October 27, 2020
“The democratization of information has made it much harder for active management,” Larry Fink, BlackRock’s chief executive officer, told the New York Times back in March 2017. “We have to change the ecosystem—that means relying more on big data, artificial intelligence, factors and models within quant and traditional investment strategies.” Mr. Fink was referring to a marked shift in his firm’s investment strategy, one that would see some $30 billion in assets (about 11 percent of BlackRock’s active equity funds) designated for a fundamental stock-picking method also incorporate a quantitative approach based on modelling and algorithms. And in doing so, the feted investment firm was adopting an investment strategy—or, indeed, a combination of two distinct investment strategies—that has come to be known as quantamental investing.
The term quantamental is, in fact, a portmanteau of quantitative and fundamental and thus refers to an investment strategy that involves combining quantitative and fundamental approaches to investing, with the aim of improving returns. By merging these two approaches to investing—computer power with human insight—it is hoped by those now adopting quantamental investing that such an approach can deliver superior returns. Indeed, this combination is often regarded as a melding of man and machine: the investor harnesses the scale and power of data and blends it with the benefits of human insight in order to unearth winning investment strategies.
Fundamental investing largely refers to the traditional bottom-up method of picking individual stocks based on the quality of their fundamentals, such as earnings and market share. By identifying securities with prices that are not reflective of the true market value of their companies, fundamental investors aim to exploit this mispricing to generate alpha—that is, the excess return of an investment in relation to the return of the market (or a benchmark index). The likes of Warren Buffett and Charlie Munger of Berkshire Hathaway represent the best proponents of this approach, using detailed analysis of company balance sheets, coupled with sound judgment and experience, to consistently generate alpha.
Quantitative investing, on the other hand, refers to the use of computer models and algorithms, as well as invariably vast quantities of data, to determine trends and patterns and thus more effectively attempt to predict future security price movements. Popular quant strategies include statistical arbitrage, whereby the model seeks out arbitrage opportunities by identifying price differences between identical securities that often tend to emerge for just a few seconds, and factor investing, in which the model identifies specific factors that influence the price of a security such as macroeconomic, microeconomic and style factors (for example, market capitalisation). And with computers making the trading decisions, quantitative investing removes the human-error component, including emotional and/or cognitive biases, that often crops up when investment decisions are made by investment managers. This represents a major advantage for the quant strategy. That said, given that the algorithms themselves are designed by humans, quantitative strategies are still limited to the skills of the developers behind those strategies.
By identifying historic patterns and comparing them to the current market environment, moreover, quant investors try to extract trading signals upon which they can act, ostensibly with more confidence. Data, therefore, plays a crucial role. Quant investors will normally attempt to procure as much relevant data as possible in order to make them better informed. This explains the current boom in alternative data—that is, non-traditional data such as satellite imagery and geolocation data that can provide significantly more granular insight into a particular investment target. For example, satellite images showing how occupied a retailer’s parking lot is over the course of a quarter can create a strong gauge of the retailer’s sales volume for that period and thus indicate how well it has performed.
Managers are now paying exorbitant amounts to obtain such data in the hope of giving them an edge over the competition. “The purpose of data is to make forecasts, and the more I have, regardless of who owns it, allows me to gain a greater conviction in existing ideas,” Ken Perry, former chief risk officer at Och-Ziff Capital Management, once told a Bloomberg panel discussion on alternative data. That said, this process currently remains in its infancy, with datasets often yielding little to no meaningful information. But data-mining techniques are also becoming more advanced all the time, with technologies such as machine learning helping programs to adapt to feedback without the need for human involvement. In turn, this is enabling investment firms to analyse more effectively big datasets and identify meaningful patterns. And with computers becoming consistently more powerful and data-storage options becoming continually cheaper, profitable opportunities for quantitative investors are growing all the time.
It should then come as little surprise that fundamental stock pickers have been under pressure for some time. With quantitative investors having entered the arena and rapidly gathered steam over the last 15 years or so, fundamental investors have faced stiff competition from this new breed of investor, meaning that their returns have been severely squeezed. Indeed, pure quant firms such as Renaissance Technologies and Two Sigma have generated some of the best returns in the industry. Active fund managers are also continuing to lose out to passive-investment strategies that simply seek to track a particular index at a comparably low cost. Much of the last decade has seen passive-investment tools such as exchange-traded funds explode in popularity by easily outperforming active strategies. This has increasingly called into question the hefty fees that active managers levy on their investors.
With research showing that broad quantitative market factors such as valuation, growth, quality and momentum have driven around 65 percent of global equity managers’ relative returns over the past 20 years, while the remaining 35 percent has been attributable to stock selection, investment managers are now increasingly seeking to supplement their fundamental approach with a robust quantitative-factor research component—and vice versa—such that quantamental investing captures a greater degree of excess returns, at both the market-factor level and the individual-stock level. The quantamental investor also benefits by having a strategy to fall back on—should one approach fail to work sufficiently well, the investor can opt for the other method, depending on the market environment.
Often quantamental investors will have initially begun as fundamentalists before beginning to absorb ideas from quantitative strategies. In practice, this would perhaps involve hiring data scientists who can develop machine-learning algorithms to glean useful investment insights from big datasets, which in turn can augment the stock-picking process. BlackRock is just one of several high-profile fundamental investing firms to have made steps in this direction, with other notable names to have made a similar transition including Point72 Asset Management and Tudor Investment. Indeed, large financial institutions all over the world have started to dip their toes into the quant world, leveraging the insights generated from big data to bolster their returns.
But there remains considerable doubt over whether fundamental investors can sufficiently adapt to absorb such a challenging discipline. Indeed, given the significant differences between the two strategies, switching to quantamental investing is likely to involve sizeable structural and organisational changes within a fundamental investment firm. For one, it will require change in the hiring policy to employ more personnel with quantitative research and development skills, in addition to significantly upgrading internal systems to cope with the strain on resources imposed by the big-data analytics process. Such changes, therefore, are usually associated with a substantial dollar investment. According to the Chicago-based international executive-search firm Heidrick & Struggles, asset managers can employ a three-step process in order to make the transition towards a more quantamental investing style smoothly.
Behavioural and cultural change may seem like an arduous challenge, but it would appear that it is becoming increasingly necessary. With technology playing a consistently more influential role in the asset-management process, and with traditional asset managers being consistently outperformed by passive strategies and quant investors, it has become imperative that fund managers adapt to this new paradigm. Obstacles do exist when it comes to taking the leap and bridging the two worlds, particularly in regards to cost. But ultimately, if they can successfully make the transition towards a quantamental strategy, it would seem that traditional asset managers will have a better chance of surviving, evolving and ultimately thriving in this new world.