Wharton Warns RWA Tokenization May Trigger Liquidity Risks With 24/7 Trading

A new report from the University of Pennsylvania’s Wharton School argues that the fast-moving promise of real-world asset tokenization could become a source of instability if ‘24/7 tokens’ are built on top of assets that can’t be priced, sold, or redeemed at anything close to that speed.

Published in May 2026 by Wharton’s Financial Policy and Regulation Initiative (WIFPR), the paper—Tokenizing Real-World Assets—frames the sector as having moved beyond pilot programs, while warning that the next phase of growth will be defined less by smart-contract engineering and more by ‘liquidity design’ and regulation that matches what a token does economically rather than how it is built.

The authors—Lin William Cong of Nanyang Technological University, Simon Mayer of Carnegie Mellon University, and Daniel Rabetti of the National University of Singapore (and a visiting scholar at Harvard Business School)—survey activity across major incumbents and crypto-native firms, citing initiatives linked to asset managers such as BlackRock and Franklin Templeton as well as fintech and DeFi issuers including Figure Technologies and Ondo Finance.

Their central claim is straightforward: tokenization does not change an asset’s fundamentals. A tokenized loan pool carries the same default risk as the underlying loans. A tokenized deposit is only as safe as the issuing bank. What tokenization changes is market access—trading venue, investor reach, settlement speed, and ‘programmability’—not the credit quality, cash flows, or intrinsic valuation anchor of the asset itself.

For that reason, the report cautions against treating ‘RWA’ as a single category. Instead, it splits tokenized assets into three broad buckets, each with distinct risk and regulatory implications. The first is already-liquid instruments such as U.S. Treasuries, gold, and large-cap equities, where tokenization is primarily about integration into onchain markets—enabling hedging, collateralization, and portfolio rebalancing without leaving blockchain rails—rather than manufacturing liquidity that doesn’t already exist.

The second bucket is money-like claims such as stablecoins and tokenized deposits, designed to hold a fixed value and function as settlement tools. Here, the defining risk is a breakdown of ‘par redemption’ during periods of stress—effectively a digital version of a run dynamic.

The third—and in Wharton’s view, most fraught—category involves illiquid assets such as private credit, loans, real estate, private funds, and other instruments whose valuation and exit timelines are measured in weeks or months. Tokenization in this domain functions as a digital form of securitization: it can create tradable claims on assets that were previously difficult to transfer. But the report stresses a critical distinction between adding ‘tradability’ and achieving true ‘liquidity.’ The former is a legal and technical wrapper; the latter is a market property that may not appear when it is most needed.

On market size, WIFPR compiles public-blockchain RWA data (excluding stablecoins) at roughly $29 billion as of April 2026, rising to around $46 billion when Figure Technologies’ tokenized home-equity and mortgage-related products are included. The paper describes the growth path as rapid: from about $2.9 billion in 2022, to $12 billion by the end of 2024, to approximately $46 billion by 2026. Industry forecasts cited in the report project the category could reach $18 trillion by 2033—an outlook that helps explain the urgency of the regulatory debate.

By segment, the report estimates tokenized loans and private credit at about $20 billion, U.S. Treasury and money market fund (MMF) exposure at $14.5 billion, commodities (primarily gold) near $6 billion, and tokenized equities around $2 billion. It also notes that Treasury and MMF products have overtaken commodities to become the largest non-stablecoin RWA segment on public chains.

The paper’s defining framework is what it calls a ‘speed-matching’ principle. The trading velocity of a token, it argues, should not outrun the speed at which the underlying asset can be priced, adjusted, sold, or redeemed. For deep and liquid markets—Treasuries, spot gold, and actively traded large-cap stocks—around-the-clock token trading is less problematic because arbitrage and reference markets can pull prices back toward fair value, even when traditional venues are closed.

The report points to empirical findings where tokenized gold tracked spot gold with a correlation near 0.99, and tokenized equities sometimes incorporated weekend or overnight information ahead of Monday’s open, effectively previewing the direction of the next traditional-session move.

The danger arises when the pairing flips: ‘fast tokens’ issued against ‘slow assets.’ If token holders can trade continuously while the underlying private credit or property portfolio cannot be liquidated or reliably marked in real-time, the system can embed a run risk resembling money market fund stress events. In a rush for exits, token prices can decouple from stale net asset estimates, and redemption promises can become discretionary or gated—precisely when investors expect them to be most enforceable.

WIFPR visualizes this in a simple 2×2: token trading speed on one axis and underlying asset adjustment speed on the other. Fast-token/fast-asset and slow-token/slow-asset combinations are considered broadly aligned. The red zone is fast-token/slow-asset—where ‘liquidity promises’ exceed what the underlying balance sheet can deliver.

Beyond maturity mismatch, the authors highlight concentration risk that complicates claims of decentralization. By issuer, the top 10 entities account for about 82% of tracked RWA value, and one Figure Technologies mortgage-related token alone represents roughly 37% of the total in the dataset cited. By chain, Ethereum (ETH) hosts about 54% of public-blockchain RWA value, followed by BNB Chain (13%), Solana (SOL) (7%), Stellar (6%), and Liquid Network (5%).

The report argues that operational and governance risk also concentrates in a small set of custodians, issuers, oracle providers, and infrastructure platforms—creating what it describes as buildup of ‘clearinghouse-like’ risk without the same supervisory perimeter found in traditional finance.

Some of the most significant vulnerabilities, WIFPR contends, are “new in appearance but familiar in substance.” One is the hybrid structure many RWA products attempt to offer: the ease of secondary trading associated with ETFs, combined with the ‘par redemption’ expectations typical of funds or bank-like claims. In practice, the report says, token markets may inherit neither the ETF ecosystem’s regulated authorized participants nor robust, rule-bound redemption mechanics. When redemptions are subject to minimum sizes, time restrictions, or issuer discretion, the stabilizing mechanism that pulls market price toward net asset value can fail during stress.

A second risk is the oracle as a single point of failure. Because real-world assets live offchain, pricing and status must be bridged onchain via ‘oracles’ and related data pipelines. The report cites the October 2022 Mango Markets exploit as a case study in how thin liquidity and manipulated price feeds can interact with automated liquidation logic to produce large losses—about $117 million in that event. It also notes that cumulative DeFi security losses have already exceeded $6.9 billion, underscoring how technical weak points can translate into systemic confidence shocks when RWAs scale.

A third theme is securitization’s moral hazard. Packaging loans into tokens can recreate the ‘originate-to-distribute’ incentive problem: if originators can quickly offload risk, underwriting discipline may weaken. The report explicitly draws parallels to dynamics that amplified credit deterioration ahead of the 2008 crisis, arguing that token rails do not automatically solve misaligned incentives.

On regulation, the authors advocate a consistent approach: supervise tokenized products by ‘economic function,’ not by the presence of blockchain. Money-like instruments, liquid risk assets, and illiquid claims should not be governed under a single template. As a benchmark for function-based rules, the report points to the U.S. GENIUS Act (enacted in 2025), citing its requirements around high-quality liquid reserves, monthly reserve verification, par redemption rights, and issuer licensing.

At the same time, WIFPR emphasizes that stablecoin-style rules cannot be mechanically applied to all RWAs. A ban on interest payments, for instance, collides with the economic rationale of tokenized Treasury and MMF products, which derive value from yield. Likewise, frameworks built around licensed, regulated U.S. issuers may not map cleanly onto structures like private credit or real-estate tokenization where custody, segregation, and liquidation mechanics differ materially. For these, the report suggests regulatory analogies closer to closed-end funds, interval funds, or auction-based mechanisms—structures designed to manage illiquidity rather than pretend it away.

Internationally, the paper compares sandbox approaches in the U.K., the EU, Singapore, and Switzerland, noting that Switzerland has progressed furthest in recognizing tokenized securities as a distinct legal category—an advantage it links to Swiss issuers’ visibility in tokenized bond issuance. Still, WIFPR argues that across jurisdictions, redemption design is too often left to issuer discretion, meaning the core ‘speed-matching’ problem remains insufficiently addressed.

One of the report’s sharpest takeaways is its relevance for markets—such as South Korea—where the most popular tokenization targets tend to be inherently ‘slow’ assets. Fractionalized real estate, music royalties, livestock-backed products, fine art, and even niche collectibles can be difficult to value frequently, information can be opaque, and liquidation can be costly or time-consuming. Wrapping these assets in tokens that trade around the clock can create the very time-lag mismatch the report warns about: a product marketed as “sell anytime” that cannot produce cash when many holders head for the exit at once.

As South Korea debates security token offerings (STOs) and secondary-market frameworks for fractionalized products, WIFPR’s message is that design priorities should not center on how finely an asset can be sliced into tokens. The crucial question is whether token trading and redemption rights are aligned with the underlying asset’s valuation and liquidity cycle. The report lists practical tools: notice periods for redemptions, ‘gates’ that allow temporary limits during stress, closed-end or interval structures, auction-based exits, and clear disclosure of valuation frequency, redemption pricing methodology, and loss-allocation rules—aimed at preventing early redeemers from benefiting at the expense of remaining holders using stale marks.

WIFPR also argues regulators should sequence experimentation. Start with liquid, money-like or highly liquid instruments such as Treasuries, gold, and MMFs—using them as a controlled ‘laboratory’ to harden custody, redemption, oracle design, and settlement workflows—before expanding into complex illiquid assets like real estate and private credit. Jumping directly into the hardest category, the authors warn, is the fastest path to a tokenization-driven liquidity shock.

Still, the report does not frame tokenization as purely a risk story. It points to evidence that tokenized equity trading can meaningfully broaden access for small investors: more than three-quarters of tokenized stock trades in one dataset were under $100, suggesting global retail participants—often facing barriers to traditional brokerage access—are using tokens as an entry point to international markets.

The broader implication, WIFPR concludes, is that tokenization can improve market access without rewriting financial reality. When ‘fast tokens’ are used to simulate liquidity for ‘slow assets,’ familiar fragilities—runs, broken redemptions, and incentive problems—can return, only at blockchain speed. The next stage of RWA growth, the authors argue, will depend on whether rules and product designs can match the tempo of tokens to the time it takes for real assets to move.