What Factor Investing Can and Cannot Do
In a sense, all equity investing is about factor investing. We can trace this all the way back to Sharpe’s Single Index Model of 1963, and his Capital Asset Pricing Model (CAPM) of 1964. In the CAPM, the factor which explains, and predicts stock price returns, is the market portfolio. Next, the identification of a further 2 factors, which along with the market portfolio, form the Fama and French 3-Factor Model of 1992. In this model, the additional two factors are Value, or the return on stocks with High Book-to-Market Minus Low Book-to-Market (HML), and Size, which is the return on stocks which are Small Minus Large (SML). A fourth factor, Momentum (MOM), or Up Minus Down (UMD), was identified in Carhart’s 4 Factor Model of 1997. Momentum is the returns on winners less the losers.
Hence, when we now talk about funds and ETFs tracking the market index, as well as a focus on value stocks, or small caps, we are basically talking about the factors which guide the construction of the portfolio. But what we as investors want to know, is really what factor investing can and cannot do for us.
Factor Investing, and what it can and cannot do
Making Sense of the Factor Zoo
While you might reasonably assume that 4 factors in constructing portfolios is enough, computing power in the last couple of decades has sparked a search for even more factors to explain and predict stock returns, resulting in hundreds of such factors, or what is known as the “factor zoo”. In a way, it makes sense to search for more and more factors, as these are the “building blocks” of stock returns. More recently, Jensen, Kelly and Pedersen (2023) show that the more than the 200 factors which they test can essentially fit into 13 categories, which really simplifies any discussion on factor investing:
The 13 Categories of Factors
Source : Jensen, Kelly and Pedersen (2023)
While the categorization of factors used to explain and predict stock returns shown above is very useful, what’s also very interesting is that these factors do not always work as effectively in every single stock market! For example, while the Size factor (one of the original factors) seems useful in emerging markets, it is the one of the worst performing factors in the US markets! And the Accruals and Short Term reversal factors do not seem to work well outside of the US. We’ll come back to this later.
What Can Factor Investing Do?
Let’s take a look at what factor investing can do for the investor. Factor investing has often been sold as a method for generating alpha for the portfolio, on top of any beta risk premium the investor is earning through exposure to the market portfolio. In most research on factors, we have the results of the total return on a portfolio attributable to each of the factors. But what does this actually do for the investor (apart from understanding where the return came from)?
A new study by Baltussen, Swinkels, van Vliet and van Vliet (2023) addresses this. Instead of looking at returns to each factor over a long period, as most other academic studies do, they look at the returns to each factor in periods of deflation, low inflation, moderate inflation, and high inflation. These results are in the table below:
Factor Investing Returns by Inflation Regime
In the table above, note that the Value factor is HML and the Momentum factor is UMD as before. the newly introduced factors are Low Risk, which is the return on low beta stocks minus high beta stocks, and Quality, which is a blend of the Profitability factor and the Investment factor (CMA), i.e. returns on Conservative, low investment stocks minus Aggressive, high investment stocks. MFE denotes a multi-factor strategy, investing in each of the four factors equally.
One result which jumps out from the table above is how strong equity returns are when inflation is low, between 0% to 4%. While the data for the table is from many years, spanning low, moderate and high inflation most equally, in the last 30 years, inflation has almost always been moderate between 0% to 4%. Hence, the strong returns to equity portfolios observed in recent decades. Which over the long span of the history for equities investing may well be an anomaly!
The second takeaway from the table above is how constant the returns to the factor portfolio are. Regardless of the inflation regime, whether deflation, low, moderate or high inflation, all the factors experience similar returns. The multi factor strategy is even more stable as a result! Which leads to the idea that the use of factor investing can add a significant amount of alpha returns to an equities portfolio, with little additional risk in returns overall.
What Factor Investing Cannot Do
So, what is there not to like about factor investing? It appears that what factor investing can do for an investor is to provide additional, stable returns to a portfolio regardless of the inflationary (and interest rate) regime we may find ourselves in. And this idea is becoming popular! Just take a look at the relevant pages from BlackRock, JP Morgan, Fidelity and Vanguard, and there is a whole list of factor investing ETFs and funds available for investors.
But this is to take the results above on face value, with no regard to how these factor strategies are actually implementable for investors. And there is some evidence that the stable alpha returns from factor investing are not so easily achievable (see here). Why? There are several reasons:
1. Factor Portfolio Construction
To achieve those juicy factor investing returns shown above, the factor strategies are implemented by constructing a long-short portfolio, with zero dollars invested. For example, for a Value factor, the strategy calls for buying stocks with high Book-to-Market values, and selling those with low Book-to-Market values. Which is the meaning of High minus Low essentially. The funds obtained from short selling are used to buy the stocks which are held.
This, without doubt, is a highly risky approach to portfolio construction, and may face an unending series of margin calls. As a result, factor portfolios are usually not constructed in this way, but only as a partial long-short portfolio, e.g. holding 130% of the portfolio value as long positions, and another 30% as short positions. Clearly, this approach will reduce the potential gains from pure factor investing by a long short!
Even if it were possible to construct a pure long-short factor investing portfolio, the returns from it will fall short of the theoretical maximum possible. Why? Because there are a lot of costs involved in doing so. Firstly, shorting stocks require the short seller to borrow the stocks shorted, and there is a cost to it. If the entire value of a portfolio were in short positions, this is a high cost indeed!
Secondly, there are costs involved in rebalancing the portfolio. Unlike equity portfolios which track the market capitalisation weighted market portfolio, factor strategies are not self-rebalancing! In a market cap-weighted portfolio, when the price of a stock goes up, its market cap also increases, as does its weight in the portfolio. But that is exactly what market cap means, and the portfolio is automatically rebalanced. And vice versa for when the stock price falls. The only rebalancing required is when there are new stocks entering the portfolio, or old stocks leaving it.
In a factor portfolio, there is no automatic rebalancing. What might have been a long position may need to be sold to rebalance the portfolio, or worse, shorted instead. And vice versa. This means that on rebalancing date, the total turnover in the portfolio may be up to 400%! That is if, all the long positions are sold, and then shorted, and all the short positions are covered, and then held.
So what is it that factor investing cannot do? In short, factor investing cannot give us the full benefit of the factor returns yet!
Conclusion: What factor investing can and cannot do
Factor investing is an idea which has been around for quite a while, and has become mainstream now through implementation in funds and ETFs. While factor investing has been touted as a technique for increasing investment returns by providing additional alpha returns, in practice it is hard to realise all of its potential, due to costs of implementation.
More concretely, the costs of shorting stocks, and rebalancing make it difficult to fully implement factor investing at the moment. And as a result, only a fraction of the benefits touted for it can be realised. Perhaps in the not too distant future, factor investing will become truly mainstream when techniques are available to reduce these costs of implementation for investors.