Coinbase Markets published a derivatives briefing this week. The headline is straightforward: roughly $1.81 billion in BTC and ETH options expire this Friday. Most desks will screen it, nod, and move on. The number is large enough to sound like an event and small enough to be irrelevant. The only genuinely informative content in the report is buried two layers down, in a pair of ratios that contradict each other.
BTC open interest Put/Call sits at 0.66. BTC 24-hour volume Put/Call sits at 0.37. Both readings are below 1.0, so both get filed under the same lazy label: bullish. That label is where the analysis usually stops. It should not. The spread between 0.66 and 0.37 is the only new signal in the document. It does not measure direction. It measures the rate of change of positioning. The standing book leans long calls. The tape leans long calls harder. New money is more bullish than old money.
That distinction is the entire thesis. Everything else in the briefing is either a restatement of known mechanics or a data gap masquerading as a conclusion.
Context: what actually happens at expiry
Crypto options are, for the most part, European-style and cash-settled. Deribit settles roughly 80% of the listed market, and its contract specification is the de facto standard: fixed strike, Friday 08:00 UTC expiry, no early exercise, cash settlement against a spot index. Coinbase's listed product follows the same conventions. This matters because the settlement mechanism determines who is forced to act.
When an option expires in the money, the buyer receives cash. The market maker who wrote the contract has been carrying a delta hedge in the underlying the entire time. As expiry approaches and the price moves toward a strike with heavy open interest, the maker's delta changes. The maker must rebalance. That rebalancing is the only channel through which an options expiry can move spot at all. It is not the expiry itself. It is the hedging flow the expiry forces.
The magnitude of that flow is governed by gamma, the rate of change of delta with respect to price. Gamma is highest near the strike, closest to expiry, and largest where open interest is concentrated. The briefing tells us where that concentration is. BTC call open interest clusters at $90,000 and $100,000. ETH call interest is spread across $3,000 to $4,000. Those are the strikes where maker hedging will be most violent if spot is nearby on Friday morning.
If spot is not nearby, nothing happens. That is the part the headline omits. A $1.81 billion expiry against a total crypto options open interest of roughly $30-40 billion is a 5% event at best. Five percent of a hedging flow, concentrated at two or three strikes, on a single Friday, is not a market-moving catalyst. It is a scheduling detail.
Core: the structure-versus-flow divergence
The 0.66 versus 0.37 gap deserves arithmetic, not adjectives. The open interest Put/Call ratio is a stock measure. It counts every contract currently alive: 0.66 means there are roughly two call contracts outstanding for every three puts, or equivalently, calls outnumber puts by about 1.5 to 1. The 24-hour volume Put/Call ratio is a flow measure. It counts only what traded in the last day: 0.37 means calls traded about 2.7 times as often as puts.
Run the comparison. The stock ratio implies a moderate call tilt. The flow ratio implies an aggressive one. The flow is more than twice as call-heavy as the stock. That is the signal. It means the marginal trader is more bullish than the average holder of the existing book. The book is drifting long. The tape is sprinting long.
The briefing also gives us the ETH side. ETH open interest Put/Call is 0.61 and ETH 24-hour volume Put/Call is 0.55. The same divergence exists, but it is compressed. ETH flow is only modestly more bullish than ETH stock. The BTC flow-stock gap is roughly 0.29. The ETH gap is roughly 0.06. That is a near-fivefold difference in the intensity of the shift.
This is the second-order reading the report never states explicitly. Capital and conviction are concentrating in BTC, not ETH. If the two assets were trading as a single risk asset, their flow-stock gaps would be similar. They are not. Something in the marginal buyer is paying up for BTC optionality specifically.
Check the math, not the roadmap. The roadmap here would be the bullish narrative: calls stacking at $100k, market front-running a breakout. The math says something more fragile. It says a large, recent cohort of traders has bought short-dated upside. Those positions decay. If the move does not come, that cohort does not get to hold. It gets liquidated, or it rolls, or it closes. All three outcomes are selling pressure or a contraction in implied volatility, not a bullish continuation.
Now the strike distribution. BTC calls cluster at $90,000 and $100,000. Both are round numbers. $100,000 is the psychological ceiling every bull market narrative uses as a marker. In 2018, when I audited the weighted constant product formula in the Bancor V2 contracts line by line, the lesson I took away had nothing to do with liquidity math. It was that round numbers concentrate behavior. Suppliers and arbitrageurs all targeted the same invariant boundaries, and the same three edge cases produced the same arbitrage losses every time. Round strikes behave identically. They are where makers hedge and where retail bets co-locate. Crowding at a strike does not predict a breakout. It predicts a magnet. Price often pins to the highest-open-interest strike as expiry nears, because that is where maker hedges are cheapest to maintain.
ETH's call interest is spread from $3,000 to $4,000. Spread interest means diffuse gamma. Diffuse gamma means less pinning, less squeeze potential, and less short-term volatility. The ETH surface is, in effect, saying the market expects a range, not a level.
Contrarian: the Put/Call ratio does not mean what the headline says
The most dangerous line in the briefing is the one nobody interrogates: calls outnumber puts, therefore the market is bullish. A Put/Call ratio below 1.0 does not measure sentiment. It measures a ratio of quantities that have no single owner. A call open interest of X can mean two completely opposite things. Retail is buying upside. Or market makers are selling upside to collect premium. The ratio is identical in both cases. The market implication is inverted.
If retail is buying calls, the tape is a bet on a breakout, and the makers are short gamma. If makers are writing calls into a flat or mean-reverting market, the tape is a bet on range, and the calls are a premium-harvesting product. The briefing never tells us who bought and who wrote. Without that field, the ratio is a coin with one face painted. You can read it, but you cannot spend it.
The same problem afflicts the missing data. The briefing provides call strike distribution for both assets. It provides no put strike distribution whatsoever. That is not an oversight. In a bullish-leaning brief, the downside map is the one chart that would complicate the story. Where is the put protection clustered for BTC? For ETH? At what level are large holders hedging their spot? We do not know. We cannot compute the skew, the downside break-even, or the reflexive floor. We are being shown half a balance sheet and asked to read it as a full opinion.
The dating problem makes this worse. The briefing carries no year and no explicit observation timestamp. From the strike clusters โ BTC at $90k-100k, ETH at $3-4k โ we can infer the window is late 2024 into early 2025. That inference is probabilistic, not confirmed. An options note without a timestamp is triage with the vitals redacted. You cannot compare it to anything.
Add the source constraint. The data comes from Coinbase Markets alone. Deribit holds the majority of listed crypto options volume and a distinctive client mix. Coinbase's derivatives franchise skews institutional. If the sample is one venue, the sentiment reading is one venue's sentiment. The report presents a market mood; it is closer to a single exchange's mood.
The deepest trap is conceptual. Complexity is the enemy of security, and this report layers several conventions on top of each other โ Put/Call semantics, open interest versus volume, gamma versus delta โ and then delivers a one-word sentiment. Each layer is defensible. Stacked, they produce a conclusion the underlying data cannot support. Audits are snapshots, not guarantees. A data briefing is a snapshot too. It captures one frame of positioning and asks you to infer trajectory.
What my own sequencing work taught me about reading derivatives data
In 2024, after the ETF approvals, I spent months computing sequencing centralization metrics across three major Layer 2 networks using on-chain data from January through June. Two of the three routed over 90% of transactions through a single sequencer. The marketing said decentralized. The transaction graph said single point of failure. The gap between the two was not fraud. It was convention. Everyone measured decentralization along a convenient axis, and the convenient axis happened to be the flattering one.
Options data behaves the same way. The convenient axis is the Put/Call ratio, because it produces a one-word answer. The inconvenient axis is the buyer-versus-seller decomposition and the put skew, because those require more data than the brief is willing to publish. When a metric is chosen because it is easy to compute rather than because it is diagnostic, you are not reading analysis. You are reading a sales pitch with decimals.
Code does not care about your vision. A settlement engine does not know that the market is bullish. It knows which contracts are in the money at 08:00 UTC Friday, and it pays accordingly. Every emotion you attach to the expiry is a layer you added on top of an arithmetic routine. The routine will run regardless.
Takeaway: what to actually watch
The $1.81 billion figure is a thermometer, not a trigger. It records the temperature of positioning at a single moment. It does not forecast the weather. Used as an input alongside spot, funding rates, total open interest, and the ETF flow tape, it is marginally useful. Used alone, it is a number that will be quoted in a headline and forgotten by Monday.
The real question is not whether the expiry moves price this Friday. It almost certainly will not, in any way you could trade. The real question is what happens to the flow ratio afterward. If BTC's 24-hour volume Put/Call climbs back above 0.8 in the next two weeks, the marginal buyer has changed its mind, and the aggressive call cohort is already gone. If it stays near 0.37 while open interest also rises, the call tilt is being reinforced, and the $100,000 strike becomes a self-fulfilling magnet. If open interest contracts sharply after expiry without a corresponding move in spot, the bullish cohort got liquidated quietly, and the reflexive unwind has already happened in the dark.
Watch the flow, not the stock. Watch the puts, not just the calls. And watch who is on the other side of every call that gets written. The expiry itself is noise. The direction of the marginal trader, after the dust settles, is the only thing that will still matter in two weeks. If you are not measuring that, you are holding a thermometer and calling it a compass.