Enactment of the Guiding and Establishing National Innovation for U.S. Stablecoins Act (GENIUS Act) and the Office of the Comptroller of the Currency’s (OCC) subsequent proposed rulemaking mark a pivotal moment for financial institutions engaging with payment stablecoins, Kevin Oden, principal of Kevin Oden & Associates, a San Francisco-based consulting firm specializing in model risk management and quantitative analysis, shares in a recent white paper. In this Q&A with ProSight Financial Association, Oden explains the law’s implications for model validation teams and the many dimensions of model risk it touches.
ProSight: How are model validation and model risk management implicated in the GENIUS Act framework?
Oden: Broadly speaking, there are eight model risk-related domains that matter under the GENIUS Act. Seven are front and center, and one is more peripheral but still relevant. Every stablecoin issuer will have to either comply with or demonstrate real competency across these areas:
- Reserve valuation: Because a stablecoin is designed to trade at or near parity with a fiat currency—typically the U.S. dollar—its reserves must be valued continuously. In practice, that means 24/7 valuation, which is inherently a modeling exercise and creates significant complexity.
- Liquidity stress testing: The GENIUS Act requires issuers to demonstrate that they can meet redemption demands when people want to exchange stablecoins back into fiat currency or another asset. Stress testing that liquidity—both intraday and over time—is fundamentally a modeling function.
- Interest rate risk: If your reserves are invested in instruments like Treasuries, you need to model how interest rate movements affect value and liquidity.
- Capital adequacy: This is closely tied to reserve valuation but called out explicitly in the Act. Ensuring that capital levels are sufficient and appropriately measured is again a core model risk management issue.
- Concentration risk: Are your reserves concentrated in one asset: U.S. dollars, Treasuries, or something riskier? Measuring and monitoring that concentration involves models.
- Monetization capability assessment: This asks whether you can reliably convert your stablecoin into fiat currency. It overlaps with reserve valuation and liquidity stress testing, but it’s not identical. Monetization is about having the operational mechanisms, technology, and market access needed to exchange stablecoins for dollars—on demand and on a 24/7 basis. Under the GENIUS Act, that expectation of round-the‑clock availability really matters.
- Smart contract validation: This is one of the most challenging areas. The GENIUS Act contemplates extensive use of smart contracts running on blockchains, and validating the algorithms behind those contracts is both a model risk and an operational risk problem.
- Operational risk: It’s somewhat more peripheral from a pure model perspective but still involves modeling components.
ProSight: Since smart contracts are especially challenging for model risk management, let’s dig in. How is this different from things like reserve valuation or capital adequacy? Is it a subset of those issues, or something else entirely?
Oden: It’s related, but it’s different in important ways. Let me give you a concrete example. In 2025, Paxos—who manages PayPal’s stablecoin, PYUSD—had an issue with an automated process embedded in a smart contract. The contract was supposed to transfer funds once balances exceeded a certain threshold. Due to a coding error, the system minted roughly $300 trillion worth of PYUSD, which obviously is far more than global GDP.
They caught the error quickly and burned (industry term for destroying stablecoins) the excess tokens, but it illustrates the risk. This wasn’t a bad economic assumption or a flawed valuation input. It was an implementation failure in a smart contract, which is a model embedded directly into technology and operating autonomously.
That’s why smart contract validation isn’t just about financial logic; it’s about ensuring the code itself is correctly designed, tested, and governed. This goes beyond traditional model validation and into technology-embedded model risk.
ProSight: Does that mean validation also includes making sure there’s a way to undo or reverse mistakes?
Oden: Yes, absolutely, but that’s more on the back end. On the front end, validation is about ensuring the smart contract is implemented correctly in the first place, with appropriate safeguards. In the Paxos example, there weren’t adequate failsafes around certain thresholds. That’s exactly the sort of thing a validation process should catch before deployment.
ProSight: Let’s turn to monetization and liquidity stress testing. From a modeling standpoint, are there parallels teams could draw from in the FX markets? Are there lessons and behaviors that institutions can leverage?
Oden: There are parallels, but also some important differences. Even the FX market, operationally, isn’t truly 24/7 for most institutions, despite being global. Stablecoins push harder on continuous operations.
A useful historical analogy is the creation of the euro. When Europe transitioned from national currencies to a single currency, there was no historical data for the euro. Institutions had to use proxies—mostly the Deutschmark, plus French francs and others.
We’re in a similar place now with stablecoins. There’s very limited history for price behavior, liquidity dynamics, stress scenarios, and interest rate sensitivity. So, while there are strong similarities to FX—especially pegged currencies like the Hong Kong dollar—the lack of historical data introduces real modeling challenges.
ProSight: Could we see a market structure emerge where some stablecoins are highly liquid, while others are more like off-the-run fixed ‑income instruments that are difficult to monetize?
Oden: Eventually, yes. But under the GENIUS Act, issuers aren’t supposed to start there. To be approved as a stablecoin issuer, you need to meet minimum thresholds across those risk domains we discussed. Stability and monetization near parity are core expectations.
Over time, though, you’re absolutely right. You’ll see differences in liquidity, and there will likely be firms stepping in to make markets in certain stablecoins on a regular basis.
This may not happen immediately, but I can easily foresee stablecoin-backed bond issuance and mortgage issuance, too, where you give dollars or stablecoins and get interest paid in stablecoins. By design, these will be smart contracts, which means stablecoin transactions will happen autonomously.
ProSight: Is there a chicken-and‑egg problem here? You need liquidity to issue, but issuance helps create liquidity.
Oden: From a regulatory standpoint, issuers don’t have to prove a secondary market already exists. But from a governance standpoint, as a bank or quasi bank you have to convince your board that demand is real.
Then you still have to satisfy the OCC, which the GENIUS Act designates as the primary federal regulator. That means demonstrating reserve composition, interest rate risk management, liquidity planning, operational readiness—the whole package. If no one wants your stablecoin, of course, all of this falls apart.
ProSight: Finally, what about technology risk more broadly? How big is that from a model perspective?
Oden: It’s enormous. People often emphasize that blockchains are distributed and resilient, but they haven’t truly been pressure‑tested over decades.
The bigger risk, as we discussed, comes from technology-embedded models, especially smart contracts. These contracts execute autonomously. A coding error can instantly mint or burn massive amounts of value before humans can intervene.
We’ve seen analogous events in traditional markets like automated trading systems contributing to sudden market crashes. With stablecoins, contagion could spread even faster. An erroneous minting event could trigger instantaneous buying, selling, or redemption activity across platforms. So, from a model risk standpoint, technology risk, smart contract validation, and rapid contagion effects are critical considerations under the GENIUS Act.
By: Michael Bender
