Deviate at your own risk
The recent controversy about factor investing has probably not caused any distress to the investment industry, let alone the wider public, but it is a fundamental one.
Proponents of a purist approach argue that only factors as defined in the academic literature have been shown to provide systematic explanations of stock returns, and that deviating from those definitions is dangerous. Providers of factor analysis tools contend that those definitions do not reflect investment practice and that in reality practitioners use more nuanced definitions.
“History has shown that any manager with sufficient skill can identify patterns of stock behaviour over a protracted period”
Take the value factor as an example. The academic literature uses a stock’s price-to-book ratio as a proxy for value. In contrast, providers such as MSCI and Style Analytics use more complex definitions based on measures of company performance such as earnings, cash flows and enterprise value.
The debate matters because investors see factor investing as a tried-and-tested systematic investment strategy. They assume it exploits regular patterns of stock behaviour that can be observed through time. If a value factor exists, then it should be possible to define it, explain its economic basis and observe it over long periods of time in a sufficiently accurate manner.
History has shown that any manager with sufficient skill can identify patterns of stock behaviour over a protracted period. However, unless these can be defined accurately and rigorously tested, they cannot be called ‘factors’. Some call them ‘signals’ instead. In essence, factor investing is about replicable results rather than temporary signals.
It is easy to conclude that investors should rely on tools that are based on rigorous academic research. It is also true, however, that some definitions can change. Today, less quantifiable measures of value, such as a company’s brand or reputation, are increasingly important in defining a stock’s intrinsic value. It is difficult to argue price-to-book is still a good proxy for value. Yet the classic value factor remains the best explanation of systematic stock returns from an academic perspective.
Surely the best approach is to use whatever factor definition seems best based on the available evidence while remaining aware of the risks. Awareness by investors, and transparency by providers, are crucial in this area. Using non-standard definitions could give different results than expected, precisely because it has not been tested to the same extent.
But that may not be such a bad thing. After all, investing necessarily involves taking risks. As long as risks can be properly understood and quantified they should reward those who take them.
Carlo Svaluto Moreolo,
Senior Staff Writer