At the core of smart beta
Charlotte Moore attempts to pin down the characteristics at the heart of factor investing
At a glance
• Despite its mathematical roots, factor investing is a surprisingly nebulous concept engendering frequent discussion and debate.
• There are six core factors: value, momentum, quality, size, high dividend and low volatility, according to MSCI.
• To determine whether a factor is a viable investment strategy, investors need to determine if there is rationale for the factor, examine the historical evidence and be confident it will persist.
• Practical considerations and valuations are important considerations. Investors should be convinced that returns can still be generated despite market frictions, and strategies should not be bought at any cost.
If only investment theory was more like mathematics. Then a pension scheme would know exactly what concepts and variables a fund manager was referring to when discussing factor investing. But while this strategy has its roots in quantitative investing, its terminology is far from precise.
Despite the different interpretations and impassioned debates evoked by factor investing, some basic concepts can be prescribed. At its simplest, a factor is a particular characteristic of a group of stocks that helps to explain their return and risk profile.
Although several different factors can be identified, investors are only interested in those characteristics that give a particular group of assets a long-term risk premium. By selecting stocks with these attributes, investors hope to generate better returns than the market.
Academics laid much of the groundwork for factor investing, analysing a number of different investment characteristics and identifying those with the potential to generate long-term returns. Supply and demand for such products has grown in tandem. Developments in the asset management industry have made it possible to turn these ideas into cheap and transparent investment products.
Joop Huij, head of factor investing research at Robeco, says: “After the global financial crisis, investors became much more sceptical about active managers’ performance and fee structure.” This drove demand for straightforward index-based solutions.
The core investment factors
According to MSCI, there are six investment factors that generate long-term risk premia, although there is disagreement about the precise number:
• Value. These stocks have low prices relative to their fundamental valuation but outperform higher-valued stocks over time. For example, their stock price will be lower than their net asset value. Several different metrics can be used to find stocks with this characteristic, including price-to-book, price-to-earnings and various cash flow metrics.
• Small cap. Smaller companies generate higher returns over time than their larger-cap cousins. These stocks can be easily identified by looking at market capitalisations.
• Momentum. Stocks which performed well in the past will continue to outperform those that did not. Looking at the relative performance of stocks over different time periods highlights those with this factor.
• Low volatility. Shares with lower historical volatility will perform better than their more volatile counterparts over time. This can be identified by looking at the standard price deviation.
• High dividend. Those stocks with higher dividends outperform those with lower dividends. This is easily spotted using dividend yields.
• Quality. Companies with low debt and stable earnings growth outperform less-high-quality companies. Looking at return on equity, stability of earnings and balance sheet strength reveals these characteristics.
But investors should be aware that there are alternative views on the number of factors in the academic research. Practitioners also differ.
Phil Tindall, director of investment at Willis Towers Watson, says: “Quality and low volatility could well be slightly different expressions of the same factor, as there is considerable overlap between these characteristics.” Companies which have dependable earnings and cash flow could fall into either or both classifications.
While some are trying to narrow down the number of investment factors, others are trying to expand the range. Willem van Dommelen, head of multi-asset systematic strategies at NN Investment Partners, says: “The increasing popularity of factor investing has resulted in an explosion in the number of documented factors.”
This adds to the confusion for investors: which are valid investment factors and which spring from the imagination of an asset manager’s overzealous marketing department? A good dose of common sense can help to clarify the situation.
Tindall says: “Many of these seemingly new factors are merely a different flavour of a core investment factor.”
How to identify an investment factor
Investors can apply a simple set of criteria to determine whether a particular investment characteristic has the potential to generate long-term returns.
“For an investment characteristic to qualify as a factor, it needs to have a number of qualities,” says Remy Briand, global head of research at MSCI.
There should be a rational explanation for the existence of the premium. Briand says: “It could be a risk explanation.” For example, smaller companies are a riskier investment because they have a narrow business model and are more prone to going out of business if just one thing goes wrong. As it is a higher risk investment, investor should expect to earn greater returns.
“Or there could be a behavioural reason why these stocks outperform,” says Briand. Investors tend to choose more volatile stocks because the investment narrative of these companies is more compelling, but tend to disappoint over the longer term. “As a result, less-volatile stocks perform better over the longer term,” he adds.
Another explanation for the existence of an investment factor is constraints on investor behaviour and resultant market flows which shape the performance of some stocks relative to others.
Huij says: “For example, most managers are rewarded according to the size of their assets under management.” This form of incentive encourages managers to take on more risk, as outperformance will draw in more assets. Huij says: “This results in strong performance-chasing behaviour, which explains why high-momentum stocks tend to outperform.”
Not only does an investor need an explanation about why a particular asset should earn a premium but there should be empirical evidence for this investment factor. Briand says: “This where the academic literature is important to provide long-term evidence.”
For newer factors, however, there might not be sufficient academic evidence. Tindall says: “The investor, or their adviser, would need to carry out their own robust testing of a particular factor over a long period of time.”
Finally, the investor needs to make an assessment about why the factor will continue to exist. Briand says: “Investors need to be convinced this factor is persistent and will not be arbitraged away.”
This is especially true of those factors shaped by structural market imbalances (see Structural factors). Van Dommelen says: “These imbalances need to be properly understood to ensure they are not simply ephemeral trends.” The investor needs to be convinced the market participants who are being forced to act this way will not change their behaviour in the future.
Before choosing to invest in a factor, an institutional investor needs to satisfy all three criteria – that there is: a rationale for the factor; clear evidence that it exists; and a strong conviction that it will persist.
Briand says: “If an investor cannot satisfy these three criteria that is a clear signal that a particular characteristic is not a factor but merely a spurious explanation of some market element.”
Even if investors can satisfy those three criteria, they need to remember there is a considerable difference between the factors described in academic research and the reality of investing in actual financial markets.
Academic literature uses hypothetical portfolios which have little resemblance to actual funds. Huij says: “For example, researchers assume dividends are fully paid out and immediately re-invested; investors can trade the closed price and, critically, there are not trading frictions, taxes or costs.”
The reality, however, is different. There are trading frictions, taxes and costs. Dividend payments and re-investment are not instantaneous. Huij says: “These all have the potential to erode the returns on a strategy using a particular investment factor.”
Investors need to differentiate between hypothetical and actual returns of different investment factors. Huij says: “Incorporating all the realities of running an actual investment portfolio paints a very different picture.”
Huij adds: “For example, it’s difficult to generate the hypothetical returns from a small company portfolio as laid out in the academic literature.” Trading restrictions and costs erode a significant proportion of those returns.
Accounting for the realities of implementation, Huij says there are only four viable investment factors out of MSCI’s universe of six. “These are value, momentum, low volatility and quality.”
These practical considerations can also be applied to a new investment factor which appears on the investor’s radar. Huij says: “Once the realities of implementing an investment idea have been taken into account, often a newly identified investment factor no long seems viable.”
Van Dommelen concurs: “An investment factor has to do more than look good on paper: it needs to be translated into a profitable strategy.”
Valuation is still king
As with any other investment strategy, buying at the right valuation is an important consideration. Like other popular investment strategies, there are signs that factor investing has started to become a victim of its success, with valuations increasing sharply.
Rob Arnott, chairman and CEO at Research Affiliates, says: “The point of investing in a particular investment factor is to produce better returns than the market.” The old adage of buying an asset cheaply and selling when it is expensive is just as relevant for factor investing as it is for any other form of investment.
Before an institutional investor decides to allocate asset in an investment factor, it behoves them to examine the current valuation. Arnott says: “Is it realistic, then, to expect a factor to generate better returns than the market after it has already become expensive?”
If the investment factor has become too expensive, it would be unrealistic for investors to expect it to outperform the market. Instead, investors should focus their attention on those factors which are currently undervalued.
Arnott says: “At the moment, the cheapest factor available is value.” The value factor can also be implemented in an easy and cost-effective manner. And it has been shown to work. Tindall adds: “The value concept is as old as the hills, dating back to Benjamin Graham in the 1930s.”
“Value has been the worst-performing wfactor for the last decade,” adds Arnott. This would indicate the value investment factor is particularly good value at the moment.
Structural imbalances can occur across a broad range of different asset classes. Van Dommelen says: “Many passive commodity index trackers roll their forward contracts at a similar time, which creates an imbalance in demand for specific future contracts at a particular time.”
Regulation can also create opportunities. For example, before the financial crisis, when an insurance company wanted to dispose of a portfolio of bonds, an investment bank would purchase the entire portfolio. The bank would then slowly release that portfolio into the market, making a profit from its role as market maker.
Luc Dumontier, head of factor investing at La Française Investment Solutions, says: “But new regulatory capital requirements force the bank now to dispose of the portfolio rapidly.” This increased flow creates a supply and demand mismatch, causing the price of the bond to tumble.
Dumontier says: “The creditworthiness of the bonds, however, has not changed.” As result the equivalent credit-default swaps (CDSs) do not change in value. The difference between the two can be captured by buying both the bonds and CDSs. Dumontier says: “This is an example of structural investment factor which can be exploited by investors.”