Jamie Dimon Wants This Guy to Disrupt JPMorgan

Meet David Hudson, tasked with building innovation inside the bank to protect it from outside threats. His title might as well be chief disruption officer

From

David Hudson exhaled slowly and decided to take the leap. He’d just been promoted to a new job that put him in charge of spearheading disruption within JPMorgan Chase & Co.’s trading businesses. Now he stood before Chief Executive Officer Jamie Dimon in the 48th floor boardroom of the company’s Park Avenue headquarters in Manhattan to see what he could get away with. If the bank really wanted to anticipate clients’ needs and attempt risky moonshot projects, he told Dimon, his budget had to expand: “I want an extra $50 million.”

After five minutes of questioning, Dimon gave the go-ahead. Then came the hard part. For the first moonshot that Hudson, 42, had in mind—a platform to liberate asset managers and ­smaller banks from having to employ traders—JPMorgan had to switch into aggressive hiring mode. Recruiters would be needed, at least 10 of them, to find the coders and traders to build and operate the system. Office space had to be found to house the group and salesmen to hawk the thing once it existed.

David Hudson
Photographer: Ana Cuba for Bloomberg Markets

In getting his bosses’ backing, Hudson assumed an enormous responsibility. “I only figured out later that it was a psychological game, that by them saying yes to everything I asked for, they made it my problem,” he says. “In other words, ­money’s not the problem. My capacity to deliver is.”

His problem is a high-class one. If you had to pick one winner among banks in the post-financial crisis era, it would be ­Dimon’s JPMorgan, a behemoth with $2.55 trillion in assets and operations in 60 countries around the world. Its Wall Street division is a top-three performer in most major asset classes, including government bonds, currencies, corporate credit, and equities—and last year it produced a 16 percent return on equity, among the highest of its peers. The markets businesses share a lot of resources, so keeping them all humming is key to the company’s profitability. Perversely, this means JPMorgan has the most to lose should clients suddenly decide to do business in a different way—say, if a startup creates a cheaper or easier way of matching buyers and sellers.

QuickTake: Robots Make Trading Speedier With Magnified Mistakes

This is where Hudson comes in. The South African-born coder-turned-manager was named global head of markets execution last year, which essentially means he’s responsible for making sure that if new technology upends trading patterns, it’s JPMorgan that does the disrupting.

There’s a simple reason Wall Street banks are adopting the talk and practices of Silicon Valley. The barriers preventing potential usurpers from storming the gates of finance have collapsed because of a convergence of technologies from machine learning to cloud computing. For decades, it took investments of $50 million for a startup to reach breakeven, Goldman Sachs Group Inc. Chief Financial Officer Marty Chavez told a Harvard audience in January. Now that figure is $3 million. The company he most wants to emulate: Google.

Hudson, formerly finance chief of the markets business, reports to Daniel Pinto, head of the corporate and investment bank and mastermind of the division’s strategy. Both are based in London. A phrase scrawled on a whiteboard in Hudson’s Canary Wharf office encapsulates his mission: “It’s the incumbent’s job to find the innovation before the innovators find distribution.” Anyone who makes a living in the markets today should take heed.

Hudson’s education in the vagaries of technological disruption began on a pleasant London day in May 2002, when his ­employer at the time, an electronic foreign exchange platform called Atriax, suddenly folded. He’d left Johannesburg in 1999 after learning Cobol, a computer language for business, in the hopes of making his fortune fixing the Y2K bug. He dabbled in mobile text messaging and later found himself a senior developer at Atriax. The startup was supposed to have an edge against rivals because it was backed by the world’s biggest players in FX, including Chase Manhattan Bank, Citigroup, and Deutsche Bank, but it was poorly run and outmaneuvered by hungrier rivals.

The day pink slips were handed out, Hudson had just bought his first home, a flat in Bedfordshire an hour’s commute from work. Suddenly he was the last man standing in an empty office, kept on for a few months to sell whatever wasn’t bolted to the floors. When it came time to unload the computers and office furniture, he piggybacked onto a public bankruptcy auction held for Enron Corp. “That really stuck in my mind,” he says during an interview in a corner office on JPMorgan’s New York trading floor. “This company should’ve won, but we didn’t because we weren’t really trying that hard. I sold most of the kit on the Enron auction, which felt somehow appropriate.”

Not long after that, JPMorgan, which had been an Atriax investor through its merger with Chase Manhattan, asked Hudson to join as a business manager in its London foreign ­exchange business. It was 2003, and the bank was just coming to terms with its failure to adopt electronic trading in currencies, the biggest and most liquid of the world’s markets. While electronic brokering systems were faster and less ­error-prone than people, offering audit trails and better transparency, JPMorgan was heavily invested in traders and salespeople whose primary technology was the phone. The bank slipped from a top ranking in FX to ninth by 2006.

So Hudson came aboard to help incorporate electronic trading in the fixed-income business, and then he had managerial stints in equities structured products and fixed-income hybrids, two units that were wound down after the financial crisis. He left to be chief operating officer for equities sales at Nomura Holdings, then returned to JPMorgan’s foreign exchange unit to help global head Troy Rohrbaugh, ultimately becoming CFO of the trading businesses until his promotion last year.

QuickTake Q&A: Why Machine Learning Models Often Fail to Learn

In the years after the financial crisis, JPMorgan gained share in fixed income and equities trading, becoming the world’s top Wall Street company by revenue. By 2014, Pinto, who started his career as an FX trader in Buenos Aires at a predecessor firm of JPMorgan’s, had become sole head of the bank’s Wall Street ­operations. Last month, a global survey of currency users ranked the bank at No. 1 in FX. Despite JPMorgan’s lofty perch, the memory of previous missteps still stings in the minds of Pinto and Hudson—lodged deep in the ancient, reptilian-era center of the brain responsible for loss ­aversion—who are determined to prevent complacency from settling in.

“The fear is we’ll wake up one day and we’ll be like we were in 2003 in FX, not embracing change, focusing on the things we’re good at instead of the things we’re not good at,” Hudson says. “We used to want things to stay the same because we’re leaders in most spaces. That’s just completely naive because someone who has nothing to lose will do whatever they can to build market share. So our job is to continually be that person and disrupt ourselves.”

As for the pace of that change, JPMorgan would rather be too fast than too slow, according to Pinto. “I’d rather make the mistake of accelerating the change than by waiting too long,” he says in an interview. “We’re not going to try to squeeze the last drop from the success of the past, compromising the future.”

 

In person, Hudson looks the part of an anti-trader: horn-rimmed glasses, fuzzy beard, and his hair in an outgrown Caesar, like an architect or a liberal arts adjunct professor. He speaks the dispassionate language of Silicon Valley, dropping references to The Innovator’s Dilemma, Clayton Christensen’s seminal book on why corporations rise and fall, and Jeff Bezos’s annual letter (he’s read the latest one 20 times). His London home is, of course, wired so that he can run every function from his phone.

He splits his job into three segments: making sure the tools used by JPMorgan’s trading personnel are up to snuff in an age when expectations are set by Uber Technologies Inc. and ­Amazon.com Inc.; doing the same with client-facing products; and making those moonshots.

David Hudson
Photographer: Ana Cuba

When it comes to big, radical ideas, Hudson asks his staff every Monday morning: If you had $100 million and WeWork space, how would you take on JPMorgan or lessen its grip on clients? In practice they become hackers, testing JPMorgan’s business ­model for weaknesses. But instead of exploiting them, they build a solution, and if it catches on with clients, it gets embedded in the organization. (This isn’t Alphabet Inc.; nobody’s trying to cure death or create fuel from seawater.)

Their first project, one that’s being built now and may go online later this year, is a trading platform for asset managers and regional banks, institutional clients who are suffering as investors flock to low-fee index funds and regulatory costs squeeze margins. (A chunk of the $50 million Hudson requested is going toward this effort; the bank won’t disclose exactly how much.) Hudson holds up a paper printout that distills the activities of both types of companies into a dozen or so categories. Only a handful, such as gathering assets, is something these companies truly need to do, leaving the rest—things like executing trades and back-office settlements—better handled by someone like JPMorgan.

“An asset manager or regional bank can call us up and say, ‘Buy us €100 million of government bonds,’ ” Hudson says. “We’ll find the best price on the market, which may or may not be from JPMorgan, and buy it on their behalf, so now they don’t need traders. We’ll charge them a commission, do their best execution reporting, provide that as a service.” In other words, the pitch is: Fire your traders, stay in business.

Another idea being studied would be to create products powered by the bank’s internal risk management platform, known as Athena. Regional banks could plug into JPMorgan’s services to, say, value a portfolio or receive data feeds.

People tend to overestimate change in the near term and underestimate it over the long term, Hudson says. In fact, life will be pretty similar a year from now but probably really ­different in 10 years. (Recall that the first iPhone made its debut in June 2007; today there are more mobile devices than human beings.)

A business that’s ripe for change is debt capital markets, where companies borrow money by creating bonds. An ­automated auction technology could someday replace the current, ­human-heavy process for issuers who have high credit ratings, Hudson says. “If there is a way for us to build Bonds ‘R’ Us, then we want to be a part of that rather than someone else,” he says. Such a future would be devastating for the investment bankers who currently make a tidy living performing just such a function.

This story appears in the June / July 2017 issue of Bloomberg Markets.
Cover artwork: Zachary Walsh

Still, in an industry hyperfocused on the next quarter, it’s difficult to say whether Hudson’s more ambitious bets will ever pay off. The bank could decide to scuttle its new trading service, or the service might simply not catch on with asset managers. In 2014, Hudson was part of a team that created an aggregation platform to help trade interest-rate swaps that clients haven’t touched. “My group is an area where it’s OK to fail,” he says. “I guess if I fail too much, Daniel will decide if the model is OK, or if it’s me.”

Automation, Hudson says, is likely to reduce the number of workers across finance and change the composition of those who are employed to include more technologists and ­fewer ­traders. Last year the bank created a machine-learning program to interpret commercial loan agreements, eliminating 360,000 hours of legal work a year. Even more modest advances, such as big-­data-powered tools for institutional salespeople that Hudson is working on, will have an impact. “People are free to spend more time with clients and add more clients, or we’ll need fewer of them,” he says. “Either way, the result is you’ll get efficiency gains.”

As we wrap up our hourlong conversation, Hudson mentions that his parents were journalists in South Africa. Feeling a pang of anxiety, I ask him if machine-learning programs could automate the writing of financial news. His brow furrows, and he’s quiet for a minute. His answer? Feed the machine my interview notes, and it could probably produce a rough draft.

Hudson’s ideas may go nowhere, or they could change the world. In the meantime, as long as his boss is willing to sign $50 million checks, a tiny seed of disruption is planted wherever his mind takes him.

Son covers finance at Bloomberg News in New York.