Dimon flags new risks

Jamie Dimon used JPMorgan Chase’s annual letter this year to do what he often does: talk like a banker who thinks the real danger is not the thing right in front of markets, but the thing markets are pricing as if it cannot happen. In the letter released April 6, he put geopolitics at the top of that list. He pointed to the wars in Ukraine and Iran, broader conflict in the Middle East, terrorism, and rising tension with China. He warned that oil and commodity shocks could keep inflation sticky, push rates higher than investors expect, and expose how vulnerable high asset prices look when the world stops cooperating. (jpmorganchase.com) That part was familiar. Dimon has spent years arguing that markets are too calm about hard problems. What felt newer was the way this year’s warning bundled old macro risk with two market-structure stories that are much more specific to Wall Street right now: artificial intelligence and private markets. CNBC reported that Dimon’s letter singled out all three areas together, which matters because they are not separate themes. They are starting to collide inside the same firms, on the same balance sheets, and in the same compliance systems. (cnbc.com) Private markets are the easiest place to see the shift. Big banks and asset managers have spent the past few years pushing deeper into private credit, private equity, and other assets that do not trade on public exchanges. JPMorgan’s own commercial and investment bank described private capital as part of its growth agenda for 2025, alongside AI adoption and digital assets. That is the important context for Dimon’s warning. Private markets promise growth just as public markets look crowded and expensive. They also bring thinner price discovery, less transparency, and a much harder surveillance job when information moves unevenly. (jpmorganchase.com) AI makes that surveillance problem stranger, not simpler. JPMorgan has been expanding AI use across its businesses for years, and its executives now describe the technology as something being scaled across operations rather than tested at the edges. That means more automation in research, trading workflows, client service, and internal decision systems. It also means more chances for a bank to create products that look less like a traditional account and more like a live information engine. Once that happens, the old distinction between a market, a media feed, and a betting interface starts to blur. (jpmorganchase.com) That is why Dimon’s separate comments about prediction markets landed so hard. In a CBS interview aired March 31 and reported by Yahoo Finance and others, he said JPMorgan could one day offer something in that area. He immediately added limits. No sports. No politics. Strict rules on insider information. He also said that, for the most part, prediction markets are more like gambling than investing. Still, he did not dismiss the format. He said the bank is studying how it might work. (cbsnews.com) That is the real story here. The biggest bank in the country is not rushing into prediction markets because it has suddenly become a casino. It is studying them because event contracts are turning into a new financial interface. They pull in retail attention, react instantly to headlines, and create obvious temptations around nonpublic information. The CFTC has been leaning into that reality, issuing a staff advisory on event contracts in March and an enforcement advisory in February after cases involving misuse of nonpublic information and fraud tied to prediction markets on Kalshi. (cftc.gov) So Dimon’s annual letter and his prediction-market comments are not two unrelated messages. They are one message in two forms. The world is getting more unstable. Finance is getting more real-time. Banks are moving into businesses where the next edge may come from private assets, machine-driven analysis, or contracts that let people trade on whether something happens at all. If JPMorgan ever builds that kind of product, the hard part will not be attracting users. It will be deciding which events are safe enough to turn into a market, and then proving that nobody inside the bank knew the answer first.