Prediction Market Compliance: Why Banks, Fintechs, and Society Are Being Forced to Choose

By M. Mahmood | Strategist & Consultant | mmmahmood.com

TL;DR / Summary

As I was researching Prediction Market, I came across two interesting facts: On August 13, 2026, a Financial Times report revealed that JPMorgan had quietly ended its banking relationship with Polymarket back in October 2025 over regulatory concerns, and that same day a King County Superior Court judge in Seattle ordered Kalshi to geofence Washington state residents out of its sports, election, and mentions contracts entirely, with $120,000 in daily fines starting September 2 if the company fails to comply in time. It's very clear to me that neither of these are policy debates happening somewhere in the abstract, since both are institutions, a bank and a court, independently deciding they can no longer treat prediction markets as an ordinary line of business the way they might treat any other fintech client. 

Since, nothing makes sense with out facts, here is a data point that should worry every consumer lender in the country. A recent survey of 1,000 active prediction market traders found that 51% funded their trading positions with credit cards or other borrowed money, and among that group, 88% of them ended up losing money.

The Bank That Said No, and the Judge Who Agreed

JPMorgan told Polymarket to find a new banking partner back in October 2025, and the decision only became public knowledge this month once the Financial Times reported on it, a detail Reuters independently confirmed through a separate source close to the matter. Polymarket told CNBC it still maintains a close, active relationship with JPMorgan across multiple entities and operational integrations, which matters here because it shows the bank never actually abandoned the category as a whole, choosing instead to exit one specific account relationship while quietly preserving enough of a foothold to potentially underwrite Polymarket's future IPO, a possibility several outlets have since confirmed the bank is actively exploring.

A judge in King County Superior Court reached a strikingly similar conclusion through an entirely different process and for entirely different reasons. Back on July 21, 2026, that judge ruled that Kalshi likely violates Washington's existing gambling laws, following a lawsuit brought by the state's Attorney General, and by August 13 that finding had hardened into a binding order requiring Kalshi to block Washington residents from trading sports, elections, politics, entertainment, culture, technology, science, and mentions contracts specifically, while still leaving commodities, climate, and finance contracts untouched. Kalshi asked the state's Court of Appeals to pause the order and was turned down, and the compliance deadline of August 19 lands just before the NFL season kicks off, which is not a coincidence given how much of the platform's trading volume tends to run through sports contracts during football season. Washington is not an isolated case either, since a Nevada judge banned Kalshi from operating in that state back in March 2026 for lacking a required gambling license, meaning two separate state courts have now reached the same practical conclusion even while the CFTC continues to insist its own federal jurisdiction over this category is exclusive.

The Debt Problem Nobody Is Pricing Yet

The part of this story that has received far less attention than the bank exit or the court order is arguably the part that should worry consumer lenders the most. A survey of 1,000 active U.S. prediction market traders found that 79% of them lost money over the past year, with 27 percent losing more than $500, and more strikingly, 51% said they funded their trading positions using credit cards, personal loans, or other borrowed money, a group whose loss rate jumped all the way to 88 percent compared with 69% among traders who used their own cash. Consumer finance expert Erica Sandberg, who worked on that study, summarized the problem bluntly by noting that borrowing money to place a bet, whether through a credit card or a personal loan, is a universally bad financial decision regardless of how tempting the platform makes it feel in the moment.

That survey data lines up closely with what bankruptcy attorneys are already seeing walk through their office doors. Consumer bankruptcy lawyers interviewed by Business Insider described a wave of young clients, mostly men in their twenties and thirties, running up tens of thousands of dollars in credit card debt tied directly to online betting and prediction market trading. North Carolina attorney Ed Boltz described clients arriving with $20,000 to $40,000 in credit card debt accumulated in as little as a single year, while Florida attorney Chad Van Horn described one client in his mid-twenties who built up $25,000 in credit card debt before filing for Chapter 7, and another in his early thirties who filed for Chapter 13 after amassing roughly $50,000 in debt, most of it tied to online betting, before eventually falling behind on rent as well. Van Horn's explanation of the underlying mechanism captures why this happens so quickly, since the debt builds fast precisely because people are not gambling with cash but with borrowed money instead, which makes the path to maxing out a credit line almost a straight one.

Bank of America's own research team flagged this exact risk before it ever became a bankruptcy court statistic. A research note the bank published in November 2025 warned that the gamified interfaces built into prediction markets and sports betting apps drive rising credit balances and increased loss severity, particularly among young men and lower-income borrowers, and the bank's researchers found that access to mobile sports betting correlates with an average 2.75-point drop in personal credit scores. A separate Federal Reserve study cited in the same reporting found that states legalizing mobile sports betting saw credit delinquency rates spike specifically among borrowers under forty, and that the shift from in-person-only betting toward online wagering raises the likelihood of a personal bankruptcy filing by roughly 25 percent.

This pattern is not confined to some small, fringe group of users either. A Northwestern Mutual survey found that 32% of Gen Z respondents and 24% of millennials said they participate in, or are seriously considering participating in, sports betting and prediction markets in 2026, with many citing a general feeling of being financially behind as their main motivation for taking on that risk. Billionaire investor John Arnold, who built his fortune trading energy markets long before any of this existed, has personally funded research into what he calls debt addiction tied to these platforms, telling Bloomberg that the seamless, mobile-first design of these apps, combined with direct links straight into a user's bank account, specifically raises addiction risk among young men who may already be financially vulnerable.

What This Means for Business

A bank underwriting payment rails or deposit accounts for a prediction market platform is not simply accepting regulatory risk tied to the CFTC or a state attorney general, since it is also quietly accepting a customer base where roughly half fund their trading with credit cards, and where the borrowers who do so lose money at an 88% rate. That is a credit risk profile a bank would flag immediately in almost any other lending context it encountered, and Bank of America's own research team has already put that concern in writing well before JPMorgan's decision became public. The bank's choice to exit Polymarket's operating account while retaining other, less visible ties may well reflect exactly this kind of layered risk calculation rather than a single, isolated regulatory concern.

For fintechs building payment infrastructure into this category, the Washington and Nevada rulings make clear that state-by-state compliance is not optional in any meaningful sense, and a platform that misses a geofencing deadline like Kalshi's August 19 target faces immediate, quantifiable financial penalties running to $120,000 per day. Any fintech partner processing payments on behalf of these platforms inherits exposure to that same compliance timeline whether it planned for that exposure or not. For consumer lenders more broadly, the delinquency and credit score data coming out of Bank of America and the Federal Reserve should already be showing up somewhere inside underwriting models built around younger borrowers in states with legal mobile betting and prediction market access, and if it is not showing up there yet, it likely will be soon given how quickly bankruptcy attorneys say these balances tend to accumulate.

What This Means for Society

The credit card debt numbers here describe a genuine behavioral pattern rather than a hypothetical harm dreamed up by critics of the industry. A product that resembles investing closely enough to attract people who describe themselves as financially behind, in Northwestern Mutual's own survey language, but produces an 88% loss rate among those who borrow money to participate, is not really functioning as an investment vehicle at all. It is functioning much closer to the sports betting apps that attorneys like Van Horn and Boltz are already watching drive their bankruptcy clients into Chapter 7 and Chapter 13 filings on a regular basis.

The Washington and Nevada rulings suggest that state courts, unlike federal regulators still working through their own rulemaking process, have already reached a practical judgment on this question for at least the sports and politics categories of these contracts. The Washington judge's own words in the July ruling were direct on this point, finding a likelihood of actual and substantial injury to Washington consumers from illegal gambling activity absent an injunction blocking it. That is a court applying existing consumer protection law and arriving at essentially the same conclusion that Bank of America's own credit researchers reached separately using nothing but loan performance data.

Where This Leaves You

If you run a bank, a fintech, or a consumer lending operation, the decision in front of you now has two separate, independently sourced data trails converging on the same basic conclusion. JPMorgan and two state courts have already decided that specific categories within this industry carry too much regulatory risk to bank without meaningful restriction, while Bank of America's own researchers and a direct consumer survey have already shown that the credit behavior of active users in this category resembles gambling-driven debt accumulation far more than it resembles ordinary investment activity. 

Treating this as a single yes-or-no decision misses both of those threads entirely, while treating it instead as a series of specific, evidence-based exposure decisions, made contract by contract and state by state, is what the banks and courts moving this week have actually chosen to do.

That is the kind of layered judgment call I help executive teams work through directly. If your board is wrestling with a decision like this one, my MD-Konsult advisory practice works with leadership teams on exactly this kind of high stakes, high ambiguity call, and if you are building or investing in a fintech that touches this space, my Entrepreneurship Book walks through how to make hard calls before a regulator or a court ends up making them for you instead.