Who Quotes a Stock Token at 3am: Hedging, Spreads and the Closed-Market Problem

The question

A stock token is a claim on a share held somewhere in a custody chain. The token moves on a blockchain that never closes. The share moves on an exchange that is open roughly 32 hours a week, five days out of seven, holidays excepted. So when someone sells a tokenised share at three in the morning on a Sunday, who is on the other side, what are they holding afterwards, and what do they charge for the privilege. That is the single most load-bearing question in tokenised equities, because the answer determines the spread, the size you can trade, and what happens on Monday morning.

What the market maker is actually left holding

Buy a stock token at 3am and the dealer who sold it to you is now short a claim on a share they cannot buy. They have taken on the risk that the stock gaps up before the primary market opens. The reverse is worse: a dealer who absorbs your sell is long an asset whose reference market is shut, carrying full exposure to an open they cannot trade into. In a normal equities session that risk lasts seconds and is closed out on the lit book. Over a weekend it lasts up to 65 hours.

This is not a new problem. Futures markets, FX and contracts for difference have priced overnight equity risk for decades. What is new is that the retail-facing instrument itself keeps trading, rather than the retail-facing instrument being shut while professionals price the gap between themselves.

The four ways the risk gets hedged

There are only a few genuine hedges, and each trades one thing for another.

The first is an index or futures proxy. Equity index futures trade nearly around the clock, so a dealer who is long a basket of single-stock tokens can short the index future and neutralise most of the market-wide move. What survives is idiosyncratic risk, the company-specific news, which is exactly the risk that matters when a stock gaps. A proxy hedge is cheap and deep and does nothing at all about an earnings leak.

The second is an onchain perpetual on the same name. Where a perp on the single stock exists and has depth, the hedge is close to exact and settles in the same place as the inventory. The cost is funding, which can turn sharply against a dealer precisely when everyone wants the same side, and the perp's own mark is usually derived from the same closed market, so both legs can be stale together.

The third is inventory limits, which is less a hedge than an admission that there is none. The dealer simply quotes small size overnight and widens as the position grows. This is why tokenised equity books often look fine for a few thousand dollars and thin for a hundred thousand, and why screenshots of weekend spreads tend to be taken at the smallest clip.

The fourth is primary market access, the ability to create or redeem tokens against real shares. That is the only hedge that ultimately clears the position, and it only works when the share market is open. It is the reason weekend inventory is a liquidity problem rather than a solvency problem: the risk has a fixed end date, which is the next open.

What this does to the spread

Spread on an overnight stock token is the dealer's estimate of the gap distribution, plus funding or hedging cost, plus a margin for the possibility that the person trading at 3am knows something. That last component is the one to watch. Classic market-making theory splits a spread into inventory cost and adverse selection, and adverse selection rises when the informed trader has somewhere to go and the uninformed do not. A quiet weekend book is mostly informed flow, so the width is not greed, it is the cost of being the only counterparty to a trader with a reason.

The practical shape of this is predictable. Spreads are narrowest in the hour overlapping the primary session, wider in the hours after the close, widest on Saturday and after a Friday close that preceded scheduled news. Any venue that advertises stock trading twenty-four seven without disclosing session-by-session spread and size is advertising uptime, not liquidity.

Where it breaks

Three failure modes recur. The first is the stale reference price. Many onchain venues derive a stock token's fair value from an oracle that reports the last primary close, which means a dealer quoting around it is quoting around a number that stopped updating. A real event over a weekend, a takeover report or a regulatory action, moves the token while the oracle does not, and automated quoting either widens to uselessness or gets picked off.

The second is the halt. If the underlying stock is halted when the primary market reopens, the dealer's exit disappears and the overnight position becomes an indefinite one. The token meanwhile keeps trading, which means price discovery for a halted security migrates to a venue with no halt mechanism. Rules for that case, whether the token venue mirrors the halt or lets trading continue, are a disclosure worth reading before you need them.

The third is corporate actions landing outside the session. Splits, special dividends and ticker changes have a record date and an effective moment, and if the token's handling of that moment is manual, the weekend is when the mismatch shows up on the screen.

What to watch

Four things tell you whether an overnight stock token market is real. Whether the venue publishes quoted size, not just spread, broken out by session. Whether the quoting mechanism is an automated market maker working off an oracle or a request-for-quote system where a dealer prices each trade and can decline. Whether create and redeem is open to more than one party, since a single authorised participant means the hedge of last resort has a single point of failure. And whether there is a declared policy for halts, gaps and corporate actions, written down before the event rather than announced after it.

The honest framing is that weekend equity liquidity is not free and never has been. Tokenisation does not remove the cost of carrying an unhedgeable position across a closed market, it moves that cost into a spread a retail trader pays directly and can see, instead of a session they were simply excluded from. Whether that is an improvement depends entirely on whether the spread is disclosed.