3.3 Liquidity for Historically Illiquid Markets
What illiquidity actually costs an investor, what the liquidity claim for tokenization asserts, why trading data shows liquidity being inherited rather than created, and what would have to change for tokenized private markets to trade.
Tokenization is said to make illiquid assets liquid. It is the most economically significant claim in the category, because if true it reprices trillions of dollars of private assets by removing a discount they have always carried. It is also the claim with the clearest evidence against it, and the evidence comes from the on-chain market itself rather than from sceptics.
Section 3.2 found that the access claim survives in a narrower form than usually stated, because the binding constraint is legal rather than technical. The liquidity claim fails differently. Nothing legal prevents a tokenized asset from trading. The constraint is that liquidity is a property of markets, not of instruments, and a token inherits whatever market its underlying asset already had.
What Illiquidity Actually Costs
The cost is measurable, and it is large enough to explain why the claim attracts so much attention.
Nadauld, Sensoy, Vorkink and Weisbach studied the private equity secondary market using proprietary data from a leading intermediary, and found that most transactions occur at a discount to net asset value 1. Their central result quantifies what the seller gives up. Buyers average an annualised Public Market Equivalent of 1.023 against 0.974 for sellers, implying buyers outperform sellers by a market-adjusted five percentage points annually 1. For the most common transaction, stakes in funds four to nine years old, the gap narrows to about three percentage points 1.
Five percentage points a year is the price of needing to get out early. The authors attribute both the NAV discount and the buyer-seller return gap to factors associated with asymmetric information and market depth 1.
Note those two causes, because they determine whether tokenization can help. Asymmetric information means the buyer cannot easily verify what the asset is worth. Market depth means there are too few buyers to absorb a sale without moving the price. Neither is a problem of divisibility or settlement speed.
The Liquidity Claim
The tokenization argument runs that an asset which can be split into small units, transferred without intermediaries, and settled in minutes will trade more readily than one requiring a negotiated transfer, a subscription agreement, and a multi-week close. Remove the friction and trading follows.
The friction reduction is genuine. What the argument assumes without stating is that friction was the binding constraint. If the reason a private credit stake does not trade is that transferring it is cumbersome, tokenization solves it. If the reason is that nobody can price it and few want to own it, tokenization changes nothing that matters.
Liquidity Is Inherited, Not Created
The Q1 2026 trading data separates these hypotheses cleanly, because tokenized assets now span categories whose off-chain markets differ enormously in depth.
CoinGecko's RWA Report 2026 puts total tokenized RWA market capitalisation at $19.32 billion at the end of Q1 2026, up 256.7 percent from $5.42 billion at the start of 2025 2. Within that total, the trading activity is distributed in a way the liquidity thesis does not predict.
Tokenized commodities reached $5.55 billion in market capitalisation, dominated by the gold tokens XAUT at $2.52 billion and PAXG at $2.32 billion 2. Spot trading in tokenized gold reached $90.70 billion in Q1 2026 alone, exceeding the $84.64 billion traded across all of 2025 2. That is roughly sixteen times the category's own market capitalisation turning over in a single quarter.
Tokenized stocks are smaller and trade harder still. The category held $486.69 million in market capitalisation against $15.12 billion of Q1 spot volume 2, a turnover of about thirty-one times in three months.
Set that against tokenized Treasuries, which crossed $10 billion for the first time on 11 February 2026 and ended Q1 at 67.2 percent of the tokenized RWA market 2. The largest category by far generates little secondary trading, because holders subscribe and redeem rather than trade. Tokenized private credit, at $7.96 billion by RWA.xyz's stricter measure 3, shows the same pattern more sharply.
The pattern is consistent and it is the opposite of the claim. The tokenized assets that trade heavily are gold and equities, which were already among the most liquid assets in the world before anyone tokenized them. The tokenized assets that barely trade are Treasuries held for yield and private credit, which is exactly the illiquid category tokenization was supposed to transform.
Why Gold Trades and Private Credit Does Not
The explanation returns to the two causes the NBER authors identified.
Gold has a deep, continuous, globally quoted reference market. A tokenized gold unit can be priced to the cent at any moment by reference to spot, and it is perfectly fungible, so one unit is indistinguishable from another. Asymmetric information is near zero and market depth is inherited from the underlying market. The token is a wrapper on an asset whose liquidity already existed, and the wrapper makes that liquidity accessible on a new rail.
A private credit position has neither property. Its value depends on the performance of specific loans to specific borrowers, which the seller knows more about than the buyer. It is not fungible, because one pool's credit quality differs from another's. Valuation depends on a model and a reporting cycle rather than a live quote, which is why Section 2.4 treated oracle-supplied NAV as a dependency rather than a detail.
Tokenizing that position hands a prospective buyer the same information disadvantage in a faster wrapper. Settlement improves. Price discovery does not.
The Redemption Channel Is Not Secondary Liquidity
There is a real liquidity improvement in this market, and it is worth separating from the one being claimed, because conflating them is how the thesis survives contact with the trading data.
A holder of a tokenized money market fund who wants out generally does not sell to another investor. They redeem with the issuer, and in the better products that redemption runs continuously rather than on a cut-off schedule. Compared with a traditional fund redemption that settles the next business day and cannot be initiated at all on a weekend, this is a genuine improvement in the holder's practical access to cash, and it is the mechanism most tokenized Treasury holders actually use.
But issuer redemption is a different economic object from secondary market liquidity, and the difference matters under stress. Secondary liquidity means another investor takes the position at a price. Redemption means the fund sells underlying assets and returns cash, so the liquidity available to holders is bounded by the liquidity of the underlying portfolio and by the issuer's operational capacity. It is a claim on the issuer rather than a market clearing between two investors.
That distinction is precisely the one that money market fund regulation has spent decades on, because redemption channels concentrate rather than distribute stress. When many holders redeem at once, the fund becomes a forced seller of its own assets, and the tools that exist for that situation are gates and fees rather than prices. The FSB notes that custody and redemption of tokens issued on a distributed ledger may require third-party service providers 4, which adds an operational dependency to a channel that is already the fragile one.
Continuous redemption is therefore a real benefit with a real caveat. It improves the holder's experience in normal conditions and it does not create the two-sided market that would let a holder exit when the issuer cannot accommodate them. Chapter 5 treats that asymmetry as a risk.
The Regulators' Verdict
Securities regulators have reached this conclusion from their own evidence. IOSCO's November 2025 final report found that many of the promised benefits of tokenisation, particularly concerning secondary market liquidity, appear not to be fully achieved for now 5. Its example is precise: secondary trading of tokenized money market funds is limited and restricted to whitelisted investors 5.
That last clause identifies a compounding problem. Even where a tokenized instrument could trade, the compliance layer that makes it legally issuable restricts the eligible counterparty set. A whitelist is a deliberate reduction in market depth, and market depth is one of the two things the NBER research identified as driving the illiquidity discount in the first place.
IOSCO named the structural obstacles to scaling as the lack of interoperability across blockchains and the lack of high-quality settlement assets 5. Both are liquidity problems. Interoperability determines whether buyers on one chain can reach sellers on another, and a settlement asset determines what the other side of the trade is paid in. A market fragmented across chains, each with a partial whitelist and no common settlement instrument, is a market engineered for thin order books.
What Would Actually Create Liquidity
The honest version of the liquidity thesis is conditional, and its conditions are nameable.
Liquidity requires a reference price that both sides trust, which for private assets means valuation infrastructure rather than tokenization. It requires a counterparty set large enough to absorb size, which means either wider whitelists or interoperability that pools them. It requires a settlement asset that a seller will accept, which is why stablecoin settlement matters to RWA markets beyond convenience. And it usually requires committed market makers, who quote two-sided prices because they are compensated to, not because the rails are efficient.
Tokenization contributes to several of these. Continuous settlement makes market making cheaper to run. Common token standards make interoperability tractable. On-chain settlement assets exist and work. What tokenization cannot supply is the demand side, and no engineering decision creates a buyer for a loan pool that a buyer could not previously price.
The result is that the liquidity benefit is real where an underlying market already exists and largely absent where it does not. That is close to an inversion of how the benefit is usually sold, and it means the categories with the most to gain in theory, private credit and real estate, are the ones showing it least in practice. Section 5.4 returns to thin secondary markets as a risk rather than an unfulfilled benefit.
- Illiquidity carries a measurable cost. NBER research on private equity secondaries found most transactions occur at a discount to NAV, with buyers outperforming sellers by a market-adjusted five percentage points annually, driven by asymmetric information and market depth.
- Neither of those causes is a friction problem, so neither is addressed by divisibility or faster settlement. The liquidity argument assumes friction was the binding constraint without establishing it.
- Trading data inverts the claim. Tokenized gold turned over roughly sixteen times its $5.55 billion market cap in Q1 2026 with $90.70 billion of spot volume, and tokenized stocks about thirty-one times, while tokenized Treasuries at 67.2 percent of the market and private credit at $7.96 billion barely trade secondarily.
- Liquidity is inherited from the underlying market. Gold and equities bring deep reference pricing and fungibility to their tokens; private credit brings neither, and a faster wrapper leaves the buyer's information disadvantage intact.
- IOSCO found secondary market liquidity not achieved, with tokenized money market fund trading restricted to whitelisted investors, and named interoperability and the absence of high-quality settlement assets as the structural blockers. A whitelist is a deliberate reduction in the market depth that drives the discount.