Asset managers that employ standard portfolio management techniques have largely been playing the same game over the past several decades, with managers and their staffs building investment models manually -- gathering traditional financial data from public sources and inputting them in a relatively basic model.
While some of the tools used by asset managers have advanced, conventional active management approaches are reaching the limit of their ability to generate alpha. The industry is defined by a handful of large fund complexes that have failed to innovate, but data science and artificial intelligence could drive a profound and positive transformation to the industry.
Advances in computing, coupled with ready availability of data on virtually every aspect of the global economy and everyday life, have led to a proliferation of opportunities for artificial intelligence to upend traditional approaches while driving considerably better outcomes and empowering investors to invest more confidently.
To place the opportunity set into proper perspective, imagine the captain of a ship speeding to an island 200 miles offshore. A GPS device will provide certain inputs that surely will prove useful in navigating the way, such as precise position, bearing and speed.
Yet what is more striking is the critical information a GPS may not provide that could prove useful — wind speed and direction, for example, or data on weather or ocean currents.
There are ways to gather that information, and a seasoned navigator knows which data points should carry more weight as she integrates them into her guidance of the craft. The likelihood that the ship reaches the island safely dramatically increases with the increased number and quality of relevant inputs. Machine learning can have a similarly transformational impact on asset management.
More information on human activity exists today than has ever existed in history, and the connectedness of our digital infrastructure provides deeper and broader data sets than ever before available, while making that data available on a real-time basis.
But no matter how astute any human fund manager is, no person or team of persons is capable of processing, prioritizing and analyzing all of this data on a sufficiently rapid basis.
Just as a greater breadth of navigational and environmental information help our imaginary ship’s captain make better decisions, more data and smarter algorithms can help human asset managers build investment products that independently execute precisely the right trades to achieve desired outcomes.
One important area where modern data science and machine learning can transform asset management is in redefining and properly assessing risk. Until now, risk and return have long been thought of as inextricably linked – to shoot for high returns, investors must be willing to take on greater risk, and to have lower risk, they must live with more modest returns.
Operating on the basis of this premise, asset manager approaches to tackling the risk-versus-return question invariably fall under one of the following categories: diversifying between asset classes and specific investments; investing across a basket of already diversified funds; and relying on more active management in the form of aggressive trading and rebalancing.
The problem with each of these approaches is the insufficient recognition that, above all else, risk is a drawdown on any investment portfolio — either with realized losses or unrealized losses in the form of downward volatility along the way.
Artificial intelligence and machine learning can foster the ability to more rapidly and deeply mine a broader set of data sources for inputs that can feed into exponentially more sophisticated trading algorithms today, executing on a more precise trading approach that ladders up to investor goals without adding unnecessary risk.
Arguably, one of the bigger roadblocks to adoption of AI across the asset management sector is attitudinal, with many industry participants suffering from the misplaced fear that AI will completely displace people or will make decisions that we do not understand and cannot monitor.
Approached correctly, with understandable algorithms, and aligned with client objectives, the adoption of AI across asset management will accelerate the ability of investment professionals to better serve investors and create new sources of value, making them even more essential.
Just as a bank of sophisticated instruments, in addition to a GPS, enables a captain to navigate more effectively, artificial intelligence and machine learning can empower asset managers to decouple risk and return more effectively than ever before, to the benefit of investors everywhere.
Andreas Roell is chief executive of AlphaTrAI, a leading venture-backed and data science-driven asset management firm.
Medicare scam, pandemic benefit theft, offshore tax evasion — federal prosecutors are casting a wide net.
Report finds that pension income acts as a financial lifeline for retirees facing late-life shocks and raises urgent questions about the DC-only future.
Nine-month electronic trading freeze and share lending program at the center of dismissed claim.
Meanwhile, Rossby Financial's leadership buildout rolls on with a new COO appointment as Balefire Wealth welcomes a distinguished retirement specialist to its national network.
With a smaller group of companies driving stock market performance, advisors must work more intentionally to manage concentration risks within client portfolios.
As technical expertise becomes increasingly commoditized, advisors who can integrate strategy, relationships, and specialized expertise into a cohesive client experience will define the next era of wealth management
Growth may get the headlines, but in my experience, longevity is earned through structure, culture, and discipline
