In "People Are People," I ended with something close to hope. The kind that has been pressure-tested, that has read enough history to be suspicious of itself, but hope nonetheless. The idea that better infrastructure, properly designed, might produce better outcomes for more people.
I want to complicate that.
I should say where I'm writing from. I work at an RWA tokenisation platform. I believe in what we're building — the legal structure, the mechanics, the yield being real. This is not a hit piece on the space or on any team within it. What follows is a diagnosis. But a diagnosis needs to be honest about what it finds.
And what I've been finding, watching this market grow from $5.5 billion at the start of 2025 to over $19 billion today, is a pattern worth naming.
The thesis is clean. Real-world assets — property, credit, treasuries, private equity — are illiquid, expensive, and structurally unavailable to most of the world's capital. Tokenise them. Put them onchain. Fractionalisation lowers minimums. Composability opens new financial primitives. The person in Kuala Lumpur gets access to the same asset class as the family office in Geneva.
It's a good thesis. I still believe it.
The numbers, taken at face value, seem to bear it out. The market has tripled in fifteen months. Hundreds of institutions are building. The language of democratisation is everywhere. But look at who's actually showing up. The newest wallets created specifically to hold tokenised assets — purpose-built, first-time onchain addresses — are predominantly holding institutional-grade instruments. Private credit. Specialty finance. Structured products. Retail categories trail. The data tells you something about who tokenisation is currently serving, even when the marketing doesn't.
The gap between the story and the structure is where this essay lives.
Most protocols handling regulated assets require compliance verification to mint or redeem. KYC. Accredited investor status. Jurisdictional eligibility. The reasons are legitimate: securities law, AML requirements, the basic reality that putting a regulated asset onchain does not make it unregulated.
But the effect is specific. The primary market — where the asset is created and redeemed, where NAV is anchored, where price discovery actually happens — is closed to most people.
The accredited investor threshold doing most of this work is not a new idea. The SEC established it in 1982. $200,000 in annual income, or $1 million in net worth excluding your primary residence. Four decades later, largely unchanged. No inflation adjustment. No fundamental rethinking. That threshold already excludes over 90 percent of American households from private offerings — a ratio that maps, proportionally, onto much of the world.
Tokenisation inherited this architecture because it had no choice. The assets being tokenised exist within legal frameworks that blockchain did not rewrite. The compliance layer is old infrastructure. What's new is only the rails it sits on.
The token is the same for everyone. Worth saying clearly — retail and institutional holders can look at the same contract address, the same onchain record, the same underlying representation of the asset. The difference is not in what the token is. It's in what your wallet is permitted to do with it.
Whitelisted wallets — those that have passed the compliance gate — can interact with the primary market. Mint. Redeem. Arbitrage when secondary prices drift from NAV. Their position is always anchorable to the underlying.
Non-whitelisted wallets can buy and sell on secondary markets, and compose with the token in DeFi — use it as collateral, integrate it into yield strategies, move it permissionlessly. That composability is real value. In some dimensions it exceeds what traditional finance offers retail at all.
But without whitelist access, you cannot mint or redeem. You cannot close the gap when secondary prices dislocate from NAV. The secondary market frequently trades below the underlying's actual value — and when it doesn't, it's usually because sentiment is running ahead of it. Either way, pricing is shaped by whoever has the liquidity to move it, not by the asset itself.
Same token. Different permissions. The gate isn't written into the contract. It's written into the whitelist.
I should be honest about something else. Most people who enter crypto retail don't arrive as believers. I didn't.
The first instinct, for most, is extraction. Find the yield. Rotate the liquidity. Take the airdrop and move on. The early behaviour of retail crypto — mercenary by design, finite by temperament — gave builders and compliance teams every reason to be reluctant. They've watched liquidity farm and exit. They've watched token launches gutted within weeks by the same community that evangelised them days before. The wariness is not paranoia. It's pattern recognition.
And here is the uncomfortable truth the access argument has to sit with: the compliance gate exists, in part, as a structural response to that behaviour. Mercenary retail didn't just make tokenised asset projects nervous. It handed the people maintaining those gates a defensible rationale for keeping them closed.
But here is what the gate cannot read. People change.
The person who arrived mercenary — chasing yield, rotating positions — sometimes stays. Something shifts. They read deeper. They sit with the infrastructure longer. They begin to understand what the underlying asset actually is, what the protocol actually does, what the long-term thesis requires. They move, quietly, from finite to infinite. Still building their wealth. Still below the threshold. Still locked out.
The gate never updated its read. It sorted them the same as the day they arrived — same compliance check, same exclusion, same net.
That's the problem the access argument hasn't been honest enough about. The compliance filter is static. People aren't. And the converts — the ones who did the work, who moved from extraction to conviction — are precisely the ones most likely to be caught in a net designed for someone they no longer are. The gate was built for the mercenary. It keeps out the believer too.
This is structural, not incidental.
Traditional finance built the compliance architecture governing private market access over decades. The stated intent was investor protection — shielding people from instruments too complex to evaluate, from risk they couldn't absorb. That intent was not entirely dishonest. But over time the architecture became something else: a moat. A mechanism concentrating sophisticated returns within a closed loop of sophisticated capital. Wealth as proxy for sophistication. Sophistication as the price of entry.
Tokenisation doesn't dissolve that moat. It builds a bridge to it and installs the same gatehouse at the entrance.
The tragedy is not that the gate exists. The tragedy is its resolution. Wealth is a blunt instrument for sorting human intent. It cannot distinguish between the mercenary and the convert. It cannot read trajectory. The filter was built to separate the sophisticated from the unsophisticated — but it ends up separating the wealthy from everyone else, and calling that the same thing.
Every project that tokenises a regulated asset, markets it as democratised financial infrastructure, and restricts primary access to accredited investors is navigating this contradiction. Some with full awareness of it. Some without asking the question. The charitable reading: this is transitional. Regulatory frameworks will evolve. Singapore and Switzerland are already moving. The architecture is shifting.
The less charitable reading: the capital that needed efficient rails, faster settlement, and composable yield arrived early and got what it came for. The broader access story is a longer horizon, a harder problem, and a less profitable one to solve.
The mercenary culture of crypto retail and the institutional capture of tokenisation are not opposites. They are two sides of the same finite game. And the people who stepped out of it — who became something more patient, more committed, more infinite — are still waiting at the same gate.
The infrastructure is sound. The compliance layer is the gate. The question was never whether tokenisation works — it's whether we're willing to redesign who it works for.
I haven't resolved it. I'm not sure you should either.
IN.
Previously: "Ringgit, Riba, and Remittances"