Authored by Patrick Feeley via Substack,
The code most American banks run on was designed in 1959, the year Alaska and Hawaii became states. A committee of government and industry people wrote COBOL so that business programs could be read by people who were not mathematicians, and a good part of it was modeled on FLOW-MATIC, an earlier language from Grace Hopper, a Navy officer. I doubt anyone on that committee thought it would still be running banks in 2026. In 2017 Reuters estimated that about $3 trillion of daily commerce still ran through COBOL. In April 2020, when unemployment claims in New Jersey overwhelmed the state's forty-year-old system, Governor Phil Murphy went on television and asked for volunteers who knew COBOL. The state had to go on TV to find programmers for its own unemployment system.
I bring this up because of a clip I posted last night of Bill Ackman talking with Shane Parrish on The Knowledge Project. Ackman said Cognition, the company behind the coding agent Devin, can rewrite a bank's COBOL "in a matter of days as opposed to many months." Someone replied asking me what I meant when I said commoditized lenders would compete the savings away. It is a fair question and I could not answer it in a tweet.
I do believe Ackman that the savings are real. Inside a large bank the core ledger still runs in batch. The balance a customer sees on the app at noon is an estimate (bankers call it memo-posted). The actual accounting happens overnight, when a mainframe works through a queue of jobs in a set order, posting transactions, accruing interest, charging fees, and producing files that every other system reads the following morning. The programs share data through copybooks, which are record layouts where a field is known only by its position. Cognition described one client where a single taxpayer ID showed up under dozens of different names across thousands of programs. Most big banks were put together through acquisitions, and each acquired bank came with its own core system that management was usually too nervous to shut off, so the old systems just piled up.
.@BillAckman tells @shaneparrish that AI will sharply cut what big banks spend on tech. Rewriting legacy COBOL (decades-old bank code) with @cognition now takes days instead of months. Ackman is right that the cost savings are real. The open question is who keeps them. Banks with …
— Patrick F. Feeley (@PFFeeley) September 30, 2026Replacing all of that has gone badly more often than well. Commonwealth Bank of Australia spent five years and more than A$1 billion replacing its core, and people in the industry consider that one a success. TSB in the UK moved customers onto a new platform in April 2018 and the platform did not work. Customers were locked out, some could see other people's accounts, and service was not back to normal until December. TSB ended up paying £32.7 million in redress and £48.65 million in fines. Cognition's own figure is that roughly two-thirds of COBOL modernization projects fail. With odds like that most banks built layers around the old core and left it alone. JPMorgan expects to spend about $19.8 billion on technology in 2026, and its CFO told investors in February that the priority had moved to "modernizing the underlying application code and data." I would guess a large share of that budget still goes to keeping the layers standing.
Cognition is fairly careful about what its agents can do today. Devin is good at documentation, refactoring, and batch jobs, which are the parts of a migration where you can give the agent yesterday's inputs and outputs and let it keep trying until the new code matches the old results. Cognition estimates batch is 30 to 50 percent of a typical migration. The real-time systems (card authorizations, for example) are still out of reach. Banks also have a security reason to hurry. Anthropic's Mythos model, which can find and exploit software vulnerabilities, had bank regulators in the U.S. and Europe holding urgent calls this spring, and Reuters quoted security experts who named legacy bank systems as especially exposed.
Ackman's harder point came a little later in the conversation. "The problem with money generally is it's a commodity," he said. For loans I agree with him. A company that wants a five-year term loan will collect six or seven term sheets and take the cheapest one, and a bank whose costs just went down will give up some spread to win it. Deposits have never really worked like a commodity, and I think that is where his argument is missing a piece.
The best explanation I have read is from three NYU economists, Itamar Drechsler, Alexi Savov and Philipp Schnabl. Their paper argues that banks have real market power over deposits. When the Fed raises rates, banks raise what they pay depositors slowly and only partway. Keeping that power costs money for branches, bankers and technology, but almost all of the cost is fixed. So deposits end up behaving like long-term fixed-rate funding, which is how a bank can hold thirty-year mortgages without being wiped out every time rates go up. It is also why the industry's net interest margin has barely moved over several decades of rate cycles. The FDIC has it at 3.32 percent today.
Bankers measure this with the deposit beta (the share of a rate increase that gets passed along to depositors). Checking accounts have low betas. Online banks have high ones because, as the St. Louis Fed put it, their customers are looking for yield. During the 2022 hiking cycle the New York Fed found that super-regional banks passed through more than small banks did, while the very largest banks passed through less than either. After Silicon Valley Bank lost $42 billion in deposits in one day, money moved toward size, and the biggest banks did not have to pay more to get it.
Meta's Muse goes right at this. It launched September 8, the same day Cognition announced it had raised more than $2 billion at a $48 billion valuation with run-rate revenue near $900 million. Muse is a personal agent that reads accounts at more than 12,000 U.S. banks and financial apps through Plaid. On Tuesday, September 22, Schwab fell 6 percent, LPL fell 7 percent, JPMorgan and Wells Fargo each fell more than 3 percent, and XLF, the largest financials ETF, was down 2 percent. On Sunday Torsten Slok at Apollo put out a note titled "Is an Agentic bank run coming?" He pointed out that the average checking account pays about 0.1 percent while Revolut, SoFi, Wealthfront and others pay between 3.3 and 5 percent, and he warned that banks "could lose a large share of the cheap deposits they rely on to make loans."
Muse cannot move money yet. The Plaid connection is read-only, and Meta deserves to have that said. I still would not want to be running a bank's treasury desk this month. Most people leave savings at a tenth of a percent because switching is a hassle. Opening a new account takes an afternoon, and nobody wants to be the person who breaks their own direct deposit. If an agent already sees every balance and can fill out the forms, most of that afternoon goes away.
A hypothetical helps here. Bank A has $10 billion of deposits that are really savings, money the customers do not need next month and have not looked at in a while. Rates are at 4 percent, and agents push Bank A's beta on those balances up by ten points. That costs Bank A about 40 basis points on $10 billion, or $40 million a year. Say Cognition saves Bank A $15 million a year on code maintenance, which is my guess and not a published number. Bank A is behind by $25 million, and that is before counting any spread it gives up to keep borrowers. My numbers could easily be off in either direction, but for a bank funded mostly by savings I do not see the code savings covering the deposit cost.
Corporate treasurers dealt with this a long time ago. A company keeps enough in its operating account for payroll and suppliers and sweeps the rest into money market funds or Treasury bills, and banks price corporate deposits knowing somebody is watching. An agent gives an ordinary family something like a corporate treasurer. Next month's bill money will stay in checking. The surplus that has been sitting there since the pandemic probably will not stay at 0.1 percent, and I expect it to reprice slowly at first and much more visibly in the next rate cycle.
Some banks are safer than others. Deposits that run a business's payroll and payables, carry a line of credit, or belong to an owner whose banker actually picks up the phone will not move for half a point, and banks holding those should keep most of what Cognition saves them. Banks that fund themselves with rate-shopping savings and win loans on price are in a worse spot, since they will likely pass the savings on to borrowers and pay more to depositors at the same time.
This part relates most to my own work. Ackman was talking about big institutions that own their code. Most American banks rent theirs. The Kansas City Fed found that Fiserv, Jack Henry and FIS together served more than 70 percent of banks in its 2022 survey, and 61 percent of banks had used the same core provider for over ten years. When code gets cheaper for a community bank it gets cheaper for the vendor first. Whether any of it reaches the bank depends on a contract that may run for years and on how hard those three companies compete at renewal. Agents do not have to wait for any contract, and they will reach a community bank's depositors the same week they reach JPMorgan's. On September 22 the market sold banks with deep local relationships about as hard as it sold the ones without them, and in some cases I think that was a mistake.
If I were looking at a bank stock this month I would not spend much time on efficiency ratios, since nearly everyone's will improve. I would look at what the bank's deposits did from 2022 to 2024, which is the closest thing to a live test of stickiness the industry has had. I would want to know how much of the deposit base is operating money and how much is savings nobody has checked in years, and I would want the renewal date on the core processing contract. At Sargasso Capital Management we spend most of our time on small and mid caps where there is a wide gap between what AI can do and what the company has actually put to use. With banks that gap only matters if the customers are still there when it closes. Bill Ackman is right that the code is about to get a lot cheaper. I am just not convinced most banks get to keep much of it.
A note on sources. Bill Ackman's comments are from his September 2026 appearance on The Knowledge Project with Shane Parrish. Cognition figures come from its September 8 funding announcement as reported by Reuters, TechCrunch and SiliconANGLE, and from its April 2026 post on COBOL modernization. Muse details come from Meta's launch materials and coverage by TokenPost, CNBC and Zacks. The Torsten Slok note was reported by CoinDesk on September 28. COBOL's origins and New Jersey's April 2020 call for COBOL programmers are widely documented. Banking data comes from the FDIC Quarterly Banking Profile for the second quarter of 2026, the Federal Reserve Banks of New York, St. Louis and Kansas City, the Federal Reserve's review of Silicon Valley Bank, JPMorganChase's February 2026 Company Update, the FCA's TSB enforcement notice, Reuters reporting from April 2017 and April 2026, and Drechsler, Savov and Schnabl, "Banking on Deposits" (Journal of Finance, 2021). Bank A is a hypothetical example.
This post is for informational and research purposes only and does not constitute investment advice or an offer to buy or sell any security.
Tyler Durden Wed, 09/30/2026 - 16:20
For years, conspiracy theorists have insisted that the Federal Reserve is run for the benefit of a foreign power. It turns out they were - at worst - only half wrong: it's just that the foreign power in question wasn't getting the FOMC minutes through the front door, but via a lovesick Fed economist, an online dating site, a set of silk scarves, and a gym-obsessed "Ph.D. student" who, according to US officials, was actually a Chinese intelligence officer.
Readers may recall the name John Harold Rogers, the former senior adviser in the Fed Board's Division of International Finance whose arrest we first reported back on Feb 1, 2025, and whose case we then revisited in May 2025 under the only headline the facts allowed: "Teaching Or Treason?" In February a DC jury gave its answer (sort of): guilty of lying to investigators, not guilty of conspiracy to commit economic espionage. And in July, as we reported on July 16, Judge Dabney Friedrich sentenced him to 38 months in federal prison.
That, it seemed, was that. But now a lengthy CNBC investigation, built on text messages, investigator audio recordings, trial exhibits and copies of the Fed documents in question, fills in the gaps. And the gaps, it turns out, were where all the good stuff was hiding.
"The facts of this case read like a spy novel. There was sex. There was money, manipulation, secret meetings with shady characters in China, fake identities, lies, trickery, and deceit. But this was real life." — Federal prosecutor Nicholas Hunter, in his address to the jury Meet "Hummin Lee"The story begins in 2013 at a conference in Shanghai, sponsored by, of all things, the Federal Reserve Bank of Dallas, on the campus of Fudan University. During a session break, a tall, broad-shouldered young man introduced himself to Rogers as Hummin Lee, a Ph.D. student. US authorities believe his real name is Jin Chuan, and that he is a Chinese spy.
A photo of Hummin Lee flexing his muscles that Lee sent to John Rogers in a text message on Aug. 4, 2018; Source: Department of JusticeWhat followed was textbook slow-burn cultivation. Lee sent birthday wishes and Christmas presents (including, in a detail no novelist would dare invent, a copy of Sun Tzu's The Art of War), invited Rogers to lecture in China, and kept tabs via a Facebook account featuring shirtless gym photos and, in total, six friends. After Trump won in 2016, the gift of a traditional Chinese painting arrived with a question attached: would Trump label China a currency manipulator? Rogers, to his credit, dodged.
Then came New Year's Eve 2016. Rogers, a divorced father of four grown children who later said he "kind of felt like a loser" that night, logged on to AsianDating.com and met Yu Liu, known as Yu Yu, a young woman in Shanghai 24 years his junior. Her messages were beautifully written in English. Only later did Rogers learn she didn't speak English at all.
Yu Liu and John Rogers are seen at a restaurant in a photo from a text message Rogers sent to Hummin Lee on Nov. 20, 2017; Source: Department of JusticeRetired 35-year CIA veteran Ralph Goff, who had no role in the case, told CNBC what everyone reading this is thinking:
"If this woman wasn't a plant by Chinese intelligence into that web dating site, then I'm sure that they had a meeting with her soon after that contact."CNBC notes it is unclear whether Yu Yu has any links to Beijing; she did not respond to questions. What is clear is that from that moment on, Rogers suddenly very much wanted to travel to China, and his generous friend Lee was only too happy to arrange it. "Your related expenses in China will all be covered by us," Lee wrote.
The full-service handlerLee didn't just book flights. He planned itineraries and arranged lectures. When Yu Yu needed cash, Lee wired her 5,000 "of my own money. Don't worry." When her then-husband demanded 80,000 yuan to sign divorce papers, Lee told Rogers to "go to sleep and take a good rest man," and seven hours later texted a photo of a freshly issued divorce certificate. When Rogers married Yu Yu in Hong Kong in March 2018 (she became pregnant on his second trip to China), Lee organized the logistics and witnessed the prenup. When the baby arrived that summer, Lee met Rogers at the airport and went with the couple to the hospital.
Rogers described him to Fed investigators as a "super nice guy but so nice it was a little bit spooky." Which is a useful working definition of a case officer.
"I owe everything to him. The baby wouldn't have happened, the relationship with my wife wouldn't have happened... And I love him like a brother." — John Rogers, to investigatorsLee also introduced his "boss," a "Professor Cui," who pressed Rogers for Fed rumors and gossip and, per Rogers, offered him packets of cash (which Rogers said he declined). When the FBI searched Rogers' home years later, agents found $55,000 in cash, including some in a white plastic grocery bag at the bottom of a closet under a pile of women's clothing. Investigators could not establish where it came from.
The "questions"With Rogers now thoroughly in his debt, Lee's curiosity about the Fed grew. The "classes" Rogers taught in China increasingly took place in hotel rooms with a handful of attendees. At one point Lee sent over a typed list of questions, among them:
"I'm wondering if you could help us collect answers from your colleagues or documents and teach us when we meet," Lee added, helpfully. Rogers testified these were "bread-and-butter" textbook questions. Which, to be fair, they mostly are; it's the "from your colleagues or documents" part that is less textbook.
Rogers then emailed Fed colleagues asking for material, remarking with remarkable self-awareness: "Part of it, I think, is that they want me to reveal some kind of Fed secrets they believe exist." One document he received was marked "Nonconfidential/Internal FR." According to a federal agent's testimony, Rogers changed the heading to "Nonconfidential/External" and forwarded it to Lee. On another occasion he asked a colleague for Fed thinking on trade policy uncertainty, adding: "If there is something in writing that is not allowed to be shared, please send that just for me to catch up." He received a document on washing machine tariffs, forwarded it to his personal Gmail, and from there to Lee.
Then, in June 2019, Rogers printed two documents: one from a category generally covering the staff forecasts prepared for the FOMC and open market operations (per the indictment, "Class II" FOMC information), and a set of bullet points for a briefing to Fed governors ahead of an FOMC meeting. The next day he flew to China. Three days later, he photographed a "class" in a Beijing hotel room: Lee, another man and a woman, posing by a big video monitor. Prosecutors later said both men worked for Chinese security and intelligence services.
How valuable was any of this? That was the heart of the trial. The defense argued Rogers passed nothing of value and was "a naive academic." Prosecutors argued, in their sentencing memo, that the restricted information "could allow [Beijing] to make untold sums of money by trading with its roughly $1.5 trillion in U.S. Treasury securities and related instruments." The jury, notably, was not persuaded on the espionage count. Then again, as one might note, you don't need a Fed briefing book to front-run the Fed; you just need to read Jerome Powell's press conference... or a Nick Timiraos article (well, during the Powell years at least... it appears he isn't on Warsh's Christmas Card list, or Bessent's for that matter).
Undone by... InstagramIronically, it wasn't China's MSS that brought Rogers down, but garden-variety sextortion scammers. After his marriage imploded (the Vienna, Virginia police were called after a physical fight between Yu Yu and a masseuse Rogers had invited to live in the house; "I swear to God we're just friends," Rogers told officers), he went back online.
On Instagram he traded nude photos with accounts showing attractive women, photos taken with his Fed-issued phone, sometimes in the gym at Fed headquarters.
The people behind the accounts then demanded money, and threatened to send the photos to then-Fed Vice Chair Richard Clarida. They also threatened to kidnap his 18-month-old daughter. On Super Bowl Sunday 2020, Rogers told his superiors, who referred him to the Inspector General, where the Fed's cyber team was already uneasy about his China trips and his friendship with Lee.
In a 2.5-hour recorded interview, OIG agent Alan Hershkowitz asked if he had ever taken a nude photo of himself. "No," said Rogers. Pressed, he conceded: "I can't rule it out." Asked whether he had ever shared restricted Fed information outside the Board, he answered: "Never." Prosecutors proved he had sent such a document to a co-author at Fudan. That single word is what cost him three years.
Perhaps the most damning detail, though, is about the Fed itself: Rogers was quietly allowed to resign in 2021, the circumstances were never made public, and he promptly went to teach at, yes, Fudan University, where according to prosecutors he earned at least $900,000 in salary and grants between February 2022 and his arrest. It then took more than three years for the FBI to show up at the bus stop.
"All of the book smarts he had, all of the articles in his resume, they don't prepare you for spies." — Stephen Saltzburg, Rogers' attorney Is the Fed a sieve?If the Rogers case were an isolated incident, one could shrug it off as a lonely man getting played. But the Fed's own watchdog suggests otherwise:
So: a Fed that can't catalog its own critical assets, can't stop a retiree from USB-sticking FOMC files out the door ahead of a trip to a restricted country, and is targeted by Chinese hackers from the outside and (allegedly) cultivated insiders from the inside. Asked by CNBC whether China is still trying to penetrate the Fed, Goff's answer was one word: "Absolutely." The Fed and its IG declined to comment.
We leave the last word to Rogers himself. Asked at trial who he believed Lee really was, he replied: "A spy." And how did that make him feel?
"Duped. He was my friend."
Which, come to think of it, is also how most savers feel about the Fed.
Tyler Durden Wed, 09/30/2026 - 16:15