Last Updated on 27 January 2021 by F.R.Costa
The traditional way of splitting assets into asset classes, according to their key features, is inefficient. Instead, we should look into the hidden characteristics that drive returns. These are the risk factors an investor gets compensation for being exposed to, which are highly correlated with the phase of the business cycle. While a pure risk-factor portfolio is difficult to build, individual investors can easily build portfolios with proper tilts to the prevailing risk factors.
No 60/40 Stock-Bond Portfolio
The most common division of the market puts stocks at one side and bonds at the other side. The traditional view is that bonds are safer investments and do better during recessions while stocks provide a better performance during the recovery and expansion phases of the business cycle but come with a lot more risk.
While stocks and bonds do provide a simple approach to risk, investors usually go deeper in the division. There are domestic stocks, foreign stocks, high-yield bonds, emerging market equities, money market, futures, and many other classes of assets. Investors usually associate them with a level of risk and potential reward.
After allocating assets to their proper categories, investors must then set their own goals in terms of tolerated risk and desired reward. The higher the risk tolerance, the higher the proportion of stocks, high yield bonds, emerging market equities. The lower the risk tolerance, the higher the proportion of bonds in general, domestic stocks (instead of foreign stocks), money market instruments, and other assets considered safer.
So far so good. But, unfortunately, the financial crises of the last few decades, in particular since the 1990s, have shown that asset classes are highly correlated at times, providing poor diversification. When that is the case, the diversification based on asset classes will fall short of the aim of limiting risk. We must then find alternative ways of categorising the assets.
Identifying Risk Sources
The return on some asset is a compensation given to its owner for being exposed to certain risks. In order to maximise long-term returns while keeping risk at a minimum, an investor should correctly identify the risk sources driving those returns.
The finance literature has identified several risk sources since the introduction of the Capital Asset Pricing Model (CAPM), with the most relevant of all being just the market itself. When buying a stock or a bond, an investor is compensated for bearing market risk. Later, in the 1990s, Fama and French developed a 3-factor model, identifying another two risk sources: value and size. An investor is additionally compensated for holding low book-to-market value stocks and small stocks.
The literature on factors explaining returns is vast and open-ended. Academics define factors in a way that they can be used to explain returns and evaluate performance. The industry define factors in a way they can be turned into investable assets. One way or another, a risk factor should be an independent source of risk explaining the returns of an asset over the long-term.
If a risk factor drives returns over the long haul, then an individual investor should try to identify the key risk factors and match them to the business cycle. The next step would be to build a portfolio with the desired risk factors. The process isn’t much different from the classic asset class way of looking at investment. The main difference lies on the fact that risk factors are hidden characteristics that need a special lens to be identified. But, don’t worry, there are simple tools that can help here – accounting ratios.
The most widely used risk factors in the finance literature are: the market, value and size. Additionally, momentum has been added to many models and empirical tests. The industry has often considered quality and volatility as another two important factors. While there are other potential factors, I believe that the deeper we dig and split factors into new ones, the higher the correlation among them. We want factors that are as uncorrelated as possible and then I believe the pointed 6 factors are more than a good start for the sake of building a portfolio.
Let’s take a look into each factor:
- Market – In most circumstances, this factor explains the largest part of returns. The original CAPM is a single factor model, which dictates a linear relation between the return of a stock and the return of the market. But, research has consistently proved the CAPM incomplete. The market factor captures the exposure to macro events like the business cycle, interest rates, and government policy. To capture this factor, an investor just needs to invest in the whole market. Such task can be achieved by buying a large selection of stocks or a broad market index.
- Value – It captures the returns from underpriced stocks, those that are priced below fundamental value. In general, the value factor is captured by investing in stocks with low price multiples like price-to-book, price-to-earnings and so on. In the Fama and French world, this factor is captured by the book-to-market ratio, an equivalent to the price-to-book ratio. A portfolio filled with stocks with low price multiples would be tilted towards the value factor.
- Size – This factor captures the returns from small capitalisations, usually identified as growth stocks. In general, ranking stocks using market capitalisation is enough to build a portfolio exposed to the size factor. If an investor chooses the lowest quantile (stocks with smaller capitalisations), the portfolio is then tilted to the size factor.
- Momentum – When past returns are positive, sentiment among investors rises, which further pushes prices higher. When the economy is performing well and markets are rising, investors tend to follow the trend. This results in the consistent phenomenon reported by Jegadeesh and Titman that stocks that have outperformed the market in the recent past tend to exhibit strong returns going forward. This factor is captured by selecting stocks with high relative performance for the past 3, 6 or even 12 months.
- Quality – Companies with low debt, stable profits, strong corporate governance, and stable dividends are seen as quality companies. When the economy enters periods of turbulence, investors tend to flight-to-safety (flight-to-quality), which means looking for companies that provide better safety metrics. When investors start cutting their past lunar projections for growth stocks, the better established businesses tend to outperform. Some metrics used to capture this factor are a low debt-to-equity, a high return-on-assets, a record of dividend growth stability, low financial leverage, and other indicators of balance sheet strength.
- Volatility – In general, over the long haul, less volatile stocks tend to outperform highly volatile stocks. Investors often believe that just because a stock is highly volatile, it may reward them a big prize, but as far as the theory goes, it is often preferred to invest in less volatile stocks and use some leverage instead. High volatility is many times the result of idiosyncratic risk, which can be diversified away. A metric that can capture the volatility factor consists in ranking stocks by average daily volatility for the past 1 to 3 years and then select the lowest values.
Matching Factors With The Business Cycle
As of recently, factor investing and smart beta strategies have been growing in popularity. Investors may now easily adopt these strategies without much effort, as there is a large ETF offer proxying risk factors.
Some claim that factor investing outperforms the market. It may be true. But it is also true that factors go through prolonged periods of underperformance. While the global economy has been growing over time, there is a business cycle that repeatedly turns expansion into recession and recession into expansion. During each phase of the business cycle, the range of outperforming factors changes. A trend following strategy like momentum doesn’t work during a recession phase. A value strategy also doesn’t work well during an advanced stage of expansion.
Matching factors to the phase of the business cycle is key to avoid prolonged periods of serious underperfomance.
We can split the business cycle into four different phases and try to match them with prevailing risk factors, as follows:
- During the early contraction phase, investors realise that stocks are overvalued as growth projections turn more realistic, many times due to a deterioration of economic conditions. At this stage, investors discard momentum stocks and go for quality stocks. Strategies based on value, volatility and quality tend to outperform. The market as a whole, momentum and size should be avoided. Regarding the size factor, what happens is a reversion of it. Higher capitalisations are now safer than the smaller capitalisations.
- During the late contraction phase, value continues to outperform, but investors are now upgrading their growth projections. Quality may still work well, volatility starts underperforming and size is picked up again.
- During the early expansion phase, the market, value and size factors are usually the best. Momentum starts appearing and improves as a strategy when the trend solidifies. Quality and volatility are unfavourable at this point because they are defensive strategies.
- During the late expansion phase, stocks are highly valued but continue to perform well. The trend is very well established at this phase, which benefits momentum strategies the most. The broad market as a whole is a good performing factor as well. Value, volatility, quality and size usually underperform.
Building a Factor Portfolio
The simplest way of building a portfolio out of factors is to choose the prevailing factors at each time and build a portfolio with equal proportions from each. As an example, let’s assume we’re in the late contraction phase. According to what I stated above, during the late contraction phase, the prevailing factors are value, size, and quality. We can then allocate one third of the funds to each factor.
The next step consists in building the factors themselves. The lazy way is to buy some pre-packed factors. There is a huge offer of factor ETFs. For the sake of capturing the performance of value, size, and quality, we could choose, for example:
- iShares Edge MSCI World Value Factor UCITS ETF – to capture a global value factor.
- iShares Edge MSCI World Size Factor UCITS ETF – to capture a global size factor.
- iShares Edge MSCI World Quality Factor UCITS ETF – to capture a global quality factor.
While investing in ETFs is simple, it many times falls short of the objective. These ETFs are just tilts towards the factors and are far from pure factor investing. At the same time, considering the amount of funds they need to invest, they can only invest in larger market capitalisations, otherwise they would move the price substantially anytime they attempted to buy and sell shares. Holding less liquid stocks would increase the risk significantly. In the perspective of an individual investor these restrictions rarely exist. While an ETF may pile £10 million in a single stock, an individual investor may be satisfied with just £10 thousand.
Another option open to an individual investor is to use accounting ratios to select stocks. When this is the case, an investor must find accounting ratios that are good proxies for each of the factors and then select the top rankings (or bottom, depending on the perspective). Let’s say the aim is to build a value portfolio. If the price-to-book is a good proxy for value, investors could rank stocks from lowest to highest on price-to-book (as lower is better) and then select the top rankings until filling the number of desired stocks. This task could be repeated for other factors to then build the final portfolio. This kind of multi-factor portfolio is often referred to as a top-down combination portfolio.
Empirical evidence sometimes point to the fact that stocks ranking high in one factor may rank low in other factors. Let’s suppose that we want to build a value-size multi-factor portfolio. We start by selecting the top rankings for the value factor and then the top rankings for the size factor. If, for some reason, the companies with the best value rankings are also the companies with the worst size rankings and the companies with the best size rankings are the companies with the worst value rankings, we would end building a combination portfolio that would be more or less neutral to the value-size theme. Why? Because we would have the best and worst value companies and the best and worst capitalisations in the same portfolio. Another option of building a multi-factor portfolio follows a bottom-up approach. Instead of ranking stocks one factor at a time, we can attribute stocks an overall ranking. Only after we select the top rankings, which in essence are top rankings for the multi-factor theme.
Some Final Words
Before building a portfolio, investors should identify the risk sources behind returns and try to match these with the phase of the business cycle. While there is an ETF offer for almost anything; due to their size, ETFs aren’t always the best way of getting exposure to the market. In the particular case of factor investing, ETFs don’t constitute more than a slight tilt towards a factor, as they have many limitations. But, individual investors are less limited and can build the factors by themselves.
A true factor portfolio (in an academic sense) is a long-short portfolio that is neutral to every other factor. When we build a long-only portfolio, we may be exposed to more than we desire. Market risk would always be present. But, as selling short has its own limitations, in particular for individual investors, we must accept the shortfalls. A way of trying to get rid of the market factor would be to buy a short/inverse ETF on global equities, but, again, it increases the complexity and risk of the final portfolio.