EU RWA market size in 2026: the marketed number vs what is actually investable.

The figure quoted across the industry in mid-2026 is roughly 25 to 35 billion dollars in tokenized real-world assets, excluding stablecoins. The exact number depends on the data source (rwa.xyz, DefiLlama, and various asset-manager reports all count slightly differently) and on what gets included. Stablecoins are usually excluded because they are a money-transmission instrument rather than an investment, though some reports fold them in to reach larger headline figures.

Here is the thing to understand before the number means anything: a market-size figure is only useful if you know what is inside it. "25 to 35 billion dollars in tokenized real-world assets" sounds like a diverse economy of tokenized property, infrastructure, credit, and commodities. It is not. The composition is heavily skewed, and the skew is the story.

The honest framing is that the RWA market is, today, mostly a tokenized money-market fund market with a diverse-real-assets story told on top of it. That is not a criticism of tokenized treasuries, which are a genuine and useful product (covered in detail in the BUIDL vs BENJI vs USDY vs USDM comparison). It is a correction to the impression the headline number creates.

Break the headline figure into its actual components and the picture changes completely. Approximate composition of the tokenized RWA market in mid-2026, by category:

The dominant fact is the first row. The large majority of tokenized RWA value is US Treasury and money-market exposure. This concentration exists for a simple structural reason: US Treasuries have near-infinite supply, deep institutional appetite, a uniform and well-understood risk profile, and require essentially no per-asset diligence (a Treasury bill is a Treasury bill). Tokenizing a billion dollars of Treasury exposure is an operational exercise. Tokenizing a billion dollars of real estate requires sourcing, diligencing, and structuring hundreds of individual properties, each with its own risk.

So the market concentrates where tokenization is easy and the asset is uniform, not where the narrative is most exciting. The "real-world asset" framing implies the diverse real economy. The actual value sits in the single most uniform, most liquid, least diverse real-world asset there is: short-term US government debt.

This is the filter that most RWA market-size discussions skip, and it is the most important one for an allocator. In late 2025, rwa.xyz formalised a distinction that the data had been quietly muddling: distributed assets versus represented assets.

Distributed assets are genuinely tokenized in the sense an allocator expects: transferable between wallets, tradable on a secondary venue, with price discovery and real liquidity. BUIDL, BENJI, USDY, and a properly structured tokenized property SPV are distributed assets.

Represented assets are tokenized only in a recordkeeping sense: the asset's notional value is recorded on chain, but the tokens cannot be moved to wallets outside the issuing platform or transferred between wallets. The blockchain is a database, not a distribution channel. There is no secondary market, no price discovery, no real liquidity.

The practical takeaway: when you read any RWA market-size figure, ask what share is distributed versus represented. A headline that includes large represented-asset notional is quoting a number that overstates the investable market, sometimes by billions. The distributed-asset figure is the one that matters for an allocator, because it is the only part you can actually trade.

Tokenized real estate is the category the RWA narrative leads with. "Own a fraction of a building with 100 dollars" is the headline use case in nearly every introductory article. It is also one of the smallest categories by actual value.

Combined tokenized real estate, retail platforms (RealT at roughly 130-150 million, Lofty at roughly 89 million, Reental at roughly 71 million, GromaCoin at roughly 68 million) plus accredited tokenized property SPVs, sits in the low hundreds of millions to low single-digit billions of dollars globally in 2026. Against a 25 to 35 billion dollar headline RWA figure, real estate is a single-digit-percent slice.

The reason is the same structural fact that explains the treasury concentration, seen from the other side. Real estate does not scale like Treasuries. Every tokenized property requires sourcing, legal structuring, diligence, custody, and ongoing management, and each property is a unique asset with unique risk. The per-deal cost and the diligence load mean tokenized real estate grows building by building, not in billion-dollar institutional blocks. That is not a flaw, it is the nature of the asset, but it means the category will likely always be a small share of total tokenized RWA value even as it grows.

For the allocator, this reframes the opportunity. Tokenized real estate is not where the big money is, and it is not going to be. It is a niche of direct, named-asset exposure for allocators who specifically want it. The three-tier structure of how that niche works (public REITs, retail platforms, accredited SPVs) is in the tokenization vs REITs guide, and the deeper retail comparison is in the RealT vs Lofty vs REIT piece.

Now narrow the headline figure to what a European accredited allocator can genuinely access, and the number shrinks again. Start from 25 to 35 billion dollars and subtract:

• US-person-only tokenized treasuries. BUIDL is qualified-purchaser-only (US institutional). BENJI's primary retail access is US. A European allocator cannot directly buy the largest single chunk of the RWA market.
• Represented assets. Untradable by definition, for anyone.
• Jurisdiction-restricted deals. Products structured for and restricted to other markets.

What remains genuinely accessible to a European accredited allocator:

The accredited EU real-asset SPV segment specifically, the part this desk operates in, is a small fraction of the headline RWA number: measured in low single-digit billions across all of Europe, concentrated in energy infrastructure, light industrial property, and credit. But it is the part that is genuinely investable, genuinely transferable, and backed by real-economy assets that produce real cash flow, rather than by US government debt with a token wrapper.

That is the honest reframing. The investable European real-asset RWA market is far smaller than the headline implies, but it is also far more real: every deal is a specific asset with a specific operator and a specific cash flow, not a notional figure in a market-size report. The 9-category deal-flow map shows what is actually available.

Alongside the current market-size figure, the industry quotes future projections: tokenized RWA reaching anywhere from several hundred billion to tens of trillions of dollars by 2030, depending on which consultancy or asset manager published the forecast.

The single most useful thing to notice about these projections is that they vary by an order of magnitude between sources. When credible institutions publish 2030 estimates that range from, say, 600 billion to 16 trillion dollars, the spread itself is the signal: these are assumption-driven forecasts, not measurements. A projection with a 25-fold range between the low and high credible estimate is not telling you what the market will be. It is telling you that nobody knows.

The honest reading has two parts. The direction is real: tokenized treasuries are genuinely growing, institutional infrastructure (BlackRock, Franklin, Securitize, regulated custody) is genuinely being built, and the regulatory frameworks (MiCA in the EEA, evolving US guidance) are genuinely maturing. Something real is happening. The specific numbers are marketing-grade forecasts that should not anchor any allocation decision. An allocator who sizes a position today based on a 2030 projection has confused a forecast with a fact.

Treat the projections as evidence that serious institutions believe the category has a future, which is useful context. Do not treat them as a measurement of anything, because they are not.

A market-size figure is context, not a decision input. Used correctly, it tells an allocator three useful things and zero misleading ones.

What it legitimately tells you: that institutional infrastructure is being built, that the category is real enough for BlackRock and Franklin Templeton to commit, and that the regulatory direction is toward clarity rather than prohibition. That is genuine directional context worth having.

What it does not tell you: whether a specific deal is good, whether a specific platform is trustworthy, whether a specific yield is sustainable, or whether a specific operator can execute. None of those questions is answered by a macro number. They are answered by deal-level diligence: the asset, the operator, the structure, the waterfall, the exit. The market-size figure and the allocation decision live at completely different levels.

The reason this distinction matters is that the RWA narrative is engineered to move from the macro to the micro without you noticing: the deck opens with the big market number, then pivots to a specific deal, and the size of the market is offered as implicit validation of the deal. It is not. A 30 billion dollar market can contain excellent deals and terrible ones in equal measure. The market size tells you the category is real. It tells you nothing about the deal.