This article is a guest contribution by Robert Horrocks, CIO, Matthews Asia.
Risk and uncertainty are part of everything we do as investors. Much energy goes into trying to understand these elements, and many a computer keypad has been worn out in attempts to write about them. There is a vast body of academic literature that discusses risk and uncertainty. Describing their role in finance has earned more than one Nobel Prize. And yet there remains some distance between what the academic theories define as risk and what the experience of the practitioner is, particularly when it comes to how we think about risk when managing Asian equity portfolios.
From an academic point of view, risk should be something that varies in proportion to returns, the riskier something is, the higher return we should get for investing in it. Risk should also be calculable, in order to know what the trade-off between risk and return is, we have to be able to measure it. Risk that cannot be calculated is called uncertainty. Risk is what you can try to manage; uncertainty is what you can never fully know or calculate. Branching out from this theory, academics tend to treat risk as the volatility of returns, how much an investment's prices "bounce around" its trend value. Simplistically, the more volatile the security, the higher return needed to justify buying it.
The only trouble is, it doesn't seem to work that way in reality. Several studies have shown that volatility does not have this suggested relationship with returns. In my own work in the Asian context, investors are rewarded with better returns for taking on volatility only up to a point. After that point, returns to the highest volatility stocks on average can be worse than those for stocks of average, and even lower, volatility. So, it pays to push the risk curve only so far, and that point is somewhere a little beyond average volatility. Nor does it seem that high beta portfolios generate, on average, better returns than low beta portfolios.1
Seeing this problem, academics and practitioners have added other risk factors to their models. These began with market sensitivity (beta), company size (market capitalization) and value (book-to-market), but the list of risks was progressively expanded to include factors as diverse as leverage, volatility, size, valuation, momentum, return on capital, growth and earnings sustainability. This seems sensible at first sight, but is not that intuitive because in many cases it is not clear that taking on extra risk (as one would commonly define it) yields better returns.
Investing in smaller companies may be more risky and, therefore, investors should get a premium. But value? Why should I get an excess return from buying cheap stocks? Surely, if it is a risk factor, I should get the excess return from the risk of buying expensive stocks. The same can be said of many "quality" measures. I have found in Asia that financially sound companies do not return, on average, less than financially weak ones. Of course! (You might think.) But that is a problem for the theory of risk, surely, most investors feel they should be rewarded for taking on more risk and not for playing it relatively safe. Many of the risk models out there are actually return models, they describe what drives stock prices, rather than where the dangers are. If you think that focusing on cheap, high quality, small companies is a good generator of returns, you should be maximizing this exposure, not minimizing it.
