Lessons from Five Years of Real Estate Tokenization: Why Only Two Models May Succeed?

Real estate tokenization has only reached $200 million. This article analyzes the successes and failures of five models, indicating that credit and index products are the way forward.

Written by: @FourPillarsFP

Compiled by: AididiaoJP, Foresight News

Real estate is the largest asset class globally, yet it is one of the smallest asset classes on-chain. According to RWA.xyz data, by July 2026, distributed tokenized real estate is approximately $203 million, while the entire tokenized asset market is about $35 billion.

The gap is significant, with many attempts and numerous failures, yet capital continues to flow in. This week, Index Industries emerged from stealth mode, completing a $50 million seed round to create a single index token for U.S. housing prices.

Will real estate tokenization be the next growth engine for RWA? The author believes it will, but growth will not come from the models built by most in the past. True growth lies in credit and index products. As for fractional ownership of a single building, this is the model most people think of first, yet it has the worst performance record and is the most complex in structure.

1. Doing Financial Business in Real Estate is Actually Easier to Succeed

Under the label of "real estate tokenization," there are actually five types of businesses. They offer different things to holders, leading to varying degrees of success.

Fractional ownership sells shares of a company or trust that holds a property. Holders receive rental income and wait to sell, while the platform relies on listing fees and management fees for each small building to maintain operations. These fees do not cover operational costs, making this the weakest of the five models. RealT, Kasa Korea, and Funble have all shut down.

Housing credit puts mortgages on the ledger, allowing investors to earn interest paid by borrowers. Revenue comes from lending and services. Figure reported a net revenue of $393 million in the first half of 2026, with a net profit of $132 million. This is the most mature model among the five, but it requires scale. Pineapple Financial has moved over $1 billion in mortgage records on-chain, yet reported only $2.1 million in revenue over nine months.

Title and settlement involve the verification of ownership, custody, and contract registration to transfer property from seller to buyer. It does not provide investment products itself. Propy records contracts and transactions, charging a fee for each settlement, and is currently using a $100 million financing limit to acquire title companies. Revenue is stable, but growth relies on acquiring licensed companies, and the author expects expansion to be slow.

Government registration effectively incorporates the government into the fractional ownership model. The land registration department lists token holders as registered owners, giving them the strongest legal claim among the five models. The Dubai platform Prypco Mint charges fees for issuance and trading, currently open only to residents. The business itself is credible, but it can only follow the pace of one regulatory body.

Price exposure completely removes the property itself. Parcl offers leveraged contracts based on urban indices; USHP plans to launch a supported token in Q4 2026 based on the Index Industries housing price index. Revenue will come from the asset scale attracted by the index and the resulting trading volume. Daily indices can also be made into perpetual markets, similar to what Hyperliquid's HIP-3 has done for stock indices.

The pattern is clear: models that earn money from existing cash flows like loan interest and settlement fees have already achieved scale. Models that require investors to bet on a building or a manager have failed the most. Each building has its own tenants, maintenance, and legal issues, making standardization impossible; the platform loses money on one building while making money on another.

2. Why It Hasn't Taken Off

Investors buy real estate typically for two reasons: stable income and price appreciation. Tokenization promises to lower the barriers, similar to how U.S. REITs make expensive, illiquid assets more affordable and sellable.

On both counts, on-chain products have not delivered. The author identifies four issues: yields are unsustainable, platforms are losing money, tokens are rarely traded, and price increases are too small.

Yields may look high, but few can withstand scrutiny.

The annualized income advertised or paid by real estate tokenization products ranges from 3.4% to 16%. Few have been audited, making the source of each number more important than the number itself.

The platforms with the highest advertised yields are often the ones that have already failed. RealT advertised rental income of 6% to 16%, but stopped most payments by the end of 2025, officially halting distributions in February 2026, and announced voluntary liquidation in July.

The remaining high figures are mostly forecasts or self-reported averages. Dubai's Prypco Mint advertises a net annual return of 8% to 12%, which combines rental income and capital appreciation predictions. Lofty claims that as of May 2026, 111 properties had an average rental yield of 9.2%, but did not disclose the calculation method.

Credit products pass on the money paid by borrowers. Figure's home equity pool showed a yield of 9% in February 2026. The PRIME token based on the same pool shows a native yield of 6.01% on DefiLlama.

The lowest figures are often the most reliable. Lucentblock's building in Daejeon paid out about 5% annually before selling in October 2025. Mitsui's AEON Omiya land token is expected to yield 3.4% pre-tax under Japanese securities law.

Regulated, realized, or third-party tracked figures generally fall between 3% and 6%.

Liquidity remains poor: Japan issues the most but trades the least.

Japan is the largest regulated market for tokenized real estate. In FY2025, ¥140.8 billion in real estate security tokens were issued, accounting for about 85% of all security token issuances. BOOSTRY estimates that cumulative public offerings will reach ¥530 billion in FY2026.

The market can grow because it is embedded in the existing financial system. Trust banks hold properties, and brokerages like SBI and Nomura sell tokens to retail clients, with blockchain replacing the back-end record-keeping.

This same design also explains why it has not developed into an on-chain market, with three prominent limitations.

Secondary trading is virtually nonexistent. The Osaka Digital Exchange launched eight tokens in March 2026, valued at ¥33.6 billion, but settlements still take two business days.

The network is closed. Most tokens are issued on private consortium ledgers, blocking foreign capital and on-chain markets. Progmat only recently moved tokens to Avalanche.

Essentially, it is a savings product. A yield of around 3.4% is attractive to Japanese savers but offers limited appeal to global investors. Thin trading indicates that most buyers hold until maturity.

Japan illustrates that regulated distribution can scale tokenized real estate; it also shows that merely issuing tokens does not create a liquid market.

Price increases are insufficient.

Housing prices are not rising fast enough to attract capital on-chain. BIS data shows that over the ten years ending in 2025, the best-performing U.S. markets rose by 84%, averaging about 6% per year. Singapore saw about 4% annually, Japan 3%, South Korea 2%, and mainland China and Hong Kong nearly returned to their starting points.

These returns appear meager compared to the yields already available to on-chain investors. Housing credit can provide yields of 6% to 9% without exposure to housing price fluctuations. Tokenized government bonds have attracted $15.9 billion, while real estate has only reached $203 million.

Homeowners have better returns than indices because they borrowed most of their purchase price and live in the asset. Token holders only benefit from unleveraged price changes. Therefore, price exposure must be leveraged or combined with income to be competitive.

3. Lessons from Past Failures: Move Prices and Debt On-Chain, Not Specific Buildings

Past failures stem from attempts to move the buildings themselves on-chain, resulting in issues with property management, narrow exit strategies, and platform costs. Two models may succeed because they leave the buildings in place and only move prices and debt on-chain.

The first model is to create a venue for leveraged trading and hedging.

There is genuine demand for exposure to housing prices, but building a usable market is challenging, similar to the Pre-IPO market: thin liquidity and no recognized fair price. Parcl failed to seize this opportunity.

Real estate has an advantage: public indices provide reference prices. Trading venues can leverage this to offer leverage to investors who feel that 6% per year is too low, and also provide hedging for owners, developers, and lenders who default on long positions in real estate.

The second model is to create on-chain private credit within a composable ecosystem.

Mortgages themselves generate cash flow, and Figure has recorded about $20 billion in mortgages on its books. Once loans are tokenized, they can be sold, staked, or split. Figure Connect sells loans to external buyers, and its PRIME token, based on its home equity pool, serves as collateral in the Solana lending market.

Both models do not require the platform to manage tenants or maintenance, which is precisely where early platforms failed.

4. Conclusion: Growth Comes from Credit and Indices

Real estate can become the next growth engine for RWA, with growth coming from products built on credit and indices.

Housing credit already has about $20 billion on-chain. RWA.xyz counts it as private credit, not included in the $203 million real estate figure. USHP had commitments of up to $10 million before its launch. Its trading in the first few months will test whether index exposure can find liquidity for single buildings that has never existed before.