I keep seeing the same sentence every quarter. Billions in Bitcoin options expired today. It sounds like a vault door opened and a pile of coins slid from one desk to another. That is not what happened. A large book of contracts reached a deadline. Some finished with value. Many finished with none. The rest of the story lives in premiums paid weeks ago, hedges already on the books, and two different settlement currencies that do not behave the same way.
The Big Number Was A Snapshot, Not A Receipt
Deribit’s September quarterly Bitcoin options expired at 08:00 UTC on Friday, September 25. The delivery price is not the last print at that second. It is an index time-weighted average from 07:30 to 08:00 UTC. That window is the reference for automatic exercise. Everything else is commentary.
A September 23 preview put Bitcoin options open interest near $16.1 billion, inside a combined Bitcoin and Ether estimate around $18.1 billion. An earlier mid-month figure sat near $16.6 billion across both. Those numbers describe positions still open at an earlier clock time. They are not a settlement bill. Positions get closed, rolled, or opened right up to the cutoff. The dollar translation of BTC-denominated contracts also moves when Bitcoin itself moves.
I’ve found that readers treat open interest like a cashier’s check. It is closer to a guest list. It tells you who still had a ticket. It does not tell you who got paid at the door.
Three Ledgers Hide Inside One Headline
There are three different tallies, and mixing them is how a market event starts to sound like a mass transfer of wealth.
- Open interest counts outstanding contracts before expiry.
- Intrinsic settlement measures options that finish in the money at the official delivery price.
- Net trading profit adds premiums, fees, and any hedge put on earlier.
Those three numbers can live in different assets. Inverse Bitcoin options settle cash flows in BTC. Linear options tied to USDC can produce USDC cash flows under a different contract design. A single dollar sum across both is a scale measure. It is not one pot of dollars waiting to move at 08:00.
An advertised notional is a size label. Settlement is a payoff formula applied to whatever was still open at the cutoff.
Open interest also double-counts if you treat every long and every short as separate piles of wealth. One contract has two sides. Counting both as money that “changed hands” inflates the drama. Counting every dollar of notional as a payout makes a second mistake. An option can expire worthless. It can also finish only a sliver beyond its strike.
What An Option Actually Pays At The Bell
A call benefits when the delivery price sits above the strike. A put benefits when it sits below. That relationship sets intrinsic value. It does not set anyone’s profit. Premiums were paid earlier. Hedges were placed earlier. Fees exist. Rounding exists. Account-level netting exists.
Take a simple illustration, not a claim about today’s print. Suppose the delivery price is $85,000. A call struck at $80,000 has $5,000 of intrinsic value per BTC of underlying before premiums. A call struck at $90,000 has none. A put struck at $90,000 has $5,000. If the holder paid $6,000 for that first call, the exercise still leaves a $1,000 loss after the fact. The strike exposure of $80,000 is not transferred. The gross exercise value is the smaller number.
Convert that $5,000 at $85,000 and you get about 0.058824 BTC on an inverse cash-settled path. Change the hypothetical delivery price to $80,100 and the same call is worth $100. Leave it at $79,900 and it is worth nothing. The pre-expiry notional can look similar in all three cases. The money that posts to accounts does not.
Near-the-money concentration matters more than the giant headline. Distant cheap puts can pad a chart of outstanding risk without producing a payout anywhere near spot. I’ve watched that pattern repeat. A $20,000 put in a market trading many times higher can look huge in notional terms and still settle empty.
Inverse Contracts And Linear Contracts Do Not Share A Wallet
This is the part most round-number posts skip. Inverse Bitcoin options are cash settled in BTC. Cash settled, on a crypto venue, means a ledger credit or debit in the contract’s unit. It does not mean a truck of coins left a warehouse. It does not mean every winning call bought one Bitcoin in the spot market.
Linear USDC options follow another path. Documentation for that design has the product exercise into a future, with the resulting position cash settled in USDC. Calling every Bitcoin option a BTC payout is simply wrong. The instrument code decides the unit. The clearing system nets by account. Public open interest does not show that netting.
There is also an accounting trap. If you convert BTC credits at one price and later revalue them at another, you invent a second dollar figure. A claim that an exact amount “changed hands” has to name its unit and its valuation time. Otherwise the headline mixes reference exposure, coins credited, dollar-valued settlement, and ordinary exchange volume.
Max Pain Is A Model, Not A Clearing Result
Max pain is the strike where aggregate intrinsic value of the snapshot book would be smallest under simplified assumptions. It can be recalculated when positions change. It does not drag Bitcoin to a target like a magnet bolted to the chart. It is not an amount transferred. It also assumes holders share similar interests, which they do not.
Real books mix strikes, futures, spot inventory, and exposures on other venues. A seller who wants one strike out of the money may already be long another option or short futures. A crowded call wall does not tell you whether the holders are hedging inventory, running a spread, or sitting naked. Public strike charts do not name beneficial owners.
Perhaps the most interesting aspect is how often people treat max pain as a forecast and then treat any nearby print as proof. Price can land near a popular strike for a dozen reasons that have nothing to do with options dealers. Macro news, ETF flows, leveraged futures, and ordinary spot demand all sit in the same window.
A Trade, An Exercise, And A Hedge Leave Different Fingerprints
Before the cutoff, a trade exchanges an option at a premium. That can open, close, or transfer a position. If the original buyer sells the contract before expiry, that buyer never sees settlement. If both sides close, open interest falls even while volume spikes. Rolling a September call into October is not a cash gift equal to the September notional. The old position is offset. A new premium is paid or received. That is trading activity. It is not the final exercise amount.
At the cutoff, in-the-money options get contractual value against the delivery price. Out-of-the-money options expire with no intrinsic value. The writer may have collected premium earlier. The short side’s liability pairs with the long side’s receipt through venue clearing. Collateral and netting decide what actually posts.
A hedge is a third event. A dealer who sold a call and bought Bitcoin earlier already executed a spot trade. After expiry that Bitcoin might be sold, held, or reused against another short. An unrelated fund can buy spot in the same minute. A candle around 08:00 is the net of every motive in the room. Direction alone does not prove “expiry flow.”
In my experience, the cleanest way to keep this straight is to ask one question: which ledger am I looking at? The option book, the settlement blotter, or the spot tape? They are related. They are not interchangeable.
| What people say | What it usually is | What it is not |
| Billions expired | Outstanding notional before the cutoff | Cash or coins transferred |
| Max pain level | A snapshot minimization model | A binding price target |
| Cash settled | A venue ledger credit in BTC or USDC | Automatic fiat wires or spot purchases |
| Dealers bought after expiry | A possible hedge unwind | A fact proven by open interest |
Why The September 23 Figure Could Not Be Today’s Payout
That snapshot sat almost two days before expiry. Two days is a long time in options. Traders close risk. They open new risk. They roll into later months. Bitcoin’s price also changes the dollar label on BTC contracts. A figure drawn from one venue does not automatically include every other listed book, over-the-counter inventory, or futures used as hedges.
Even inside one venue, the dollar presentation can use an underlying price that is not the eventual delivery price. To audit a headline you need the observation timestamp, the currency, the contract universe, and the calculation method. Without those, you are reading a weather report from Wednesday and calling it Thursday’s rainfall.
Earlier expiries make the same point. After an August book near $9.6 billion, the market still traded on funding, spot flows, and broader news. A large maturity arriving on a calendar does not isolate price impact. It does remove those dated rights and may change how dealers hedge what remains.
What Would Actually Verify The Amount Paid
A reproducible tally starts with the published September 25 delivery price and a timestamped inventory of expiring instruments just before 08:00 UTC. Each row needs option type, strike, contract size, denomination, settlement currency, and outstanding count. Apply the correct payoff. Sum gross intrinsic value. Keep BTC totals and USDC totals separate before anyone converts them into dollars.
A second layer would check exercise records, trading pauses, fees, and clearing balances. If the question is net transfer by customer or liquidity provider, account-level books are required. Public strike totals will not do. If the question is profit, historical premiums and hedges must enter the math.
As of this writing on September 25, a complete independently verified 08:00 contract-level payout tally was not sitting in public view. So I will not dress the September 23 $16.1 billion Bitcoin estimate as a settlement total. I will not invent a cash-flow figure by applying some assumed percentage. The $85,000 example above is mechanics only.
- Read the official delivery price from the 07:30 to 08:00 UTC window.
- Compare final open interest by strike with earlier snapshots.
- Split inverse BTC options from linear USDC options.
- Estimate gross exercise value and label it as gross, not net profit.
- Look at timestamped spot and futures flow before blaming a later candle on expiry.
Did Anyone Have To Buy Bitcoin After 08:00?
Maybe. Maybe not. Public notional does not show dealer net positioning. Some books get flattened days early because everyone knows the calendar. A huge option notional can vanish at 08:00 while spot barely shrugs. A smaller book can matter if it sits near spot and hedges are twitchy. Both outcomes fit the mechanics. Neither can be predicted from the largest headline number.
Delta hedging can produce buying or selling before the delivery window. As price walks toward a crowded strike, sensitivity can change fast, especially near expiry. Direction depends on whether the dealer is net long or short those options and what else sits in the book. Open interest is not a sign map of those books.
Venue context matters too. Ownership changes and institutional migrations do not rewrite payoff rules. You still have to identify the product, margin currency, and settlement method before adding amounts. Other venues and over-the-counter books can expire on different clocks. A claim about the entire Bitcoin options market needs an explicit universe.
The Honest Answer To “What Changed Hands?”
Mechanically, expiry extinguished dated rights and obligations and posted contract-defined BTC or USDC results for positions that qualified. Numerically, the aggregate transferred across all accounts cannot be pulled from a pre-expiry notional estimate. Some contracts finished without intrinsic value. Those in the money received account credits according to instrument design. Earlier premiums and hedges belong to other transactions and other times.
Two investors can both exercise a winning option and still walk away with opposite net results after premium. A dealer can lose on the option and make it back on the hedge. Count only exercise and you report the contract correctly while misstating who gained from the whole trade. Count the headline notional and you are not even in the right room.
The expiry cleared contracts. It did not clear the headline billions from one side of the market to the other.
That is the part I wish more coverage would say out loud. Size still matters. Traders had a reason to watch a narrow settlement window. Size does not tell you how much was paid, who profited, or whether an observed Bitcoin move was forced hedging. The useful post-expiry update is smaller and more boring: delivery price, strike-level inventory, estimated intrinsic value split by calls and puts and by BTC and USDC, plus a label that says gross settlement, not net investor gains.
Questions Worth Asking After Every Quarterly Expiry
Did the option finish in the money relative to the official window, not a fleeting touch on a chart? Was the product inverse or linear? Was the credit BTC or USDC? How far were the crowded strikes from the delivery price? What share of the book was already closed or rolled? Did spot volume around the window look unusual against similar Fridays?
Those questions will not give you a cinematic number. They will keep you from treating a guest list as a wire transfer. And they travel well. Next quarter the notional will be large again. Someone will say billions changed hands. The contract terms will still say otherwise.
If you trade these events, treat the calendar as a risk date, not a jackpot date. If you only watch the tape, treat expiry as one more flow among many. If you write about it, separate the snapshot from the settlement. That last habit alone would save readers a lot of confusion.
A Practical Way To Read The Next Headline
When the next preview lands, look for four details before you share the number. Timestamp. Venue universe. Whether the figure is Bitcoin only or mixed with Ether. Whether anyone tried to estimate intrinsic value at a stated delivery price. If those four are missing, you are looking at scale, not settlement.
Then watch the window itself. A 30-minute average can differ from a dramatic wick. A strike that got tagged for ten seconds can still finish out of the money. That gap is where a lot of social posts go wrong.
After the print, ignore victory laps that convert every in-the-money call into “forced buying.” Recipients of BTC can sell, hold, or already be offset. Recipients of USDC may have no extra Bitcoin at all. Inferring spot pressure from settlement rules alone is a stretch.
Expiry reading order: 1. Official delivery price 2. Final strike inventory 3. Product and currency split 4. Gross intrinsic estimate 5. Actual trade flow, if you have it
None of this is a reason to shrug at a large book. Large books can still matter. They can change dealer hedges. They can pull liquidity during the averaging window. They can leave the market with a different options surface on Monday. They just do not liquidate the advertised notional as if it were a single invoice.
So what actually changed hands today? Contractual credits and debits for the options that finished with value, in the units those contracts specify, after netting. Not $16.1 billion in Bitcoin marching from losers to winners. The smaller answer is less exciting. It is also the one that matches how the market is built.
