How a Stock Token Prices Itself When the Stock Market Is Shut
Tokenized equities trade around the clock, but the shares behind them trade roughly a third of the week. This is what fills the gap, and what it costs.
A tokenized share is a claim on a security that, in most cases, sits with a custodian and only changes hands during the hours its home exchange is open. The token does not observe those hours. It sits in an automated market maker or an order book that runs continuously, and anyone with a wallet can hit it at three in the morning on a Sunday. The question that follows is simple and badly understood: when the reference market is closed, what is the price on the screen actually made of?
The short answer
It is made of inventory risk, not of information about the company. During market hours, a market maker quoting a stock token has a clean hedge. It can buy or sell the underlying share, or a listed option or future on it, and carry essentially no directional exposure. Its profit is the spread and it can recycle its capital many times a day. When the exchange closes, that hedge disappears or becomes expensive, and the quote stops being a mirror of a price and becomes a bet on where the price will reopen. The spread widens to pay for the bet.
What actually fills the gap
Three things, in descending order of reliability. First, correlated instruments that do trade. Index futures run nearly around the clock, and for a large-cap name with a high beta to the index, a desk can hedge most of its exposure with futures and wear the residual, the idiosyncratic part specific to that one company. Second, overnight and extended-hours venues, which exist in some jurisdictions and have real but thin liquidity. Third, nothing at all, which is the honest description for a mid-cap name on a Saturday afternoon.
The residual risk is the whole story. Hedging a single stock with an index future leaves you exposed to the company itself, and that is precisely the exposure that moves on earnings, guidance cuts, regulatory letters and acquisitions. Those are the events that arrive outside market hours by design. So the hedge works best when nothing is happening and fails exactly when something is.
Why the spread widens, mechanically
A quoted spread covers three costs: the cost of processing the trade, the cost of being picked off by somebody who knows more than you, and the cost of holding inventory you cannot immediately offload. Only the third of these changes much between weekday and weekend, and it changes a lot. Holding a thousand shares of exposure for two minutes until you can hedge is a rounding error. Holding it for sixty hours until the exchange reopens is a position, and it has to be sized, funded and capitalised like one.
This is why weekend quotes on stock tokens are not a liquidity failure so much as correct pricing of a genuinely worse product. A trade you can execute when nobody can hedge it is worth less to the taker than the same trade at ten on a Tuesday, and the spread is the invoice for the difference.
The gap risk that nobody can hedge away
The sharpest version of the problem is the weekend halt. If a company announces something material on a Saturday, the token is the only place that opinion can be expressed, and it will move. When the exchange reopens on Monday, the official print may be well away from where the token traded all weekend. Anyone who made a market in between has eaten the difference. Automated market makers are structurally worse at this than human desks, because a constant-function pool does not know the news and will keep selling into it at algorithmic prices until arbitrageurs have drained the inventory. That loss has a name in the literature, and it is borne by the liquidity providers rather than by the issuer.
There is also a settlement asymmetry worth holding onto. The token leg settles in seconds and is final. The hedge leg, if it is a real share, settles on the traditional cycle and can fail. A desk running this book is short settlement finality on one side and long it on the other, and the funding of that mismatch is a real line item.
Where the mechanism breaks
Three places. Corporate actions, because a token whose reference share has split or paid a dividend is mispriced until the issuer patches it, and the patch is not instantaneous. Halts, because an exchange halt freezes the reference price while the token keeps trading, which converts the token into an uncollateralised opinion about whether the halt resolves up or down. And redemption windows, because the arbitrage that keeps a token honest depends on somebody being able to create or redeem against the real share, and if that window is only open during business hours in one time zone, the token can drift for a long weekend with no mechanical force pulling it back.
What to watch
The useful tell is not the mid price, it is the spread and the depth at a fixed distance from mid, sampled at the same clock time across weekdays and weekends. A venue whose weekend spread is a small multiple of its weekday spread has real hedging capacity behind it, probably via futures. A venue whose weekend book simply thins to nothing has none, and its around-the-clock trading is a marketing claim rather than a market. The second tell is the Monday reopen: compare the first official print to the token's last weekend trade. A venue that consistently reopens near where the token left off is discovering something. One that consistently gaps is not.