On-chain RWA crossed roughly $31.7 billion. As of mid-2026, verify current figures, but the number is real and the trackers agree on it. So let me give you the answer up front, before the celebration slides start: almost none of that $31 billion is a real-world asset in the sense the marketing keeps promising you.

Not a building. Not a warehouse. Not a solar plant or a fleet or a production line. When you break the number open, two categories are the overwhelming majority of it, and neither of them is a real business tokenizing something it owns to raise capital.

The category is real. The label is a lie of omission. Let me show you the inside of the box, because once you have seen it, you cannot unsee where the actual opportunity is.

Open the box

Here is what is actually in the $31.7 billion, as of mid-2026, verify current figures.

Tokenized US Treasuries: on the order of $13 billion. Tokenized private credit: the other giant category, and here the data gets genuinely slippery, so I will be honest about it. Trackers quote private credit anywhere from the low teens to around $19 billion, and the spread is not sloppiness, it is a real definitional gap: some numbers count active outstanding loans, others count cumulative loans originated over the life of the platforms, which is a much bigger figure. Either way, between just those two categories you are looking at the overwhelming majority of the number, before you have counted a single tokenized building. The two biggest categories are the category.

So what does that leave for tokenized real estate, tokenized equipment, tokenized commodities, tokenized real-economy business assets, the things every RWA conference deck puts on the cover? A few percent of the total. Low single digits. The pictures on the slides are the smallest part of the pie, and the part nobody photographs, government debt and credit funds, is nearly all of it.

What "RWA" actually means when you read the data

Strip the branding and read the two big buckets for what they are.

The Treasuries bucket is a wrapper around US government debt. In practice it is institutional cash-management money, the tokenized money-market funds like the BlackRock BUIDL-style products, sitting on-chain so that treasury desks and crypto-native funds can hold yield-bearing dollars without leaving the rails they already use. That is a genuinely useful product. It is also, functionally, a money-market fund with a blockchain address. Nobody raised growth capital here. An existing financial product got a new venue.

The private-credit bucket is the same shape one layer over. It is on-chain lending desks and fund structures, the machinery of finance lending to finance, packaged so the units can move and settle on-chain. Again, useful, and again, not an operating company tokenizing its own asset to fund itself. It is a credit fund that chose a blockchain for settlement.

So the honest translation of "RWA passed $31 billion" is this: Wall Street put about $31 billion of existing financial products on-chain. That is what happened. It is a real and significant thing, and it is almost entirely finance-on-chain, not real-business-on-chain.

But gold has huge volume, doesn't that count?

You will hear tokenized gold cited as proof the real stuff is arriving, and the volume number is genuinely large, on the order of $90 billion in spot volume in the first quarter of 2026, verify current figures. That is a big number and it is worth respecting.

It is also the wrong kind of number for this argument. Trading volume measures how fast a thing changes hands, not whether anyone raised capital against an asset they own. Tokenized gold is a trading instrument, a way to hold and move an ounce of gold on-chain. Nobody tokenizing gold is a business bringing in growth capital against a productive asset. It is closer to a stablecoin backed by metal than to a company financing its warehouse. High volume, zero real-economy raise. Worth counting, worth categorizing honestly.

The DTCC tell

If you want the clearest signal of where this category is really pointed, look at the plumbing the biggest players are building. DTCC, the entity that settles a vast share of US securities, has been standing up a tokenization service and a Collateral AppChain, reported with 50-plus firms involved, Chainlink-powered, moving toward production in July 2026 with a Q4 launch on the roadmap. As of mid-2026, verify current figures.

Read what that is. It is institutional settlement and collateral infrastructure for the largest financial firms on earth, so they can move collateral and settle positions faster and around the clock. It is important, it is real, and it is a very strong signal that serious money believes in tokenization as a rail.

Now read what it is not. It does nothing, at all, for a mid-sized European business trying to raise five million euros against its warehouse. The DTCC AppChain is not a channel that owner will ever touch. It is Wall Street building better Wall Street rails. The biggest, best-funded tokenization projects in the world are aimed squarely at making existing institutional finance more efficient, and pointedly not at bringing a real operating business to market. That is the whole category in one data point.

So where is the actual real-world asset?

Here is the part worth keeping, because it flips the whole story from a debunk into a map.

The category the headlines imply, a real business tokenizing a real income-producing asset to raise capital, is the part that is barely happening. The overwhelming majority of the $31 billion is finance-on-chain: Treasuries and credit funds. The real-economy raise, the property, the equipment, the plant, the earning asset owned by an actual operating company, is a few percent of the total, low single digits, a sliver of a sliver.

And an empty part of a market that everyone agrees is the future is not a reason to be cynical. It is the definition of an opening. When the loudest, best-capitalized part of a category is crowded with BlackRock and DTCC and every credit desk in the world, and the part you can actually serve is nearly empty, that gap is not a warning sign. That gap is the trade.

The reason almost nobody is doing the real-asset raise is not that it does not work. It is that it is harder, less liquid, and less scalable than wrapping a Treasury fund, so the big players skipped it and went where the easy billions were. Which means the real-economy owner is competing for capital in the emptiest room in the building.

By the numbers

Breaking the ~$31.7 billion open by category, and scoring each slice not on whether it is a good product, most of them are, but on the only question this roast cares about: is this a real business raising capital against a real asset it owns? As of mid-2026, verify current figures.

CategoryApprox. sizeIs it a real business raising against a real asset?
Tokenized US Treasuries~$13BNo. US government debt wrapped on-chain, mostly institutional cash-management funds (BUIDL-style). A money-market fund with an address.
Tokenized private creditlow teens to ~$19BNo. On-chain lending desks and fund structures. The range depends on active outstanding loans vs. cumulative originations; either way it is the other giant category. Finance lending to finance, settling on a blockchain.
Everything else (real estate, equipment, commodities, real-economy business assets)a few % of the total (low single digits)The small remainder once Treasuries and credit are counted. This is the only bucket where the answer is sometimes yes, and it is tiny. The part the marketing shows, the part the data barely records.
Tokenized gold (trading volume)~$90B Q1 2026 spot volumeNo. A trading instrument, not a capital raise. Volume, not financing, and not part of the outstanding-value total above.
DTCC / Collateral AppChain50+ firms, Chainlink-powered, prod July 2026No. Institutional settlement and collateral plumbing for the biggest firms. Does nothing for a €5M warehouse raise.
Headline "RWA" total~$31.7BThe overwhelming majority is finance-on-chain. The real-economy asset raise is a few percent of the total, the smallest, emptiest, and most open part of the whole category.

Read the last column top to bottom and the reframe writes itself. Every large bucket is a no. The one bucket that can say yes is the one nobody is filling. That is not a failure of the idea. It is a market that has been left almost entirely unserved. Compare it to the mirror-image problem in the dYdX to Arcus roast, where a token was cleanly severable from the business it funded; here the danger is the opposite, a whole category labeled after the asset that is barely inside it.

Four questions to ask the next "RWA" pitch

When someone tells you RWA is exploding and you should get in, run it through these four before you nod.

1. What is actually in the number? If the "RWA growth" they are quoting is the $31 billion top line, remember that Treasuries and private credit are the overwhelming majority of it. Ask what percentage of their number is real estate, equipment, or an operating business. The honest answer is usually a few percent, low single digits, and the honest people will tell you that.

2. Is any of it a real business raising capital? Distinguish, hard, between a financial product moving on-chain and an operating company financing an asset it owns. A tokenized money-market fund is the former. A warehouse owner raising against the warehouse is the latter. The whole confusion of this category lives in that distinction.

3. Who is it really for? The biggest infrastructure, DTCC, the AppChain, the institutional Treasury funds, is built for the largest firms in the world to settle and move collateral. That is not a channel a mid-sized real-economy business can use, and no amount of category enthusiasm changes that. Ask who touches the rail they are describing.

4. What does this mean for a real-asset owner? If you own a genuine income-producing asset, the takeaway is not "RWA is hype." It is that the part of RWA that fits you is nearly empty, which is the best possible condition to enter a market that the whole industry agrees is the future. The crowd went to the easy billions. The room you belong in is open.

The honest landing

I am not here to tell you the $31 billion is fake, because it is not. Tokenized Treasuries are a real and good product. On-chain private credit is real. DTCC building settlement rails is a serious institutional endorsement of the technology. All of that is true and all of it matters.

What is also true is that this is a finance-on-chain story wearing a real-asset name. The category as it exists in the data is Wall Street putting its own existing products on better rails. The real-economy asset tokenization the headlines keep implying, the business raising capital against something it owns and operates, is the part almost nobody is doing at scale. That is not a reason to walk away from it. It is the single clearest reason to walk toward it, because opportunity lives where the crowding is not.

If you own a real earning asset, the $31 billion is not your competition. It is proof the rails work, built and paid for by people solving a different problem than yours, and it leaves your part of the market wide open.

This is analysis and opinion, not investment advice. Figures are as of mid-2026 and should be verified on neutral sources such as RWA.xyz before any decision. Category breakdowns are approximate.

Frequently Asked Questions

What is actually inside the $31 billion of on-chain RWA?
As of mid-2026, verify current figures. On-chain RWA sits around $31.7B by trackers like RWA.xyz. Tokenized US Treasuries are on the order of $13B and tokenized private credit is quoted from the low teens up to around $19B depending on whether you count active loans or cumulative originations. Between just those two you have the overwhelming majority of the number. Real estate, equipment, commodities, and real-economy business assets together are a few percent of the total, low single digits.

Is any of it a real business raising capital against a real asset?
Almost none. The Treasuries bucket is a wrapper around US government debt, mostly institutional cash-management funds. The private-credit bucket is on-chain lending desks and fund structures. Neither is an operating company tokenizing its own warehouse or plant to raise growth capital. The category the headlines imply is the smallest slice of the category the data measures.

What about tokenized gold, DTCC, and the Collateral AppChain?
They reinforce the reading. Tokenized gold shows huge volume, ~$90B spot in Q1 2026, verify current, but that is a trading instrument, not a capital raise. DTCC's tokenization service and Collateral AppChain, 50-plus firms, Chainlink-powered, production in July 2026 with a Q4 launch, are institutional settlement and collateral plumbing for the largest firms and do nothing for a €5M warehouse raise.

If RWA is mostly finance-on-chain, is real-asset tokenization a bad idea?
The opposite. The overwhelming majority of the number is Treasuries and credit funds. The real-economy raise, an operating business tokenizing an income-producing asset, is a few percent of the total, low single digits, and the part almost nobody is doing at scale. An empty part of a market everyone agrees is the future is an opening, not a warning. For a business with a genuine earning asset of roughly €5M or more, the absence of crowding is the reason to move now.

Daniil Kozin structures tokenized real-asset deals in Europe and writes the RWA Roast series to cut through the conference slides. Previous roasts: Polymesh, Securitize, MANTRA, Plume, Justoken, Reental, GromaCoin, Centrifuge, Chainlink, Figure, Stellar, Kelp DAO, Syrup, Ondo, Canton, dYdX. Full archive at daniilkozin.com/blog.

Disclaimer: This is analysis and opinion, not investment advice. Figures are as of mid-2026 and should be verified on neutral sources such as RWA.xyz before any decision. Category sizes are approximate, trackers vary, and the private-credit figure in particular ranges widely depending on whether it counts active outstanding loans or cumulative originations. The "overwhelming majority is finance-on-chain, real-asset raises are a few percent" framing is an editorial reading of the tracker breakdown, not a precise audited figure. Do your own research.

Sources:

Data and figures as of mid-2026. Verify current figures on neutral sources before any decision.