The idea that there is a reward in bearing uncertainty – that gains are proportional to risks – is not only intuitive but a deeply-rooted assumption in the investment culture. Its explanation, however, has required a decades-long challenge to understand the drivers of investment returns. On top of this, the debate over active and passive investment – between Alpha and Beta – has unveiled that, instead of pure skill, a great portion of the excess return over a benchmark results from the exposure to a set of industry or firm-specific characteristics of financial securities, the so-called investment factors. Eventually, starting from the famous equity risk premium, many other risk premia have emerged during the years, paving the way to the modern factor-based approach to portfolio management. Nevertheless, excess returns are not going to disappear anytime soon. As we are going to see, what is changing is the dimension in which they exist, as they move from being strategy-specific to being allocation-driven.
This leaves us with a thought-provoking question: is Alpha transforming into the ability to move between Betas?
Introduction
As investing involves risk, managing its exposure and picking the right securities are an investment manager’s first concerns when it comes to building a portfolio. Indeed, an efficient allocation of capital is not just a matter of performance, but a complex set of decisions that requires returns to be constantly benchmarked against the amount of risk taken.
Simply put, instead of gross returns, risk-adjusted performance is what matters the most. Behind this simple – yet crucial – intuition lies the idea that there is no point in taking additional risks if they are not rewarded appropriately. Indeed, getting more returns out of the same level of risk is what we tend to associate with the skills of a manager, that is, delivering the so-called Alpha.
The connection between risk and return seems to be quite straightforward at first sight: higher expected returns are associated with higher uncertainty. This means that if an investor expects to earn a return higher than the risk-free rate (i.e. the return on a government bond, such as the US T-bill), the additional component must be compensation for the additional risk.
It almost feels natural to interpret such excess return over the risk-free rate as the sum of many risk-premia, one for each source of uncertainty a security (or portfolio) is exposed to, as shown by the following picture.
Start integrating AI into your investment process with the most advanced no-code solution.
Book A Custom Demo
Exposing the DNA of Returns
The first attempt to frame this relation in a mathematical model dates back to the efforts of William Sharpe and John Lintner that – in the early 1960s – developed a model that has become the cornerstone of modern finance, the Capital Asset Pricing Model, better known under the acronym CAPM.
Building on the Nobel-prize winner Harry Markowitz’s Modern Portfolio Theory, the model was intended as a tool to price securities, that is, determine their equilibrium risk-adjusted rate of return. The intuition behind the model, similar to the one represented in Exhibit 1, considered returns to be made up of two parts, a risk-free component and a unique risk-premium. Systematic risk (i.e. the risk that is not possible to reduce through diversification) is assumed to be the only source of uncertainty that markets reward. Not surprisingly, its introduction has deeply transformed traditional Investment Management and opened the debate about active versus passive investing.
Indeed, the CAPM was the first outright attempt to expose the DNA of returns. It made clear how investors could harvest the premium for investing in stocks – the equity risk premium – just by holding passively a broad and well-diversified portfolio of stocks that represented market exposure, like the S&P 500 Index.
However, the equity risk premium (i.e. the CAPM Beta) explained only a fraction of the total return spectrum. For this reason, it felt natural to interpret the unexplained component (i.e. Alpha) as a proxy for the ability to actively beat the market through a combination of timing and stock picking (or perhaps just blind luck). If we call R*asset the CAPM estimate of return on a risky asset, rf the risk-free rate, rm the return on the market and the measure of the systematic risk of the security relative to the market portfolio, we understand the logic behind the model’s most-famous equation: the degree by which an asset is exposed to the market determines its additional earned premium.
Soon the CAPM became the industry-standard for Asset Pricing, paving the way to another stream of quantitative research dedicated entirely to the exploration of alternative factors to address the remaining part of the returns. However, the debate over active investment and skill – the so-called Alpha generation – has always been a controversial topic in the financial literature. Let’s take a step back to see why.
Questioning Investing Rule No.1
When we think about the role played by markets in the financial system, we see them as the place where people trade securities, commodities or rights of any nature. Intuitively, we grasp that their correct functioning is the backbone on top of which the supply and demand for any kind of good are constantly matched.
Especially when they stop being physical and become digital venues, as it occurs to modern exchanges, their mechanisms (i.e. the market microstructure) get increasingly important in making price – the medium by which information is shared between investors – as efficient as possible. Eventually, we understand that financial markets are essentially risk-pricing units. As they close the gap between the supply and demand for any available security, they collect and aggregate the implied risk preferences of investors. With an auction mechanism, they put a price on risk so that they can – on aggregate – allocate returns among risk-takers and risk-averse agents. The more uncertainty (i.e. risk) an investor is willing to take, the higher the expected compensation he gets, and vice versa.
This suggests a fundamental – yet unwritten – rule that every market participant knows: “there is no such thing as a free lunch”. In other words, it is almost impossible to get a reward without risking something. So, what does this tell us about Alpha? Where is excess return coming from? And is its existence consistent with all this reasoning?
What appears at first sight is that the “no free lunch” argument (i.e. also known as the Efficient Market Hypothesis) holds true when we look at the big picture rather than on a trade-by-trade scale.
Indeed, even if the interplay between market participants (which buy and sell securities according to their future prospects) embeds new information into the prices of securities, the speed of adjustment of such mechanism can largely differ, causing temporary inefficiencies to live for a short period of time before getting exploited and arbitraged away, that is, eliminated by the means of trading.
Yet, this seems to be one of the root causes of Alpha. However, although it is true that it is possible to achieve an excess return, its distribution tends to be spread randomly, signaling on the one hand that it is difficult to design investment strategies that work consistently over time and, on the other hand, that maybe what we regard as Alpha may be the result of a hidden risk factor.
Supporting this thesis, a vast literature that relates overperformance to common characteristics of securities – something already hinted at by Benjamin Graham when he wrote "The Intelligent Investor" in 1949 – has contributed to shedding light on a wide range of additional risk premia.
Fact-checking Overperformance
Although CAPM suffered from some well-known limitations, it represented not only a useful way to translate risk into expected returns but it offered a tool for investors to fact-check claims of overperformance by their investment managers. After all, its broad adoption should not surprise, as it tried to solve the puzzle of over-performance, that is, to determine its amount. Indeed, as the model allowed us to break down performance into risk-based components, it was now possible to measure the degree by which a manager was effectively adding value over a benchmark.
For example, consider a US-based investor that has invested in an active domestic equity mutual fund. At the end of the period, his fund returned 15% while its benchmark – the S&P 500 – returned 12%. Assume also the rate of return on a T-Bill is 2%. At first sight, one could say the manager has overperformed its benchmark by 3%. However, this simple answer does not give any insight into the amount of risk taken by the manager during the investment period. With the CAPM, an investor could easily estimate the risk-adjusted performance of its investments running a linear regression of the fund’s return on the returns of the S&P 500 using the following formula:
If we assume for simplicity the estimated Beta parameter from the regression to be 1.2 (i.e. the fund was 20% riskier than the S&P 500), the required expected return for the fund would have been:
Thus – instead of the raw 3% – the manager would have overperformed only by 1% with respect to the increased risk he took. Eventually, the CAPM unveiled that a great portion of what managers were delivering on their portfolios was explained by broad market exposure (i.e. Beta) instead of pure skill.
Nevertheless, Alpha was still dodging the bullet. Was it structural or was it driven by hidden variables yet to be discovered? Intuitively, it felt like the CAPM was only scratching the surface of a big area of research, which – almost 50-years later – sees multifactor models such as the one developed by Fama-French leading the modern landscape of Factor Investing.
Alpha is More Than You Think
Among the many adjectives we can assign to overperformance, three appear to give a precise and objective picture of what CAPM Alpha is: ex-post, relative and model-dependent. Indeed, it is a measure of performance that indicates the amount of active return over a benchmark, net of the risk-free rate and the premium for the Beta exposure.
As the equation shows, a positive Alpha then stands for better-than-expected risk-adjusted performance. So, in a sense, clearing the air about its meaning has a twofold objective: on the one hand, it makes us understand what overperformance does not consist of and, on the other hand, it allows us to zoom into its multidimensional nature – the many shades of Alpha.
Risk-based Alpha
As a first approximation, the existence of an informational gap (i.e. between the manager and the market) is what appears to be a good explanation of the reason why we observe overperformance in the real world. More precisely, it would make Alpha only the evidence of skill, that is, the ability of a manager to time markets and pick stocks better than others.
However, skill alone seems to tell only a part of the whole story. Indeed, as we learn more about the true nature of investment returns, our models get better and we become able to look at overperformance with a new pair of glasses. What happens is that eventually we understand that a great part of what we previously thought of as Alpha turns into Beta. In other words, a great part of the excess return we observed was indeed the result of an exposure to a hidden source of return instead of pure skill (i.e. as it happened for the equity risk premium).
Yet, this observation calls for a more comprehensive explanation of the rationale for Alpha. Indeed, it seems that there are many reasons for it to exist. For instance, it can result from investment strategies that exploit systematic biases that investors face when making investment decisions – the so-called behavioral Alpha.
Behavioral Alpha
Supporting this idea, a vast literature that has married traditional economic theory with the most recent developments in behavioral psychology and neuroscience has shed light on a number of common biases that people experience when they take decisions. Indeed, the seminal contributions of Daniel Kahneman and Amos Tversky – Nobel prize winners and pioneers in the field of behavioral finance – have proposed a new framework to address the mechanisms and biases behind the human decision-making process.
Among those, the disposition effect (i.e. the tendency of investors to sell winners and hold on to losers) and the over/under reaction to new information (i.e. people tend to update their beliefs at a higher/slower pace when they receive new information) can help explain the presence of persistent market anomalies that can be exploited by rational and more disciplined trading rules, and so, deliver Alpha.
Operational Alpha
Eventually, we understand the reason why Alpha is so sought after. It is because it represents a component of return that escapes the traditional risk-return explanation in favor of being interpreted as a proxy for pure skill, informational advantage or technological edge. In recent years, technology has added an additional dimension for excess return and has gained traction as a strategic asset to many investment and asset managers. Indeed, Operational Alpha, the excess return connected with operational efficiency, emerges as the result of the benefits received from an efficient, scalable and data-driven investment process.
This, in turn, can be seen as the effect of more efficient deployment of the investment strategy and in a reduction of risk. Indeed, if, on the one hand, it means an increased level of Alpha protection (i.e. the efficient delivery of the Alpha generated by the manager to the client), on the other hand, better operations are the result of the use of technology to streamline processes and reduce the level of operational risk incurred.
Why Invest in Factors
The giant leap caused by the CAPM and the discovery of the equity risk premium made clear that the search for other drivers of excess return was only at the beginning.
Not so long after its introduction, a new theory of asset pricing – the Arbitrage Pricing Theory – was proposed in 1976 suggesting not only that securities could diverge temporarily from their equilibrium price (i.e. being mispriced), but also that the original model could be expanded to include other variables to explain securities’ returns.
However, as no hint was given on the methodology to select them, academics struggled initially to achieve consensus regarding which were the sources of alternative risk premia. It was not until the early 1990s, that the seminal paper from the Nobel prize winner Eugene Fama and his colleague Kenneth French entitled “Common risk factors in the returns on stocks and bonds” revamped the argument, uncovering what since then have been known as size and value premia.
What they observed was that – in addition to the market Beta – other exposures drove excess returns consistently across markets and geographies: the size of the company (expressed in terms of its market capitalization) and their relative cheapness (measured by the ratio between the book and market value of assets). Adding this information to the existing framework (i.e. meaning more parameters to be estimated) led the new model to explain around 90% of a security’s volatility – again, Alpha was shrinking.
Eventually, the Fama-French multifactor model was built under almost the same assumptions of the original CAPM, except that now instead of a unique risk factor, the return of a security could be broken down into multiple Betas, as shown in the following illustration.
Yet, their biggest contribution was to shed light on an alternative (and smart) method for constructing risk factors. Instead of using the excess return over the risk-free rate, their factors looked at the spread in performance between stocks sorted from top to bottom according to a given feature.
For instance, if the size of a company was a factor driving investment returns, then performance would have shown significant variability between small and big ones. Similarly, if the value factor was a driver of excess returns we would have seen a portfolio composed of a long position in cheap stocks and a short position in expensive ones consistently outperform the market.
As empirical findings initially confirmed their intuition, this methodology sparked interest in the field of Factor Investing. For the first time, Alpha started to look a bit less relevant and it seemed only a matter of time until researchers could use this methodology to uncover all the sources of excess return. With this technique, the growth, yield, low volatility and eventually the momentum factor – the excess return from buying winners and selling losers stocks, caused by investors’ overreaction – were discovered. Managers now could not only choose among a broad range of risk premia but also had a tool to get ahead of persistent cognitive biases of investors.
All That Glitters Is Not Gold
However, the road to identify a risk premium is definitely not an easy one. As one could guess, among the many candidate sources of excess return, only a tiny fraction becomes eventually an investment factor.
This happens because, in order to reach a broad consensus, several robustness conditions must be met to prove the findings are not just the result of data mining. Indeed, even in the case of statistical evidence, execution and transaction costs raise the bar even higher for them to become investable.
Nonetheless, the starting point for solid factor research is grounded in the simple yet crucial consideration that primarily investment factors should make sense from an economic standpoint. For instance, the rationale for the equity risk premium is grounded in the economic reasoning that suggests that stocks should earn a premium with respect to bonds because they are exposed to a higher uncertainty of cash flows, and investors are risk-averse.
Once a solid rationale has been established, it takes at least two additional steps to make sure a source of excess return becomes eligible: long evidence of its persistence over time (i.e. generally multiple decades of data are required) and proof of its robustness across markets and geographies. For instance, the size factor was first discovered during the early 1980s on US stocks using about 40 years of data, and later the initial findings have been thoroughly investigated in subsequent studies that confirmed its presence in the European, Asia-Pacific and Japanese markets. The following table summarizes the key findings of known and robust drivers of returns.
Exhibit 4 – Key facts about the main equity investment factors and drivers of excess returns.
Do It Right > Just Do It
Eventually, the new paradigm – using factors to harvest risk premia – has opened up many possibilities to build portfolios according to a rule-based and transparent methodology.
With a clearer mapping of investment risks in mind, now managers have a deeper look-through and can use factors as the building blocks of their investment strategies.
This, in turn, has fueled the increasing demand for factor-based investment products. According to a 2019 report by Morningstar, they have reached a record-high $800 billion in assets under management across more than 1500 exchange-traded products and this figure is expected to rise above the $1 trillion mark in the next few years. However, as we are going to see, there are some common traits of real-world factor-based investment strategies that require further examination.
Factor Design
Imagine we want to include momentum in our portfolio. We start by identifying a group of momentum-based ETFs readily available on the market. Before we choose one, we look at the cross-section of their returns in two different time frames and observe the following pictures.
Exhibit 5 – Comparison of a group of Momentum-based ETFs in two different time frames: 1 year (left) and 6 months (right). Source: Bloomberg
As the pictures show, there is significant variability in the final performance achieved by the funds. Indeed, although they move along the same trajectory (i.e. meaning they are all exposed to the same source of risk), the gap that accumulates over time reflects the slightly different composition of the underlying portfolio. Eventually, what this evidence suggests is that design is something that matters over the long-run.
Indeed, if we start from the raw momentum effect described in Jegadeesh and Titman (1993) (i.e. winner stocks tend to overperform loser stocks), there are many design choices (i.e. optimal time frame, frequency of rebalances, long/short bias, weighting scheme, etc.) that can severely impact the final performance achieved by the strategy. In this sense, it does not surprise that a very active area of research has been focused on the construction and optimization of factor-based investment strategies. In other words, knowing a factor exists is just as important as having in mind a well-defined plan to best deploy it in a financial product.
Time-varying risk premia
Besides, what tends to be less marketed (or sometimes overlooked) about the true nature of factor-based strategies is that although they are a persistent source of excess return, they are not meant to always earn a positive risk premium.
Indeed, we can imagine that at any given time, the premium we get from factor strategies has a twofold nature. On the one hand, it represents an ideal fraction of an aggregate risk premium that markets reward for being exposed against that particular risk. On the other hand, it depends on the relative crowding of each strategy (i.e. how many investors are pursuing it at the same time) that makes the premium more difficult to harvest for investors and vice versa.
In both cases, we can expect its size to naturally increase or shrink as soon as the market cycles unfold and investors’ risk preferences change accordingly.
Yet this feature – being pro or countercyclical – offers an important insight for what concerns diversification and, eventually, Alpha generation. As standalone factors alternate periods of sustained overperformance with moments of underperformance, switching between them – the ability to move between Betas – is becoming increasingly strategic for asset and investment managers as a way to allocate capital more efficiently.
A New Dawn for Alpha?
As we learn more about the true nature of risk and investment returns, it is inevitable to see in the rise of Factor Investing and alternative risk premia the new face of portfolio management – transparent, rule-based and factor-driven. Indeed, from single to multi-factor models, decades of research have uncovered how a specific set of characteristics explains a great portion of the final return on financial securities, making it clear for investment managers that having a tight grip on factors right now is not an option, but a key driver of performance and diversification.
Yet, excess return has not disappeared completely but rather shifted dimension – from being strategy-specific to be allocation-driven. While the trajectory of excess return evolves and eventually becomes Beta, delivering Alpha has remained a top priority for active managers that now have a broader toolkit to harvest risk premia and add value through technology, operational efficiency, and deeper research.
