Back to Portfolio for the Future™

Brief Thoughts on the Impacts of Autonomous Learning Investment Strategies on Investment Management and Employment

July 21, 2017
By Michael Weinberg, CFA
In our thought piece, “The Intelligent Investor in an Era of Autonomous Learning,” we explained why we believe the investment world will be transmogrified by Autonomous Learning Investment Strategies (ALIS).  In this brief follow-up, we will address a question that many of our investors and others have asked us regarding the impact on the opportunity set for investment managers and allocators like ourselves.
Recently, Blackrock made the headlines based on its restructuring of its active management business.  In its press release, it states, "Traditional methods of equity investing are being reshaped by massive advances in technology and data sciences...Asset managers who simply use the same techniques and tools from the past will limit their ability to generate alpha and deliver on client expectations." Having read numerous articles as well as the press release, we believe that they have come to a similar conclusion as us.

Our understanding is that they are effectively eliminating or replacing their large-mega capitalization fundamental discretionary portfolio managers with quantitative solutions and focusing their fundamental discretionary active management on less efficient areas of the market, such as international and sectoral investing.  The largest capitalization fundamental discretionary investing has become commoditized and there is very little value add that they can provide in this space on an active discretionary basis.

Similarly, we also believe that there are still inefficiencies globally that fundamental discretionary active managers may exploit.  For example, over the past few years we have discussed some opportunity sets that we believed have been interesting due to the greater inefficiencies and lesser competition.

Asia was one such opportunity set.  There are multiple economies; the vast majority of the universe of equities is shortable; there is less analyst coverage; regulations create market inefficiencies and higher inter-quartile spreads.  Unlike the rest of the world where hedge fund assets are at or above their prior peak, hedge fund assets are below their prior peak.  We believe these dynamics create inefficiencies, alpha and higher potential returns than the largest most developed markets and securities.  We continue to believe this thesis similarly holds in other non-Asian emerging markets where similar dynamics prevail.

Again, to clarify, we do continue to believe there are still opportunities for fundamental discretionary managers in the largest, most developed markets, such as the US. However, we believe these opportunities are in smaller capitalization securities, complex and misunderstood situations, capacity constrained and market independent strategies.   These strategies and opportunity sets, like our new ALIS ones, have always been the heart of Protege's small and emerging manager mandates.  In summary, we believe our ALIS managers are a strong, uncorrelated complement to our prior investment strategies and an extension of what Protégé has always done, which is to invest in what we believe are the best emerging managers and strategies.

Shifting gears from investment to society, if we are not wrong about the impact of ALIS on asset management, both traditional and alternative, the impact on composition and quantity of financial service employment is likely to be material.  For some time, we had a view that technology would facilitate developing world employment at the expense of the developed world.  For example, a radiologist in Asia could read x-rays over the internet at a fraction of the labor rate as a US physician.  With AL (sans the IS), we believe that a computer in the US could or soon will be shown a million x-rays of fractures and non-fractured bones and in no time will have a lower error rate than a radiologist.  Just as AL computers beat humans in checkers, chess, Go, and Texas Holdem Poker, they can be superior at non-medical image recognition compared to humans.

We are invested in an ALIS fund that has virtual analysts, traders, portfolio and risk managers.  If this were the first wave of hedge fund investing, those roles would have been filled by a team of people.  Even in the second wave of hedge fund investing, computational finance or quantitative managers likely had people in those roles.  With ALIS, those people are typically replaced by two PhDs, who tend to be the founders of the fund, and a few support programmers.  This strategy is scalable at least into the billions of AUM as it currently stands.

Twenty-five years ago, when we joined Wall Street, we remember when Trading was a superb career with strong demand, remuneration and a bright future.  Approximately 18 years ago, we remember running capital at a hedge fund when algorithmic trading was first released, at least at non-quantitative firms.  We took to it immediately.  Instead of having to go back and forth on a telephone with traders or sales-traders, one could very simply put the order in with the constraints and trade at Volume Weighted Average Price (VWAP) as well as other permutations that had different optimizations.  Moreover, the costs per share were fractional compared to the legacy telephone oriented system. Fast-forward to today and trading has largely become marginalized by algorithms.   There are far fewer trading roles and consequently demand for traders and their remuneration are both way down.

Moving back to our ALIS example, we believe it is only a matter of time, i.e. Years, before a similar path evolves with AL analysts, risk and portfolio managers.  Just as algos are lower cost and far more productive than human traders, the same will hold for many of the aforementioned roles.  At least for some time, just as we carved out an exclusion for smaller capitalization securities, complex and misunderstood situation, capacity constrained and market independent strategies, we believe those exceptions will hold here for some time too.  Irrespective of that, it does not bode well for the white-collar employment outlook on Wall Street and for top MBA students.

Just as there is already an issue in society that there is a mismatch between the unemployment rate and the skill-set of unemployed people, i.e. They do not have the Science, Technology, Engineering and Mathematics (STEM) skills that are required by our technologically driven economy, an analogous deficiency will hold for finance MBAs.  They, too, will need to re-tool to adapt to the realities of the labor force if they do not wish to face obsolescence.  Currently, there is a great deal of talk on the implications of AL in the Taxi, Limousine and Chauffeur (TLC) industry when we have driverless cars.  This is a canonical blue-collar example, however, we believe the white collar examples are under-appreciated.  Gladly, there is a solution and it is re-training.   We will save the details of that for a future essay.

For more nearly 25 years Michael Weinberg, CFA, has invested directly at the security level and indirectly as an asset allocator in traditional and alternative asset classes.  He is the Chief Investment Strategist at Protege Partners, where he is a Senior Managing Director, and on the investment and risk committees.  Michael is also an adjunct Associate Professor of Economics and Finance at Columbia Business School, where he teaches Pension, Sovereign and Institutional Investing, an advanced MBA course that he created.  He was a portfolio manager and global head of equities at FRM, a multi-strategy investment solutions provider.  Prior to that, Michael was a portfolio manager at Soros, the macro fund and family office, and at Credit Suisse First Boston.  Before that he was a real estate analyst at Dean Witter.  He has a BS from New York University and an MBA from Columbia Business School.