No flash in the pan
Factor investing strategies have gained significant traction in recent years as investors seek new ways to coax value out of a generally low-yield environment. Globally, in excess of US$1 trillion is now explicitly invested in a factor-based way.1
Yet for all its recent growth, factor investing is not new. Single-factor models focused on a stock’s beta, or sensitivity to the movement of a benchmark or the overall market, emerged in the 1960s. More recently, after a study of the performance of the government of Norway’s pension fund in the wake of the global financial crisis, investors began to understand that a major chunk of excess returns could be attributed to a combination of factors such as value (i.e. fundamentals), momentum (recent price trends) and size, even if they were not explicitly using the factor-investing method.
Choosing factors is fairly straightforward as there are a limited number that have been identified as reliable sources of risk-adjusted excess returns. For instance, in the US, value stocks -- that is, companies that appear undervalued versus their fundamentals -- generated 14.9% in annual returns from 1980 through 2017, 2.9 percentage points more than the overall market.2 As more asset managers become convinced of the benefits of the approach, factor investing-focused products have been on the upswing. Nearly 400 single and multi-factor ETFs were launched in 2017 alone.3
Factor investing’s popularity has given rise to concerns that certain factors such as low volatility (that is, shares that tend to swing less wildly than the overall market) have become too popular to ensure future performance. Yet fears of an overcrowded factor trade may fail to take into account the nature of premiums which, after all, are the compensation investors receive for taking on relatively higher-risk stocks. If risk premiums disappear, investors would likely lose their appetite for the relevant assets and pile back into more stable shares – meaning the premiums would soon be re-established. For this reason, it’s unlikely that risk premiums will simply be arbitraged away.
At the same time, factor investing should never be viewed as a consistent generator of outperformance – especially when it’s conducted passively. There are periods when certain factors will lag the broader market. A recent study looked at four factors over a period of more than three decades (1980-2014) and found that there is no one optimal factor-based portfolio, since the risk premiums associated with different factors vary widely over time.4
To take one example, the size factor-investing strategy is based on the widely held view that small cap equities offer a risk premium over the overall market. Our research into US small cap stocks over an extended period (1926-2016) found they offered an impressive annualized excess return of 1.94% - but this came with significant periods of underperformance, such as 1983-1999, when returns were mired 71% under the benchmark.
There is growing evidence that the performance of a factor-based strategy is not always replicable, and mainly driven by the choice of factors in a portfolio rather than any unique qualities. These strategies are also constrained by a lack of diversification resulting from the overlapping of superficially distinct investment styles, and vulnerability to premium-free macro risks. These issues can impede a portfolio’s performance, and add to the argument for active management that draws on a wide range of factors and considers the risk profiles and fundamentals of individual companies, to lock in diversification and mitigate exposure to unproductive risks for the long term.
