Two pillows

Jeff Davis is a veteran bank analyst. The views and opinions expressed in this piece are those of the author and do not necessarily represent the views of S&P Global Market Intelligence; Mercer Capital, where Davis is the managing director of the financial institutions group; or StillPoint Capital, where Davis is a registered representative.

I believe it was spring 1987 when Charlie Sanford and Eugene Shanks, respectively CEO and president of Bankers Trust, spoke to my corporate finance class at Vanderbilt University about the budding changes for banks on Wall Street that had seen swaths of their commercial business move to the capital markets.

In 1986, the Federal Reserve created a carve-out for a handful of Wall Street banks to engage in "investment banking" to underwrite and deal in ineligible securities (as opposed to eligible securities, which entailed US government bonds, general obligation municipals and a few other exceptions). This was done by creating an exemption to Section 20 of the Glass-Steagall Act that would be loosened a few more times before Glass-Steagall was formally repealed in 1999.

The pecking order that emerged, I think, was that JPMorgan, Chase Manhattan and Citigroup Inc. (then Citicorp) focused on investment-grade debt and, for JPMorgan, later equities. Chemical Bank, whose management team rolled up multiple money center banks, pushed into leveraged finance, and Bankers Trust focused on trading as an early pioneer in structuring derivatives for clients and trading the contracts. Bankers Trust would be acquired by Deutsche Bank in 1999, in part because its large, if not inexplicable, exposure to Russia soured when Russia defaulted on its debt.

Maybe it was hyperbole, but one of the things that Sanford or Shanks said that I remember was that new analysts received two pillows in a drawer when they started with the bank. The humor for the class — or warning — was like the old Wall Street saw that if you don't come to work on Saturday, don't bother coming in on Sunday because you won't have a job. Or, stated differently, it took a lot of effort to remake Bankers Trust.

There are similarities between the 1986 Section 20 exemption, which allowed banks to participate in securities underwriting, and the introduction of artificial intelligence.

Bloomberg recently ran a story that got me thinking about the "two pillows" story. Three men, still in their 20s, two of whom began their careers as analysts, started Rogo Technologies in 2021 to code artificial intelligence to take the "grunt work" out of being a junior banker. Somewhat contradictory, Bloomberg ran a story late last year about why we cannot quit Excel, though it seems to me AI will assume control of Excel, too, in time.

It is not a perfect analogy, but there are similarities between the Section 20 exemptions and the introduction of AI onto Wall Street. Section 20 did not immediately transform Wall Street, but it pointed to what was coming, with the business transformed by the reintroduction of the money center banks into the securities business, and the competition for who would capture the economics. The punchline is that Morgan Stanley and Goldman Sachs Group Inc. would become bank holding companies, too.

AI is directionally similar, except it is not about regulatory arbitrage but labor arbitrage. The Bloomberg pieces about Rogo Technologies and Excel are telling. Rogo and companies like it will automate the work that junior bankers have historically done manually, though the latter is a reminder that Wall Street, for now, remains wed to spreadsheets to grind numbers.

If AI works as advertised, there will be fewer bankers and other professionals for sure. Those that make the cut should be exceptionally good.

Banks have relied on armies of analysts to do the repetitive work — pitch books, models, comps — that also doubled as training. Analysts were well-paid to do the grunt work, and they were paid twice, when including the experience gained. If AI works as advertised, the traditional apprenticeship model may break in most, if not all, white-collar professions.

If so, what then? One aspect is that there will be fewer bankers and other professionals for sure. What is unclear is if this will occur via mass layoffs, a gradual reduction that follows an aging society or a combination of the two. Those that make the cut should be exceptionally good.

Another aspect is that the adoption of AI should be deflationary, maybe radically so, though the economics for those not culled may continue to be very good or even better.

That said, I do not think the "two pillows in the drawer" culture will disappear in spite of repeated efforts to end it. There is too much money at stake on Wall Street to let opportunities slip away that often have a narrow window to grasp. AI may widen dispersion among firms, much as Section 20 widened dispersion between banks that leaned into capital markets in the 1980s and 1990s and those that did not.