Last month a fund administrator I have known for eleven years called me at 11pm London time. She had spent the day reconciling a tokenized treasury product for a mid-sized pension client, and she was exhausted. "The bank says this is the future," she told me. "But nobody on my desk can explain who actually holds the keys if something breaks." Her question is the one Citigroup's new forecast does not answer.
The headline number is $5.5 trillion by 2030. Tokenized assets — bonds, money market funds, real estate, private credit — moving onto distributed ledgers at a scale that would "redefine financial markets." Citigroup, a Tier 1 global systemically important bank, published the projection through its research arm, and it was quickly amplified across crypto media. On its face, it is a vote of confidence. Underneath, it is a question about trust that has been dressed up as a question about size.
Tokenized assets are not a new idea. What is new is the vocabulary. For years the industry called this "security tokens" and struggled. Then it became "RWA" — real world assets — and suddenly every institution had a strategy. The mechanics are deceptively simple: an off-chain asset, a US Treasury bill or a share in a money market fund, is held by a licensed custodian, and a blockchain token is issued that represents a claim on it. The token trades, settles, and sometimes pays yield. The chain is the record; the custodian is the reality.
This matters because the bear market has made everyone honest about narratives. Two years ago, RWA was a talking point. Today it is one of the few sectors with real products live on mainnet — tokenized Treasury funds from Franklin Templeton, BlackRock's BUIDL, and a growing set of private credit vehicles. The growth is real but small. Excluding stablecoins, on-chain RWA sits somewhere in the low hundreds of billions of dollars by the most generous measures, and closer to $100-200 billion by stricter definitions. Citigroup's $5.5 trillion is not a description of what exists. It is a bet on a curve that has not started bending yet.
This is not an abstract concern for me. During the 2020 DeFi Summer I co-founded an educational initiative that taught non-technical users how lending risk parameters actually work, and I watched hundreds of people learn to read a protocol's governance before they trusted it with money. Later, through the FTX collapse in 2022, I ran a weekly newsletter for people trying not to panic-sell. Both experiences taught me the same lesson: adoption does not happen at the moment a product launches. It happens at the moment a person decides the system will not betray them.
I have spent the last decade auditing whitepapers and governance structures, and I have learned to read forecasts the way I read tokenomics: follow the incentives, not the headline. So let me do that here.
The core insight: the $5.5 trillion number is a market-size projection, not a feasibility argument — and the gap between the two is where every RWA failure will live.
Three bottlenecks decide whether the forecast holds.
First, asset title. Moving a bond onto a chain does not move the legal ownership; it moves a representation of it. The real asset stays inside a custodian, governed by the law of some jurisdiction, subject to the terms of a custody agreement that no smart contract can enforce. Every tokenized asset therefore carries a hidden dependency: the chain is trustless only up to the moment the custodian is asked to deliver. In my 2017 audit of ICO whitepapers, I found the same pattern — projects that promised decentralization while the actual control sat with three multisig keys and a lawyer. RWA repeats that structure, but at institutional scale, with better suits and worse excuses. The chain records ownership. The custodian owns it.
Second, composability. The value of DeFi has always been that assets can be stacked — lent, borrowed, used as collateral, rehypothecated — without asking permission. Tokenized assets are built on the opposite principle. They require whitelisting, KYC, and transfer restrictions, typically through standards like ERC-3643 or ERC-1400. This is not a flaw; it is the price of legality. But it means the "liquidity" that Citigroup's forecast implies is a permissioned liquidity — deep within a closed pool, thin across the open market. A tokenized Treasury bill that can only move between approved addresses is not liquid in the way a Treasury bill is liquid. It is liquid in the way a private club is public.
Third, interoperability. Today's RWA landscape is fragmented across chains, standards, and issuers, and none of them talk to each other by default. A tokenized fund on one chain cannot be used as collateral on a protocol on another without a bridge — and bridges are the single most exploited category in the entire industry. Scale requires solving this. Citigroup's forecast assumes it is solved. It is not.
There is a fourth dependency worth naming, because it is the one the forecast hides most effectively: stablecoins. The largest tokenized assets in existence today are dollar stablecoins, and they are the reason some analysts can wave toward trillion-dollar figures without blushing. Strip them out — as any honest accounting of "RWA" should — and the market you are actually forecasting is a fraction of the headline. I am not saying the forecast is wrong. I am saying the denominator is doing a great deal of quiet work, and readers deserve to see it.
Now, let me be fair to the bank, because I think the direction is right even where the number is soft. Citigroup is not a crypto-native shop chasing a narrative. It is a custodian, a settlement bank, and increasingly a tokenization platform operator. When it says the market will be $5.5 trillion, it is describing a world in which its own custody and settlement rails become the backbone of on-chain finance. That is not a prediction. That is a business plan wearing a prediction's clothes.
And the number itself is unremarkable in context. Boston Consulting Group has floated figures up to $16 trillion on a wider definition. BlackRock's Larry Fink has called tokenization "the next generation for markets." Across sell-side research, the 2030 range runs from roughly $4 trillion to $16 trillion — a fourfold spread. That spread is the tell. When the same market is sized at four different numbers, the disagreement is not about growth. It is about definition. Include stablecoins and you inflate the figure dramatically. Exclude them and the "pure RWA" market shrinks by an order of magnitude. Citigroup's $5.5 trillion sits comfortably in the middle, which is precisely why it is easy to publish and hard to falsify.
Here is the angle I have not seen anyone make. Tokenization does not make assets more liquid. It makes them more observable. The liquidity claim is the weakest part of the pitch, and it is the one that will decide whether the forecast is right.
A Treasury bill is already the most liquid instrument on earth. Putting it on a chain does not improve that; it adds a layer. What the chain genuinely improves is transparency — real-time settlement, auditable holdings, programmable compliance. Those are governance gains, not liquidity gains. And governance gains are exactly what institutional investors actually pay for, even when the marketing says otherwise.
The uncomfortable corollary: if tokenization's real value is observability, then the winners are the parties who control the observation — custodians, compliance vendors, and settlement networks. That is the opposite of the "DeFi eats TradFi" story the crypto community tells itself. RWA is not TradFi being disrupted. It is TradFi absorbing the chain and keeping the leverage. The infrastructure layer — custody, cross-chain settlement, identity — captures the value, and it looks a lot more like SWIFT than like Uniswap.
I will go further. Empathy is the ultimate security layer, and the RWA market has not priced it. Every tokenized product ultimately asks an investor to trust an institution they cannot audit and a chain they cannot read. The only thing that bridges that gap is a credible promise that the institution will behave when the market does not. That promise is not in the whitepaper. It is in the custody agreement, the regulator, and the reputation of the bank. Which brings us back to the fund administrator at midnight, asking who holds the keys.
This is why I keep saying: people first, protocol second. Always. The protocol is the plumbing. The person holding the pension is the point.
The $5.5 trillion question is not whether tokenization grows. It will. The question is whether the growth serves the people the system claims to include, or simply gives the old custodians a faster ledger. Track the on-chain numbers, not the forecasts. Track who holds the keys, not who publishes the chart. Trust is earned in bear markets — and a forecast is not trust. It is a promise waiting to be audited. Audit the promise before you trust the number. The plumbing can wait; the people cannot.
