Mechanical Turk

When was the last time you talked to someone at a bank? Maybe it's been years.

The biggest innovations in finance have been about automation. The ATM: automated teller machine. ACH: automated clearing house. e-signing. Automation drives modern finance.

You can view modern financial institutions as automation machines. The finance part of bank operations is sort of incidental. This allows them to do more and more business without adding customer-facing employees. Employees are expensive. At least that’s the idea.

Automation can ease the need for legions of customer service troops, but banks still need employees to keep the machine running. A lot of this has meant shifting from customer-facing employees to code-facing employees. It’s the spiritual successor to the Mechanical Turk — an “automated” chess playing machine from the 1770s that duped human opponents. The ruse was that it was person inside the machine doing the chess playing. Like the Mechanical Turk, bank employees moved inside the machine.

JP Morgan does a decent job at the automation thing. In 2024, the shop made $560 thousand per employee. Is that good?

Its increased earnings per share more than 60% in two years. Pretty good.

Could it achieve similar results just by making customer-facing employees better at their job? That sounds hard. Does anyone do this? Does anyone make money in "high-touch" businesses anymore?

Financial clearinghouse Marex tries. It specializes in commodities and clearing services and stakes its business on its people.

Giddyup

Employee-driven finance sounds like some archaic business involving buggies and horses or saddles…

Hermes makes saddles. It's a famously high-touch business. The touch being by real, live people. It trades in leather goods — some saddles, mostly bags. If you're a customer, your personal sales associate knows your name and tastes. An artisan — just one, probably French — made your entire bag. No automation. When your bag wears down, they'll fix it. The products are fantastically expensive, but they hold or increase their value over time.

Compare this to, say, another luxury bag maker, LVMH. LVMH puts less emphasis on the craft and more into marketing.

What about the economics of the two businesses? Mechanization, automation wins, right? Hermes made 30% operating margins last year compared to 23% for LVMH. Score one for the humans.

So what? What do luxury bag ateliers have to do with financial companies?

While LVMH found its lane leveraging marketing, branding and using more automated production methods to carve its luxury empire, Hermes set out on a crusade to achieve the highest quality for its customers. This costs more in artisans and less in sales and marketing.

JP Morgan and just about every other large financial institution took the LVMH route. They’ve inserted automation at every junction. Sometimes this has meant taking some shortcuts to simplify things. Marex has sharpened its focus on tailoring its products to its customers’ needs. This has meant keeping people in the loop.

Saddles to Solutions

Marex moves risk. It does this through clearing — what happens after you click "buy" or "sell" to make a trade. It does this through acting as an agent on its customers' behalf. It does this through market making — matching buyers and sellers. And it does this through its solutions business where its financial workshop lives.

If, say, a coffee producer wants to protect itself from the risk of bean prices falling before it sells its next crop. And if it wants to tie this to the times of the season it actually sells the beans, Marex does this in its solutions house. Otherwise, our bean grower would settle for an off-the-shelf product that is tied to one specific day during the season — an automated but flawed solution.

No one buys financial services from Marex because they love the brand. Customers do, though, love the service. That takes people. How do the economics of high-touch financial concierge work out?

Marex employees contribute twice as much revenue as JP Morgan’s. They do twice as much per employee as Hermes. They bring in twice as much per employee as AI automation darling Palantir.

On profitability though, JP Morgan earned 45% profit before tax last year. Marex managed just over 19%.

Does that mean the model doesn’t work in finance?

Small But Mighty

Some of the gap stems from JP Morgan’s heft. It earns more because it’s larger. It’s less efficient though. Marex hit a return on equity of 27% in 2024 versus JP Morgan’s 18%. Marex achieves consistent outperformance here. This ought to lead to punchier future growth. Investors prefer higher margins to lower ones, but growth bears much more on value. Higher growth should mean more valuable. The market values JPM and MRX about the same.

And there's reason to think Marex can do even better.

Part of Marex’s strategy is to grow by buying specialty and regional financial products companies. Some of the groups it’s brought in aren’t yet profitable. Many are simply below standard. It’s succeeded in its older vintages of deals bringing profitability up from low teens margins to mid-20 percent. It’s made progress on this front. It boosted its profit margin from 16% to 20% in 2024.

Marex’s high-touch strategy doesn’t pull the same emotional threads that Hermes’ achieves, but it does sort for a certain type of customer, ones who value service rather than ones looking for a deal.

Part of that service is working with execs at their customers to integrate their service. This takes 6-12 months — a big commitment. Once Marex embeds with its customers, switching becomes daunting and costly. This lends durability to the business.

There aren’t many places that could replicate Marex offering anyway. Customers would have to cobble together different services from different providers.

Gaps in the Seams

High profitability per employee. High-quality customers. High switching costs. These all contribute to the quality of Marex's business. The market, though, has priced Marex closer to a saddlery maker in the auto age. It's unimaginative expectations create an opportunity for an attractive investment.

Disclaimer: None of this is investment advice. It's meant to illustrate ways LCM thinks about investing. Things that LCM decides are good investments for LCM and its clients are based on many criteria, not all of which are covered here. Some or all of LCM's ideas may not be suitable for other investors. LCM does not recommend investing either long or short any position mentioned. LCM may own positions in some of the companies mentioned. Some of its ideas will lose money — investing entails risk. See full disclaimer here.