Summary. The tools of machine learning may offer active fund management firms many opportunities to outperform competitors and market indices, but the investments required in data analytics will be significant and the competitive advantage obtained many not be sustainable in the long term.
The tools of machine learning may offer active fund management firms many opportunities to outperform competitors and market indices, but the investments required in data analytics will be significant and the competitive advantage obtained many not be sustainable in the long term.
Leer en español
Ler em português
Some industry experts argue that machine learning (ML) will reverse an increasing trend toward passive investment funds. But although ML offers new tools that could help active investors outperform the indexes, it is unclear whether it will deliver a sustainable business model for active asset managers.
Let’s start with the positives
A form of artificial intelligence, ML enables powerful algorithms to analyze large data sets in order make predictions against defined goals. Instead of precisely following instructions coded by humans, these algorithms self-adjust through a process of trial and error to produce increasingly more accurate prescriptions as more data comes in.
ML is particularly adaptable to securities investing because the insights it garners can be acted on quickly and efficiently. By contrast, when ML generates new insights in other sectors, firms must overcome substantial constraints before putting those insights into action. For example, when Google develops a self-driving car powered by ML, it must gain approval from an array of stakeholders before that car can hit the road. These stakeholders include federal regulators, auto insurers, and local governments where these self-driving cars would operate. Portfolio managers do not need regulatory approval to translate ML insights into investment decisions.
In the context of investment management, ML augments the quantitative work already done by security analysts in three ways:
ML can identify potentially outperforming equities by finding new patterns in existing data sets.
For example, ML can sift through the substance and style of all the responses of CEOs in quarterly earnings calls of the S&P 500 companies during the past 20 years. By analyzing the history of these calls relative to good or bad stock performance, ML may generate insights applicable to statements by current CEOs. These insights range from estimating the trustworthiness of forecasts from specific company leaders to correlations in performance of firms in the same sector or operating in similar geographies.
Some of these new techniques produce significant improvements over traditional ones. In estimating the likelihood of bond defaults, for example, analysts have usually applied sophisticated statistical models developed in the 1960s and 1980s respectively by Professors Edward Altman and James Ohlson (notably the Z and O scores). Researchers have found that ML techniques are approximately 10% more accurate than those prior models at predicting bond defaults.
ML can make new forms of data analyzable.
In the past, many formats for information such as images and sounds could only be understood by humans; such formats were inherently difficult to utilize as computer inputs for investment managers. Trained ML algorithms can now identify elements within images faster and better than humans can. For example, by examining millions of satellite photographs in almost real-time, ML algorithms can predict Chinese agricultural crop yields while still in the fields or the number of cars in the parking lots of U.S. malls on holiday weekends.
A flourishing market has emerged for new forms of these alternative datasets. Analysts may use GPS locations from mobile phones to understand foot traffic at specific retail stores, or point of sale data to predict same store revenues versus previous periods. Computer programs can collect sales receipts sent to customers as a byproduct of various apps used by consumers as add-ons to their email system. When analysts interrogate these data sets at scale, they can detect useful trends in predicting company performance.
ML can reduce the negative effects of human biases on investment decisions.
In recent years, behavioral economists and cognitive psychologists have shed light on the extensive range of irrational decisions taken by most humans. Investors exhibit many of these biases, such as loss aversion (the preference for avoiding losses relative to generating equivalent gains) or confirmation bias (the tendency to interpret new evidence so as to affirm pre-existing beliefs).
ML can be employed to interrogate the historical trading record of portfolio managers and analyst teams to search for patterns manifesting these biases. Individuals can then double check investment decisions fitting into these unhelpful patterns. To be most effective, individuals should use ML to check for bias at every level of the investment process – including security selection, portfolio construction and trading executions.
Yet despite these substantial enhancements to investment decisions, ML has its own very significant limitations, which seriously undercut its apparent promise.
To begin with, ML algorithms may themselves exhibit significant biases derived from the data sources used in the training process, or from deficiencies of the algorithms. Although ML will reduce human biases in investing, firms will need to have data scientists select the right sources of alternative data, manipulate the data, and integrate it with existing knowledge within the firm to prevent new biases from creeping in. This is an ongoing process that requires competencies many traditional asset managers don’t currently have.
Secondly, although ML can be very effective at examining huge amounts of past data from one specific domain and finding new patterns relative to an express objective, it does not adapt well to rare situations such as political coups or natural disasters. Nor can ML predict future events if they are not closely related to past trends, such as the 2008 financial crisis. In these cases, investment professionals must make judgments about where future trends are going, based partly on their intuition and general knowledge.
Finally, many of the patterns ML identifies in large data sets are often only correlations that cast no light on their underlying drivers, which means that investment firms will still need to employ skilled professionals to decide if these correlations are signal or noise. According to a ML expert at a large U.S. investment manager, his team spends days evaluating whether any pattern detected by ML meets all of four tests: sensible, predictive, consistent, and additive.
Even when ML finds patterns that meet all four tests, these aren’t always easily convertible into profitable investment decisions, which will still require a professional’s judgment. For example, by sifting through reams of social media, ML might have been able to predict — contrary to most polls — that Donald Trump would be elected president in 2016. However, making an investment decision based on that prediction would present a difficult question. Would Trump’s election lead the stock market to go up, down, or sideways?
The bottom line is that while ML can greatly improve the quality of data analysis, it cannot replace human judgment. To utilize these new tools effectively, asset management firms will need computers and humans to play complementary roles. As a result, firms will have to make substantial investments going forward in both technology and people, although some of these costs will be offset by cutting back on the number of traditional analysts.
Unfortunately, most other asset managers have not gone far down the path to implementing ML. According to a 2019 survey by the CFA Institute, few investment professionals are currently using the computer programs typically associated with ML. Instead, most portfolio managers continued to rely on Excel spreadsheets and desktop data tools. Moreover, only 10% of portfolio managers responding to the CFA survey had used ML techniques during the prior 12 months.
Perhaps predictably, it’s the largest asset managers, like BlackRock and Fidelity, that are leading the way, nurturing relationships with information suppliers, technology providers, and academic experts. But they are unlikely to open up a large gap over competitors as scale is not necessarily an advantage in active investment. For instance, trading in large volumes can carry significant costs and firms may be constrained in the amount of overall exposure they can carry in a particular stock.
Mid-size asset managers should also be able to benefit, because they are likely to attract and retain high-quality data scientists who may see more opportunities for advancement there than in the very large firms. In addition, mid-size firms will be able to afford access to alternative data through third-party vendors, high-quality algorithms from open source libraries, and sophisticated tools from the technology companies (e.g., Amazon and Google) that already offering cloud-based services to many industries.
The losers are likely to be small firms (with less than $1 billion in assets under management). They are likely to have trouble attracting enough talent and absorbing the cost of developing the technology given the strong downward pressures on fees of active managers. Management fees for active equity managers are roughly 20% lower in 2018 than they were in 2008, in part because passive funds have become so cheap. Asset managers are also under regulatory pressure to pay their own cash for outside securities research, instead of paying with “soft dollars” by allocating brokerage commissions to good research firms. The investments required by ML, therefore, come at a difficult time generally for the asset management industry, and this will be particularly challenging for small firms.
What’s more, it is unclear whether substantial investments in ML will in fact lead to a long-term sustainable business model for active asset managers. If ML generates unique alpha for an investment firm, the firm cannot sit on its laurels for long because other firms are likely to simulate its investment methods. And if other asset managers derive similar insights from similar ML techniques, they will be buying or selling the same securities at the same time, which may have the effect of wiping out any gains the insight can generate. This has already happened on a number of occasions. Over three days in 2007, for example, several large hedge funds, using quantitative models based on the same factors, liquidated their positions simultaneously and suffered large losses as a result.
To sum up, ML may be seen initially as the savior of active investing. It surely has the potential to allow early adopters to find new sources of alpha and outperform the indexes. Yet if the insights from ML are copied by other managers as they develop ML capabilities, it may become even more difficult to find publicly traded stocks and bonds that outperform their benchmarks. Over time, will active investing augmented by ML increase the efficiency of security pricing and thereby reinforce the current shift to passive investing? If so, the costs of implementing ML will be borne by active managers, but much of the benefit will go to index funds as free riders.