Market Efficiency, AI Analysts, Human Analysts, and Wham!

If you were away from work at times in the last few weeks, you may have missed these essays:

“The Active Management Reinvention Project”:  It is time to quit defending active management as it is and to start imagining what it should be.

“A View from the Inside”:  Narratives about the culture of an organization can mask the reality of it.  Often unauthorized accounts are closer to the truth than authorized ones.

Subscribe now to receive future postings.

Market efficiency

In “The Less-Efficient Market Hypothesis,” Cliff Asness argues that “over the past 30+ years markets have become less informationally efficient in the relative pricing of common stocks.”  The paper, which Asness says is “as much an op-ed as it is quantitative research,” is full of ideas worth contemplating.  (For the full effect, don’t skip the extensive footnotes.)

Asness discusses three hypotheses as to why markets are becoming less efficient:  increased indexing, very low interest rates for an extended period, and (the one he favors) a crowd that isn’t wise:

But what if the crowd isn’t independent, but acts in unison?  Well, this has the potential not just to destroy the magical crowd wisdom but to turn it into a negative.

So, has there ever been a better vehicle for turning a wise, independent crowd into a coordinated clueless even dangerous mob than social media?

Social media algorithms “famously push people to further and further extremes,” and Asness sees it in investing as in other things.  But it’s not just “retail” going astray — institutional asset owners also travel in packs, most notably these days in private assets, so it’s good to remember that “no asset class or strategy is immune from the consequences of its own popularity.”  Plus, allocators and overseers get too hung up on line items and use evaluation windows that are wrongheaded.  Those tendencies are fertilized by providers:

The investing world is incredibly adept at explaining to you why whatever has been happening for a while is now going to happen forever.

Ultimately, a less efficient market “should be more lucrative for those who can stick with it over the long-term, but also harder to stick with.”

Asness has foresworn “pure EMH worship,” and in a good interview with the Financial Times, his former teacher Eugene Fama admitted that his efficient market hypothesis has “to be wrong to some extent.”  But how it is wrong and whether you can take advantage of that fact are important questions.

When asked about Asness’ belief that markets have gotten less efficient, Fama replies, “He’s selling products, right?”  But Fama sits on the Dimensional Fund Advisors board, and they sell products too.

AI analysts

Here are two exhibits (smooshed together into one) from a new piece from Sparkline Capital, “AI Financial Analysts.”  As examples, the report describes the use of AI in evaluating executive departures and cybersecurity patents, as well as preparing research reports.  The section on reports points out the difference in using AI for quantitative investing, where “numerical inputs are especially useful,” versus discretionary investing, where:

The text responses may themselves be the desired output.  Rather than have the AI “compress” the supplied information all the way down to a single number, we can have it stop at this intermediate step.  These text responses can then serve as fodder for further analysis by human investors.

In “reimagining the investment firm,” the report lists 29 analyst and portfolio manager functions — and judges which ones “can be completed more efficiently by a [large language model].”  The bottom line:

In our opinion, understanding clearly the complementary strengths of AI and human talent will be crucial for successfully navigating the age of AI.

(See also Seth Godin’s short posting about “Redefining a profession,” in which he asks, “In your work, are you fighting the change or leading it?  It’s hard to see us going back.”)

Human analysts

Sell-side analysts are among the most-studied investment actors, since academics go where they can find data, and recommendations, target prices, long-term growth rates, and earnings estimates offer the kind of time-stamped details that aren’t available for most other decision makers.  As natural language processing and other capabilities have become available, more papers have incorporated qualitative considerations as well.

For example, “Analysts’ use of qualitative forward-looking earnings information,” by Karthik Balakrishnan, et al., compares the risk factors disclosed in IPO prospectuses with subsequent errors in analyst earnings estimates, finding that “the magnitudes of these errors vary with proxies for analyst forecasting ability (experience), analyst reputation (All-Star), and attention (portfolio complexity), but not with analyst affiliation.”  Would an AI analyst pay more attention to those stated risks (and would a portfolio manager make different IPO allocation decisions as a result)?

Speaking of those “All-Star” analysts, why are they influential?  That’s the subject of a paper by Gil Aharoni, et al.  The authors study whether “their status itself generates influence [or] that their quality is superior, consequently yielding them more influence.”  Their work supports the quality hypothesis and finds that there are in effect two cohorts of analysts:  “Never Stars must demonstrate competence, while Current Stars must demonstrate boldness.”

Wham!co

Western Asset Management, commonly known as Wamco, has been rocked by the disclosure that Co-CIO Ken Leech received a Wells Notice from the SEC related to “past trade allocations involving Treasury derivatives.”

Wamco managed $380 billion in bonds, 65% of which was in institutional accounts according to a Pensions & Investments article, “Abrupt departure of Ken Leech raises due diligence bar for Western Asset Management clients.”  The article calls Leech the “institutional face” of the firm, with “significant influence on investments throughout the organization.”  In addition, he was listed as a manager on 25 mutual funds.

Bloomberg reported on the unusual governance arrangement between Wamco and its owner, Franklin Resources (whose stock is down more than ten percent since the news broke), which gave Wamco autonomy not accorded to other Franklin affiliates.

Institutional investors have already started bailing (some of which had the firm on watch lists due to previous personnel departures) and Morningstar has suspended its ratings on all of the mutual funds, so outflows could be substantial.  As these things go, trade allocations will now get a lot more attention during due diligence reviews (and SEC examinations) of multi-product managers, but given the wide array of affiliate relationships in the industry, more attention should also be paid to governance arrangements and expectations.

Influences

Looking back on nearly four decades in the business, Liz Ann Sonders wrote “Songs of Experience: Reminiscences of a Strategist.”  The posting summarizes the people and books that have influenced her the most — and the lessons she took from them.

Other reads

“The Corporate Life Cycle: Managing, Valuation and Investing Implications!” Aswath Damodaran, Musings on Markets.  An overview of the author’s new book.

“Mr. Market Miscalculates,” Howard Marks, Oaktree.

In other words, it’s the primary job of the investor to take note when prices stray from intrinsic value and figure out how to act in response.  Emotion?  No.  Analysis?  Yes.

“The Quant Renaissance: How Alternative Approaches Are Driving the Rebirth of Systematic Investing,” Nina Gnedin and Ori Ben-Akiva, Man Group.  An argument that the “Quant Winter” has passed, because of a normalization of Fed policy plus a proliferation of idiosyncratic approaches driven by alternative data and machine learning.

“A Growing Conflict in Wall St. Buyouts,” Andrew Ross Sorkin, New York Times.

Over the last several years, a new, insidious relationship has quietly developed between the nation’s largest private equity firms, the banks that lend them billions to fund their buyouts and the law firms that advise on these deals.

“Being right versus making money,” James Gruber, Firstlinks.  Five types of investors — rabbits, assassins, hunters, raiders, and connoisseurs — and their winning and losing habits.

“Whatever happened to the wisdom of the bond market?” Katie Martin, Financial Times.

Signal-sniffing algorithms and the greater role of speculative funds are a recipe for jerky market conditions.  A weaker post-2008 ability among banks to absorb shocks does not help.

“A Number From Today and A Story About Tomorrow,” Morgan Housel, Collaborative Fund.

The most persuasive stories are what you want to believe are true or are an extension of what you’ve experienced firsthand, which is what makes forecasting so hard.

“Insider Trading by Other Means,” Sureyya Burcu Avci, SSRN.  Do some “insiders conceal their suspicious trades by publicly reporting them (as they are required to do) in ways that confuse or discourage investigators?”

“Have the insatiable private markets eaten the small cap effect?” Christopher Schelling, LinkedIn.

The effort to find reasonably priced, rapidly growing businesses with durable revenue streams at the lower end of the private equity capitalization range is not like finding a needle in a haystack, but selecting from a stack of needles.

Look away

“The asset management industry compels us to engage with all of these distractions, when the route to better decision making is finding ways to avoid them.” — Joe Wiggins.

Flashback: Up the Organization

Robert Townsend published Up the Organization in 1970.  In one sense, it’s a period piece, especially in regard to organizational roles.  If it’s a position that matters, a “man” is right for the job, while you can count on a “girl” to be there to support him.  The commemorative edition, which came out in 2007, leaves the text as it was originally, while including footnotes from the editor commenting on such outdated language — and helping the reader with some of the contemporary references that have gone stale.

Most of the book is as fresh today as it was then.  Organized as short alphabetical chapters on a broad range of topics, Townsend delivers on the subtitle, “How to Stop the Corporation from Stifling People and Strangling Profits.”  The references to pink message slips for calls to be returned and salary levels may be dated, but the principles aren’t.  The organizational challenges Townsend addressed live on.

Famed editor Robert Gottlieb worked on the original book and in this edition said of Townsend:

He was a secular prophet — somebody who fervently preached exactly as he practiced.  There was no room for hypocrisy in his writing or his life.  Business fads may come and go, but there will always be a need for such pure voices in the wilderness.

Postings

The archives can be sorted by category if you want to narrow down the back catalog.  Here’s an example of a posting picked at random, “Models, Morals, and Management in a Trading Room,” featuring a look at Daniel Beunza’s book Taking the Floor, which has a variety of lessons that extend well beyond trading rooms:

The “performativity” of models in the financial world make them different than models in the realm of the hard sciences.  Two examples, the portfolio work of Harry Markowitz and the Black-Scholes option pricing formula, were descriptive at first and then something more than that, as their adoption shaped how market participants (and markets) behaved.  The theories behind the models became self-fulfilling.

Thank you for reading.  Many happy total returns.