I spent the last decade watching smart contracts fail for reasons that had nothing to do with code. The LibertyDAO multisig drained in 2017 wasn't a technical bug — it was a philosophical one. We built infrastructure without asking who it served or how it would behave under stress. That lesson keeps echoing through every new narrative crypto embraces. And right now, it's echoing through the tokenized treasury space, which just crossed $160 billion in issuance.
That number should be exciting. It's not. Because nearly all of it is what I call "distribution-stage" tokenization — assets minted for the purpose of being held, not used. The next phase of tokenization isn't about issuing more tokens. It's about making them do work. That work is collateral.
The shift is underway. Aave Horizon has surpassed $250 million in TVL. Figure PRIME grew over $200 million this year alone. Midas launched mWIN, a tokenized fund yielding around 6.9%, and it's now sitting in Morpho's lending markets. But here's the uncomfortable truth no one wants to say out loud: we are nowhere near ready for the liquidation scenarios that real collateralization demands.
The Context: From Distribution to Utility
Let me be precise about what changed. For two years, the RWA narrative was about issuance. BlackRock's BUIDL, Franklin Templeton's BENJI, a parade of funds wrapping treasuries into tokens. The pitch was simple: bring institutional-grade assets on-chain. The numbers were impressive. $160 billion in tokenized treasuries. But almost all of that sits in wallets, earning yield, occasionally transferring. It's not participating in DeFi. It's not securing loans. It's not doing anything that requires the blockchain at all, honestly.
Then came the "utility" stage. Projects started asking a different question: can these assets be used as collateral in lending protocols? Aave Horizon was built specifically for institutions to borrow stablecoins against their tokenized assets. Figure PRIME focuses on tokenized credit as collateral. And mWIN, issued by Midas with Wellington Management running the strategy and Northern Trust holding custody, is now listed on Morpho with PYUSD loans backed by it.
This is genuinely the right direction. I've argued for years that tokenization only matters when assets become productive. As collateral, tokenized assets can finally unlock the liquidity trapped in traditional finance. A holder of a $100 million bond portfolio can deposit their tokenized fund into a lending market, borrow stablecoins, and keep their credit exposure and yield — no sale required. That's the dream. The reality is that our liquidation infrastructure was built for a world where everything trades 24/7 and settles in seconds, not for a world where the underlying collateral settles in T+1 or T+2.
The Core: The Liquidation Mismatch Nobody's Solving
Here's the technical problem that keeps me up at night. DeFi liquidates in minutes. Traditional credit settles in days. Tokenization does not bridge this gap.
Think through the mechanics. When you post ETH as collateral on Aave or Morpho, the system monitors the price continuously. If your health factor drops below one, a liquidator steps in, repays your debt, and sells your ETH — likely within seconds or minutes. The market for ETH is deep, liquid, and open around the clock.
Now consider mWIN as collateral. The underlying portfolio holds investment-grade CLOs and other asset-backed credit. NAV is calculated periodically, not continuously. Redemptions happen on a T+1 basis. The bonds themselves only trade during traditional market hours. So when the value of that collateral drops — and it will drop, markets do that — the protocol faces a fundamental problem: it can identify the risk, but it cannot act on it with the speed its own design assumes.
The liquidation path for RWA collateral is not a technical detail. It is the entire ballgame.
Midas and Sentora, the market curator on Morpho, have tried to address this. They've set parameters based on historical NAV, market stress events, liquidity profiles, and redemption mechanics. They're using multiple competing liquidity sources instead of relying on secondary market depth. mWIN offers daily T+1 minting and redemption. All of this is thoughtful. All of it reduces risk. None of it solves the core mismatch.
Let me put this in terms anyone who's been through a bear market will understand. In June 2022, when stETH depegged, protocols that used it as collateral faced cascading liquidations because the market for stETH was shallow relative to the collateralized positions. That was a liquid, actively traded token. Now imagine that stress scenario with a tokenized CLO fund. The NAV is stale. The redemption queue is days long. The liquidator who steps in can't sell the asset — they have to wait for T+1 redemption, assuming the fund even has the liquidity to honor redemptions under stress. This isn't a hypothetical. It's the structural reality of mixing instant settlement with traditional settlement cycles.
Based on my audit experience with DAO treasuries and lending protocols, I can tell you that most risk parameters in DeFi are set based on historical volatility and correlation data. For RWA collateral, that data barely exists. We're parameterizing assets we've never seen behave under stress, using models built for assets that behave completely differently. That's not a criticism of Midas or Sentora specifically — it's an industry-wide gap.
The deeper issue is that we're still designing for the wrong standard. The article that sparked this analysis made a crucial distinction: assets built for distribution and assets built for collateral use should hold different standards. Distribution requires efficient transfer. Collateral requires frequent pricing, fast redemption, executable liquidation, and legal structures that support all of that. These are different design goals, and conflating them creates systemic risk.
mWIN's "native on-chain issuance" approach is a step in the right direction. Instead of taking an existing fund and wrapping it, Midas designed the asset from the ground up for on-chain use. Daily T+1 redemption, multiple liquidity sources, careful parameterization. This is how it should be done. But even with the best design, the fundamental tension remains: the asset class settles in days, and DeFi liquidates in minutes.
The Contrarian Angle: The Real Metric Isn't TVL
Here's where I'll push back on the prevailing narrative. Everyone is celebrating the growth numbers — $2.5 billion in Horizon, $2 billion in Figure PRIME, $160 billion in treasuries. But those numbers might be misleading us about what actually matters.
The article raised a better question: how much tokenized collateral is actually securing loans? How much stablecoin liquidity can be borrowed against it? That's the metric that tells us whether tokenization is creating real economic value or just moving assets from one ledger to another.
Consider the math. If Horizon has $250 million in TVL but only $50 million is actively backing loans, the utility narrative is overstating its progress by 5x. I've seen this pattern before. In DeFi Summer 2020, protocols celebrated TVL numbers that were mostly inflated by yield farming loops — the same capital circulating through multiple protocols, counting as TVL in each one. The real user demand was a fraction of the headline number.
RWA collateral has an even more insidious version of this problem. Institutions can deposit assets and borrow nothing. The assets sit in the lending market, earning yield, waiting for a borrowing opportunity that never comes. It looks like adoption. It's actually just storage with extra steps.
The question we should be asking isn't how much RWA is on-chain. It's how much RWA is working — securing loans, enabling leverage, creating liquidity that didn't exist before.
There's another angle here that makes me uncomfortable. The institutions involved — Wellington, Northern Trust, PayPal — bring credibility and compliance. That's good. But they also introduce a trust model that DeFi was supposed to eliminate. When your collateral's value depends on a custodian's books and a fund manager's NAV calculation, you've re-introduced the exact counterparty risk that on-chain finance was designed to remove. The blockchain is recording the transaction, but the trust is still in Northern Trust's ledger.
This is the paradox of institutional RWA adoption. You need the institutions for the assets, but their presence means the system is only as decentralized as their operational procedures. If Northern Trust has a bad day, your collateral's value is affected, and no smart contract can fix that.
The Takeaway: Utility Is a Verb, Not a Noun
We're at the proof-of-work stage for RWA collateralization. The experiments are running. The parameters are being set. The failures — and there will be failures — will teach us more than the successes.
The next twelve months will tell us whether this narrative has legs. Watch the liquidation events. Watch how protocols handle their first real stress test with RWA collateral. Watch whether the metrics shift from "assets issued" to "assets deployed."
Decentralization is a verb, not a noun. So is utility. Tokenization won't reach its potential because we mint more tokens. It will reach its potential when those tokens start doing real work — securing real loans, enabling real liquidity, creating real economic activity. That's the phase we're entering now. And it's going to be messy.
Code is law, but people are the soul. The code for RWA collateralization is being written right now. Let's make sure we're building it for the stress test, not just the demo day. The real question isn't whether tokenized assets can be collateral. It's whether our infrastructure can handle what happens when they fail.