Recently, in Bye-bye “Smart Beta” and Hello “Factor Investing” we highlighted how first-generation smart-beta strategies are now being supplanted by factor investing which uses factors to select stocks and seeks to build a portfolio of stocks with common factors that can generate returns in excess of those of the wider equity market.
Such factors measure how cheap a stock is, how profitable the underlying company is, how risky it is and how strongly the stock price is trending. Portfolios tilted towards cheap, profitable, low-risk and trending (i.e. those with momentum) stocks have tended to outperform more traditional portfolios built around market capitalisation-weighted benchmarks.
Some discretionary active styles of portfolio management have been using similar approaches for decades. Many successful active managers have relied on much the same factors to narrow down their investment universes before embarking on fundamental analysis and stock-picking. This last step is absent if a fully systematic approach to factor investing is chosen. The use of systematic approaches offers a number of advantages when it comes to disciplined investing and controlling risk.
Equally important, is the role such approaches can play in defying the very same cognitive biases that help explain why the different investment factors can be a successful basis for stock-picking. We believe that all investors should consider a systematic approach to factor investing, but it is particularly relevant to those who doubt that fundamental managers can add any value through stock-picking. For these investors, systematic factor investing offers an important alternative to the replication of market capitalisation indices.
An objective approach that is immune from cognitive biases
Factor investment is based on the concept of using pre-established criteria to select stocks. The strength of a fully systematic approach to factor investing is that these criteria can be applied objectively and rationally, allowing us to implement investment decisions without falling victim to the impact that cognitive biases have on investment decisions. In the video “How do behavioural biases impact our investment choices?”, we illustrate how cognitive biases influence our decision-making without us even noticing. This is the case in particular when we invest.
Calculation power and risk quantification
Selecting a handful of promising stocks from a universe of hundreds or even thousands can be a daunting task. It is unrealistic to expect any single investor to be able to digest all the information relating to every company before making an investment decision. Fundamental portfolio management teams must be either very heavily staffed or use a systematic approach to narrow down the investment universe to a manageable size. The use of quantitative approaches to digest all the data and generate a stock selection using criteria known to predict excess returns can be a viable alternative and should be considered, in our view.
In addition, the use of fully systematic approaches to factor investing allows for much more than just the production of a list of stocks more likely to outperform an index. For example, portfolio construction algorithms which minimise the impact of portfolio constraints and the market impact of transactions can be used to determine the optimal allocation of a weight to each pre-selected stock. The use of a fully quantitative framework also allows for better control of how risk is allocated to each of the factors, i.e. how much of the stock selection and the stock weight is derived from each factor criterion. This makes it possible to construct multi-factor portfolios with a rigorously controlled allocation to the various factors.
A understandable and transparent management process
Another advantage of the use of a fully systematic framework in factor investing is its transparency – interestingly, this is a point that is not usually associated with quantitatively managed processes because often such processes are either not fully understood or over-mystified. The reality is that in a properly built systematic framework for factor investing, the selection of a stock for the portfolio should be fully explainable. It is possible to trace back positions to the stock-picking criteria used.
The corollary of this transparency is that it becomes possible to simulate what an allocation to a given factor would have generated in the past in terms of investment return. Such historical simulations can be used for a precise analysis of a strategy’s performance profile and risk, both comprehensively over a long period, e.g. since 1970, and over specific periods of interest for stress testing, e.g. the great financial crisis on 2008.
A warning: historical simulations should be used with care
We say this because investors should be aware of naïve systematic approaches based on over-optimising the past investment returns of the strategy, while ignoring the actual significance of the criteria used or the relevance of the historical sample used for future returns. This is a serious risk because with the data and the software available now, almost anyone can build a spurious strategy destined to fail and simply based on past over-optimisation.
Factor investing can be an extremely powerful approach to investing, at the crossroads of quantitative investment, using systematic techniques based on pre-set rules, and discretionary management, helping to focus the stock selection within a manageable universe. It is typically free of the cognitive biases that are known to debilitate our ability to efficiently make investment decisions and can be deployed efficiently in broad investment universes with a great degree of transparency in making investment decisions.