Tokenization Won't Save the Underclass (Unless)

Can new rails fix a problem that was built into the old ones?

In Part 1, I argued that financial inclusivity in the traditional system is structurally impossible. The credit scoring, the risk math, the KYC requirements, the minimum balances. These are load-bearing walls. You can't remove them without the building collapsing, and so the people they lock out stay locked out.

Tokenization is the most credible challenge to that structure I've seen. And I still think it will probably fail to deliver inclusion for the people who need it most. Not because the technology is wrong. Because three problems sit between the technology and the people it's supposed to reach, and the industry is ignoring all of them.

Where it's working

Tokenized finance is already working for underserved populations in specific conditions.

Two-thirds of the $280 billion stablecoin supply is held as savings by individuals in emerging markets. People in countries with unstable currencies and limited banking infrastructure are using dollar-denominated stablecoins as a store of value, not speculative positions.

Mexico's USDC volume grew 450% and now captures roughly 8% of remittance flows. Sub-Saharan Africa has 40% mobile money account penetration, and stablecoins are building on that existing infrastructure. Up to a third of Latin American households have used stablecoins for retail payments, according to a Mastercard survey.

Standard Chartered projects that stablecoin savings in emerging markets will grow from $173 billion to $1.22 trillion by 2028.

These numbers are real. A farmer in Nigeria holding USDC on a mobile phone has access to a dollar-denominated savings vehicle that no local bank offers at that deposit size. A worker in Mexico sending remittances through stablecoin rails pays a fraction of what Western Union charges.

So why am I skeptical?

Because the places where tokenization is working for inclusion share a specific condition: weak or absent traditional banking infrastructure. Where banks don't reach, stablecoins fill the gap. Where banks do reach, the people who need inclusion most still aren't using crypto.

Problem 1: The on-ramp

A self-custody stablecoin wallet is free to create. Anyone with a phone can generate a keypair and receive USDC. No credit check, no minimum balance, no government ID required.

But the wallet is empty.

Loading that wallet with dollars requires an exchange or on-ramp service. Every exchange requires KYC. KYC requires government-issued ID, proof of address, often a bank account for fiat deposit. Brookings documented this in 2022 and it remains true: "Many crypto platforms typically require a bank account to use cryptocurrencies, so this defeats the purpose of serving the unbanked."

The person who can't open a bank account because they lack ID can't open a Coinbase account either. The person who has no proof of address can't pass the exchange's identity verification. The same documents gate both systems.

In emerging markets, peer-to-peer trading and mobile money on-ramps partially bypass this. A person can buy USDC from a neighbor or load a wallet through M-Pesa. But these informal channels work because the regulatory environment permits them. The GENIUS Act requires stablecoin issuers to have freeze-and-seize capability and comply with AML requirements. As regulation tightens, the informal channels narrow.

Deutsche Bank Research states it directly: "Stringent KYC/AML on-ramp requirements can create barriers for unbanked populations, undermining financial inclusion goals."

The on-ramp problem is the most fundamental. If you can't get into the tokenized system without the same credentials that exclude you from the traditional one, the new rails inherit the old gates.

Problem 2: The walled garden

Suppose you solve the on-ramp. You have a funded wallet. What can you access?

If the tokenized financial system looks like what DTCC is building, not much. DTCC's tokenized securities infrastructure is accessible only through regulated intermediaries. You need a brokerage relationship to touch tokenized stocks. The technology is a blockchain. The access model is the same as before.

Bank consortium stablecoins repeat the pattern. Citi and Goldman's consortium stablecoin works between member banks. If your bank isn't in the consortium, you're excluded. If you don't have a bank at all, the product doesn't exist for you.

This is the permissioned versus permissionless split, and it determines who gets access. A permissioned tokenized system serves the institutions that build it. A permissionless system serves anyone with a wallet. The industry is overwhelmingly building permissioned.

Fractional ownership of tokenized real estate, tokenized securities, tokenized commodities. Deloitte, EY, and Chainlink all document the mechanics. Lower minimum investments, broader access, 24/7 trading.

But I couldn't find a single study showing low-income populations actually using fractional tokenized assets at meaningful scale. The platforms serve institutional and accredited investors. The inclusivity is theoretical.

The walled garden problem compounds the on-ramp problem. Even if you get in, the products worth accessing are behind another gate.

Problem 3: The wrong people

A 2026 peer-reviewed study published in ScienceDirect examined who actually uses crypto. The finding: "Crypto adoption is disproportionately concentrated among underbanked, higher-income, more educated, younger, and male households, and is driven largely by investment motives."

Below $30,000 in household income, there is no statistically significant adoption difference.

The people who need financial inclusion most are not using the technology that's supposed to provide it. They're not using it because the on-ramps exclude them, because the products are designed for investors, and because the financial literacy required to navigate self-custody and DeFi protocols doesn't reach the populations that lack access to basic banking.

The BIS is blunt about this. Its 2026 Annual Economic Report notes that stablecoins' "main use cases so far have been for crypto trading and, to a lesser extent, as offshore stores of value." Adjusted stablecoin transaction values, stripped of wash trading, are less than 1% of total value.

The emerging market savings use case is genuine. But in the countries where TradFi infrastructure exists and exclusion is a policy choice, the people being excluded aren't showing up on-chain.

The conditional

So where does that leave the thesis?

Tokenization is a necessary condition for financial inclusion. Permissionless rails, self-custody, programmable savings, stablecoin yields at 5-7% instead of payday lender rates at 400%, these are genuinely new capabilities. No traditional financial system can offer them at the same cost structure.

But a necessary condition isn't a sufficient one. Three things have to hold for tokenization to actually reach the underclass.

Permissionless on-ramps have to survive regulation. The mobile money and peer-to-peer channels that work in emerging markets today could be regulated out of existence tomorrow. If every stablecoin interaction requires KYC from a licensed intermediary, the unbanked stay unbanked on new rails.

Permissionless products have to survive institutional capture. Every tokenized asset class being built right now, from deposits to securities to stablecoins, is being built with permissioned access. If the only tokenized products that survive are the ones behind regulated intermediaries, the walled garden wins.

And the products themselves have to be designed for people earning $25,000, not people investing $250,000. A fractional share of tokenized real estate priced at $50 is technically inclusive. A DeFi yield protocol that requires understanding impermanent loss, gas optimization, and bridge risk is functionally exclusive.

Where I land

I want to believe tokenization changes the equation. The emerging market data says it can. The structural problems say it probably won't, at least not for the people who need it most in countries where TradFi already dominates.

The rails can be inclusive. The question is whether the products built on them will be, or whether they'll become the same system with a faster settlement layer underneath. We've seen that playbook before. The banks' answer to stablecoins was to build their own, on their own terms, for their own customers. If that's the future of tokenized finance, the underclass stays exactly where it is.

The conditional matters. Permissionless access has to survive. If it does, there's a path. If it doesn't, tokenization is just faster plumbing for the people who already have water.