a cheaper stock gives higher returns, not necessarily with higher risk. Actually, the risk is lower.

However, the risk, as calculated by Classical Finance is higher, because it confuses risk with volatility/variability. But as I pointed out, this risk is different from real risk, which, as Buffet points out, is actually lower.

The resulting situation is almost comical. “Higher the risk, higher the return” is an old aphorism going around the markets. It seems logical, almost intuitively true………yet good value investors know that it is not so. The risk referred to here is actually not risk, but variability/ volatility. Value investors know that the highest returns actually come from 'low-risk' stocks. At this point, both variability and volatility are usually at historic lows. This is a typical aberration found commonly in markets…the Indian markets are no exception. Maybe the regular occurrence of such phenomena ensures that the “efficient markets hypothesis” is a failure.

Which investor, who manages his own money, ever compares his 'performance' against the Sensex? The actual cost of capital to an investor is his opportunity return, usually bank deposits. In other words, his real cost is inflation, not the pulls and pushes of the Sensex. This brings me back to the old argument in favour of absolute returns, rather than market outperformance.

Why is this subterfuge perpetuated? Because the mutual fund industry gets its fees from the size of funds managed, not from the returns generated. So they compete within the asset class, i.e. after the investor has put a certain share of his wallet into mutual funds, he diversifies his portfolio mindlessly, hoping that such diversification will somehow give him a better risk-adjusted return.

Regulators support this with flawed regulations. Fees based on returns are for hedge funds who need to focus on absolute return strategies in order to keep body and soul together. But the flawed logic of classical finance sees this as 'risky' and hence not for the small investor. So the poor guy lives with genuine risk, producing sub-inflationary returns on a mutual fund investment, while the smart, rich guys hire absolute return managers to take care of their portfolios.

Measurement of portfolio performance was a very good idea in its objective. It just went awry in its implementation. The application of various measurement techniques has vitiated the very purpose it was meant to achieve… the achievement of a proper, risk-adjusted return on the investor's capital.

It has perpetuated the idea that a money manager can be evaluated by a finite set of measurable, mathematically sound metrics that are standard across the entire set of managers. It assumes that the rules used to allocate your funds are repeatable, standardized and independent of their environment.

It tries to reduce to a logical evaluation, the study of process (of investment) that succeeds because of a very subjective skill. Just consider this: Your real cost in an investment is inflation and risk (i.e. variability of return and volatility of principal, not the Beta). Measuring investment performance against inflation ensures that you remain focused on absolute returns, while measuring risk ensures that you achieve genuine diversification, not the artificial, meaningless diversification achieved by trusting a larger number of strange 'mutually distrustful' faces.