The Mistake 90% of RWA Projects Make

The most common failure in real-world-asset tokenization is one of sequence: building the token before the asset. A project that raises capital against assets it has not yet acquired is running a fundraise, not tokenizing anything, and in most jurisdictions it meets the definition of a security while insisting it does not. The projects that survive invert the order, starting from a documented, ring-fenced asset and treating the token as the final layer, not the first.

Most real-world-asset projects will not survive. Not because the concept is flawed, the market too small, or regulators hostile, but because they made a fundamental error in the order of operations: they built the token before they had the asset, raised the capital before they had the structure, and promised the liquidity before they had the market. This matters beyond the individual projects, because every failure makes the next honest project harder to fund. Every investor burned by a token backed by nothing becomes harder to convince that a token backed by a documented, ring-fenced, income-producing property is a different instrument entirely. What follows is a diagnostic: what the mistake looks like, why it happens, and how to tell substance from narrative.

What does the tokenized-asset market actually look like today?

The headline growth is real. The tracked tokenized real-world-asset market, excluding stablecoins, grew from around 6 billion dollars at the start of 2025 to more than 30 billion by mid-2026, per RWA.xyz. That number is usually presented as evidence that tokenization is arriving. The composition tells a more sober story.

Close to half of that value is tokenized US Treasuries, around 15 billion dollars, and most of the remainder sits in other already-liquid, already-institutional instruments: money-market funds and commodity tokens such as gold. These are the least illiquid, most transparent assets in global finance, and they were already accessible to institutional investors before tokenization. What tokenization adds to them is operational: faster settlement, continuous availability, programmable compliance. Those are genuine improvements, not transformations.

Real estate, the asset class the tokenization narrative has most loudly promised to open up, is a rounding error. Tokenized real estate stood at roughly 70 million dollars on-chain in mid-2026, about 0.2% of the market, per RWA.xyz. The contrast with the projections is the whole point: Deloitte estimates tokenized real estate could reach 4 trillion dollars by 2035, from under 300 billion in 2024. The genuinely illiquid asset classes, the ones where tokenization would create the most value, are precisely the ones where the market has made the least progress.

What is the core mistake, and why is it a securities problem?

The most common failure pattern is also the most avoidable. A team identifies tokenization as an opportunity, designs a token, and writes a white paper describing the assets it intends to acquire. It runs a raise, selling tokens to investors who are told the assets will be bought with the proceeds. The token price reflects a narrative about future assets and future returns, not a documented income stream from any actual property.

This is not tokenization. It is a fundraise with extra steps. The token is not backed by a real asset because the asset does not yet exist. The investor is betting on whether the team will deploy the capital well, negotiate reasonable acquisitions, structure them properly, and deliver. That is venture-capital risk dressed in real-estate language.

The regulatory reading is not subtle. In most jurisdictions, an instrument whose value depends on the efforts of a third party, and whose holders expect returns from those efforts, meets the definition of a security. A project that raises capital this way without the appropriate registration is not in a grey area; it is in violation, whether or not it has been caught. The correct order is the inverse: find the asset, document it, value it independently, structure the legal vehicle around it, and only then tokenize a verified, ring-fenced interest. The token then represents something real from the moment it is issued.

What other patterns define the projects that fail?

Beyond the sequencing error, four patterns recur.

The first is treating MiCA as a compliance shortcut. The EU's Markets in Crypto-Assets regulation covers crypto-asset service providers and certain token types; it does not convert a tokenized investment instrument into something other than a security. A token that pays income and offers capital appreciation is almost certainly a security under MiFID II, whatever it is labelled. Structuring it as a utility token and claiming MiCA coverage does not change the economic function, and regulators assess function, not labels.

The second is the absence of ring-fencing. If assets are owned by the platform company directly rather than by dedicated special-purpose vehicles, investor exposure to the assets is indistinguishable from exposure to the platform. When an early-stage company hits financial trouble, as they regularly do, the assets investors believed they owned sit on a balance sheet that creditors can reach. A per-asset SPV, where the platform manages but does not own, is not expensive to implement. It is simply less convenient for a team optimising for speed.

The third is overselling liquidity. A fractional interest can technically be transferred without a full property sale, but a willing buyer at a fair price is not automatic. Depth requires participants, and participants require time. A project that launches promising liquid secondary markets, with no realistic path to creating them, sets an expectation it cannot meet, and the investors who discover they cannot exit become the loudest voices against the whole asset class.

The fourth is a utility token whose value is linked to asset performance. A token presented as offering platform access, fee discounts, and governance can be defensible, until its value is structurally tied to the performance of the underlying assets. At that point it has investment characteristics regardless of its label, and the economic-substance analysis a regulator applies reaches the same conclusion the Howey test always has: if holders expect profit from the efforts of others, it is a security. Sound design keeps the access token strictly separate from asset economics.

Why does the mistake keep happening?

The persistence comes from incentive misalignment. Building a properly structured, compliant, asset-backed platform is slow, expensive, and operationally demanding: legal counsel across jurisdictions, independent valuations, SPV formation, smart-contract audits, and regulatory engagement, all before a single token is issued. Building a token, writing a white paper, and raising capital is fast, cheap, and mostly a marketing exercise. The team that takes the fast route reaches investors first and builds market presence first, while the careful team is still filing paperwork.

In the short term, the incentive to move fast is overwhelming. In the medium term, the fast projects fail and the careful ones survive. The difficulty is that the medium term takes long enough that many investors are harmed before the correction arrives.

How do you tell a sound project from a narrative?

Six questions separate substance from story, and each demands a specific answer in writing. Does the platform already hold documented, income-producing assets, not assets it intends to acquire? Are those assets held in ring-fenced SPVs that are legally independent of the platform, with structure documents an investor can see? Has the platform obtained a formal legal opinion on whether its tokens are securities, rather than a general claim of MiCA compliance? What is the secondary-market mechanism today, at what depth and volume, not what it will become? If the platform ceased to exist tomorrow, what specifically happens to the investor's rights in the assets? Has the smart contract been audited by a named third-party firm, and what did the audit find?

A project that cannot answer all six clearly and in writing is not ready for investor capital. That is not a high bar. It is the minimum that shows the legal and operational foundations are genuinely in place.

What do the projects that survive have in common?

The platforms that will still exist in five years share a small set of traits already visible today. They started with assets, not tokens: the team had access to real deal flow before it built the tokenization layer, and the technology served the asset. They engaged regulators early, not to lobby but to build to the requirements from the start, treating compliance as the moat rather than the cost. They were honest about liquidity, telling early investors to position for yield while secondary markets develop. They built the SPV structure as the starting point, not an afterthought, so investor rights survive the platform. And they moved on verifiable proof points: each tokenized asset a documented case, each distribution an on-chain record.

The bottom line

Most real-world-asset projects will not survive, and that is a statement about how markets work, not a prediction of doom. The projects that confused narrative with substance, token with asset, and compliance with inconvenience will be filtered out as regulatory frameworks sharpen and investors learn to ask better questions. What remains will be smaller in number and larger in credibility: the platforms that treated the asset as the point, the token as the administrative layer, and the legal structure as the foundation rather than the footnote. The market will be better for the correction, because the survivors will carry the track records, the regulatory relationships, and the trust that serious capital requires.

This article is part of DeReal's research series on real-estate access, liquidity, and tokenization. It is analysis of the market, not investment advice or an offer of any kind.