Hook
Over the past week, a joint research note from Glassnode and Bybit has been circulating with a headline number that stopped me mid-coffee: options now account for roughly half of Bitcoin's open interest notional, up from about 25%. In the same document, delivery futures volume sits 97% below its 2021 level. Two data points, one implied story — the market has matured, spot leverage has given way to volatility trading, and the adults have finally arrived.
Then I found the only absolute number in the piece. Bybit's options book went from $529 million in its first month to $2.33 billion.
That is a 340% expansion — real, verifiable, and completely unimpressive in context. Deribit's single-day BTC options open interest has historically run in the tens of billions. A $2.33 billion book is a rounding error in the global derivatives complex. The gap between "share doubled" and "size is small" is where this entire narrative lives or dies. The trap isn't the data. It's the denominator.
Context
Let me lay out the plumbing, because the interesting part is mechanical.
Crypto derivatives split into three instrument families, and they are not interchangeable. Perpetual swaps are an indefinite-duration leveraged instrument; with no expiry, they anchor to spot through the funding rate — a periodic payment between longs and shorts that drags price back toward index. Functionally, they are a leverage delivery mechanism for traders who want directional exposure without rolling costs.
Delivery futures carry a fixed expiry and trade at a basis to spot. In a contango market, that basis is harvestable: buy spot, sell the future, collect the convergence. For a decade, that trade was the backbone of offshore crypto market-making.
Options are different in kind, not degree. They deliver convexity. A long option position has bounded downside and highly asymmetric upside; the buyer is paying for the right to be wrong cheaply. You cannot replicate that with a perpetual unless you dynamically hedge, and dynamic hedging imports its own friction. Options are how a portfolio expresses a view on volatility itself — the second derivative of the market — rather than on direction.
What the Glassnode–Bybit note observes is a migration of share toward that third family. Options are climbing in four of five defined market states, and climbing fastest during the long bear phase.
That last observation is the one worth defending.
I've run this forensic drill before. In 2017, as a junior analyst in Buenos Aires, I audited the tokenomics of more than fifty ICO whitepapers and found that 80% of them were funded by speculative liquidity rather than product-market fit. In 2020 I modeled Compound and Aave yield incentives and calculated that the headline APYs were largely borrowed from future token value — a structure dependent on perpetual new capital inflow. In 2022 I mapped the Terra collapse against Federal Reserve liquidity tightening and found the algorithmic failure was downstream of a macro drain, not a code bug. The pattern repeats: a market tells you what it wants to be true, and the accounting is where it confesses.
So when a report tells me options are half the market, I go looking for the other half of the sentence.
Core
Here is where the analysis has to get forensic.
The headline 25%-to-50% figure is a ratio, and a ratio has two moving parts. Options open interest notional is the numerator. Total derivatives open interest is the denominator. If the denominator is contracting — and the same document tells us delivery futures volume has collapsed 97% against 2021 — then options share can rise substantially while options size grows modestly. The cake may be shrinking and being re-sliced, not baking.
The note never discloses absolute open interest. That omission is not neutral. Percentage-of-total framing is the single most reliable way to make a small market look like a large one, and it is deployed here with surgical precision.
Then there is the ETH options share table: Bybit 32%, OKX 26%, Binance 24%, Deribit 12%.
I have to stop on that. Deribit has been the structural center of ETH options for most of the asset's history. Industry estimates of its share, particularly measured on open interest, have consistently sat far above 12% — frequently in the 60–80% band. A 12% print is not a nuance; it is an order-of-magnitude conflict with broadly held market knowledge. Four hypotheses fit. The statistic may be volume-based rather than OI-based, dominated by short-dated weekly paper where retail-heavy venues churn hardest. The four-exchange universe may exclude Deribit's institutional access channels. The note mixes coin-margined and USD-margined volume "to reduce the impact of price movements," an adjustment that sounds rigorous and can quietly introduce its own distortion. Or the window is an outlier.
Any of those is defensible. Publishing the number without disclosing which one applies is not. Volume, open interest, and revenue are three different claims wearing the same word. Moving from "28% of volume" to "market leader" requires vaulting all three gaps, and the note vaults them silently.
The gold data deserves identical treatment. Bybit reports 476 consecutive days as the top venue for tokenized gold perpetuals and 97.1% of gold options open interest. Read that again as a market structure analyst rather than a headline reader. A 97.1% share in a product category is not evidence of dominance. It is evidence that the category has approximately one participant. The higher the share, the smaller the pond — and nobody marketing a report wants you to compute the pond size. This is the only-player effect: the record exists because nobody else showed up. It carries zero predictive content about competitive strength.
The safest number in the document is also the least quoted: the $529 million to $2.33 billion options book. Real growth from a negligible base. Both halves of that sentence matter.
Why does the migration to options make mechanical sense? Because the other two families are being squeezed from opposite ends. The basis trade in delivery futures has migrated to CME, where it sits inside a regulated central counterparty with clean margin treatment and institutional custody. Offshore delivery futures are caught between perpetuals, which serve retail and high-frequency flow better, and CME, which serves institutions better. That is a pincer, and it is not cyclical.
Perpetuals, meanwhile, are structurally incapable of expressing volatility views. As allocators who need to hedge convex tail exposure arrive, options become the only venue for the trade they actually want to do. Nothing was invented here. No protocol shipped. The market simply selected the instrument matching a more sophisticated participant base.
That is why the "bear market accelerates options share" finding matters. If it holds, put demand from hedgers is driving the mix shift rather than call-chasing speculation. That would make this the first cycle where derivatives structure moved because someone was afraid properly rather than greedy loudly.
But subtract the framing bias. "Options gained share in four of five market states" also means options lost share in one. The note doesn't say which. My money is on violent upside gaps, where instant perpetual leverage is what people reach for. Selective disclosure of a 4-of-5 result is a soft form of narrative construction.
One more mechanical check that the report skips entirely: fees. Options typically price at a higher rate per notional than perpetuals, and option market makers negotiate deeper liquidity incentives. If the mix is genuinely shifting, exchange revenue per unit of notional should be rising — a testable, hard-to-fake claim. It appears nowhere. The document leans on volume, which is the most corruptible metric in the stack. Rebate programs move volume. They move open interest less. They barely move revenue at all. Choosing volume as the core lens is a choice, and it is the most favorable one available.
Contrarian
Now the part nobody publishing this note wants on the record.
The study was jointly produced by Glassnode, which supplies the data, and Bybit, which is the single largest beneficiary of every conclusion in it. Of the fourteen information points in the document, at least six directly establish Bybit's leadership across options volume, ETH share, gold perpetuals, and gold options open interest. Zero present a favorable data point for Deribit. That asymmetry is the core diagnostic for research independence. A genuinely neutral market structure report would surface the counterweights: Deribit's institutional client quality, CME's futures share, OKX's multi-asset balance.
This is not a scandal. It is the business model. Exchange-published industry research sits on the marketing line, not the research line, and is budgeted accordingly. But when a data vendor's licensing revenue depends on the exchange it co-publishes with, the circularity is structural. Data providers do not audit their customers into a loss.
The methodological opacity compounds it. The note adjusts for price movement by blending coin-margined and USD-margined volume. It never discloses weights, windows, or sample definitions. Unreproducible research is weak evidence, full stop — the claim isn't falsifiable, so it isn't verifiable either.
There is a second blind spot larger than the first. The entire analysis is arena-framed: four centralized, offshore venues. On-chain options infrastructure — AMM-based option venues, structured product protocols, vault architectures — appears nowhere. If DeFi options share is in fact growing, a four-exchange sample systematically overstates how option-ized the whole market has become. You cannot measure an industry trend using a sample that excludes one of the industry's structural segments and then generalize from it.
And a background fact conspicuously absent: Bybit's early-2025 cold wallet compromise, tied to the Lazarus group, on the order of roughly $1.5 billion. For an analysis of a derivatives venue's competitive position, custody credit and institutional trust are not peripheral details. A report that measures share without measuring counterparty trust is measuring the wrong variable. Chaos is just data that hasn't been sorted by the right axis. Here the right axis isn't market share. It's who absorbs the loss when the venue fails.
That is also where the regulatory layer enters. None of these four venues is a US-licensed derivatives exchange. BTC and ETH options sit in commodity derivative territory, so the securities analysis is clean — no common enterprise, price determined by market, not promoter effort. The instrument is fine. The venue is the risk. If US access restrictions tighten — IP gating, payment rails, enforcement precedent — share can migrate within a quarter, and these percentages become historical artifacts. Bybit's 28% volume share is precisely the kind of number that historically attracts an enforcement memo.
Takeaway
So what is actionable in a chop market where positioning beats prediction?
Ignore the share percentages. Watch three things instead. Absolute option open interest in notional terms — if it grows while total derivatives OI is flat or falling, the maturity thesis is real and the denominator effect is a footnote. CME options volume — if institutional hedging is genuine, it surfaces there first and cleanest. And Deribit's trajectory under Coinbase ownership — if a US-listed entity has absorbed the largest offshore options venue, then "crypto-native exchange" as a statistical category is already obsolete and every share table built on it decays quietly.
The real signal in this document isn't that options are taking share. It's that the derivative mix shifted fastest while price went nowhere. That is what structural change looks like from inside the chop: no candle marks it, no narrative celebrates it, and by the time it becomes consensus, the position is already priced.
The trap isn't the bear market. It's believing a ratio when nobody showed you the denominator. The illusion of infinite growth survives on exactly that omission — and the only way to check is to ask for the raw number, then watch who hesitates.