S16 E60: Jonathan Bier on Bitcoin Civil Wars & BitMEX Research
The audio keeps playing while you browse the site.
Read the transcript30,043 words, every line links to the minute it was said
Ask this episode
Jonathan Bier writes research for a derivatives exchange whose best-known feature was 100x leverage, and he is also the person I would pick to explain what the Bitcoin scaling wars were really about. That sounds like a contradiction. Over three hours and twelve minutes, live, it turned out to be a consistent worldview: look at the incentives, read the filings, distrust the narrative.
The timing was not an accident. Bitcoin is in the middle of another civil war, this one over OP_RETURN relay limits, Bitcoin Knots and a proposed soft fork called BIP-444, and Jonathan thinks it is the blocksize war on a rerun. We also covered BitMEX history, Counterparty, ASICBoost, treasury companies and Ross Ulbricht.
A Research Desk at a Degen Exchange
Jonathan met the BitMEX founders, Arthur Hayes and Ben Delo, at a Bitcoin meetup in Hong Kong, probably in 2015. He had been blogging on Bitcointalk, and the offer was to do the same thing professionally. "Okay, I'll help you guys write a blog," he remembered saying. Arthur's idea came from investment banking: sell-side research that raises the brand and brings in clients. Jonathan thinks BitMEX was the first crypto company to do it, and that many have copied it since. He called Arthur "a phenomenal blog writer in the space", and was generous about the post-FTX comeback.
Then I asked how perpetual swaps actually work, and half of the answer went over my head (I said so on air). The gist: people wanted to go long crypto with leverage and a capped downside. BitMEX's first attempt at tethering the perpetual contract to spot used Bitfinex lending rates, which failed because average leverage was about three times on Bitfinex and twelve times on BitMEX. The fix was a funding rate based on the eight-hour gap between perp and spot. BitMEX also invented the insurance fund, because with no KYC you cannot chase a bankrupt account for the money. "Only when the insurance fund gets depleted does auto-deleveraging occur," he said.
I had not thought about this before. Leverage looks like a simple product until you remember that your customers are anonymous.
Faking the Back End
BitMEX did well in the 2017 bubble because competitors like Kraken buckled under load. Then FTX arrived. "FTX copied the BitMEX front ends, and our products, but they faked our back end," Jonathan said. BitMEX's engine checked every account and every margin balance against actual Bitcoin reserves on every event, and shut down on a shortfall. That was slow, and users saw the infamous order submission error. FTX never had the reserves to check. "Turns out they didn't solve it, they just literally faked it."
Binance is the other half of the story, and Jonathan listed BitMEX's mistakes. The first was stablecoins. "We were too slow to adopt stablecoins, like Tether," he said. After the August 2016 Bitfinex hack (120,000 bitcoins lost) Arthur wrote several critical posts about the exchange. Then came Bitfinex's 0.85 distribution coefficient during the Bitcoin Cash split. BitMEX stayed skeptical and kept its contracts Bitcoin-settled, which made them awkward: an Ethereum perp quoted in a fraction of a bitcoin is a less appealing price chart. When I pointed at CoinGecko showing Binance doing around $20 billion in a day, he shrugged: "I don't know how they became so successful and became the kings, really." BitMEX showed up 47th in the derivatives ranking I had open, and he agreed "it's low relative to peers."
Bubbles, Lightning and Eight Billion People
Jonathan downloaded the blockchain around the end of 2010, when it was about two megabytes, and what hooked him was the math. He had studied elliptic curves at university and thought it was "a very novel use of elliptical curves". The moment he became a believer was the June 2011 Mt. Gox scandal, when rumors said a hacker had taken 250,000 bitcoins. "This thing is like non-reversible," he remembered thinking. (I told him about the first mobile Bitcoin transaction, on a Nokia N900 in December 2009, which I dug up for my Bitcoin Year One comic. He had not heard of it.)
He calls Bitcoin bubbles bubbles without embarrassment. "Bull market or bubbles, you can use either word." He thinks Bitcoin trades at a huge premium to any fundamental measure, such as transaction fees against market cap, but that it "deserves to be trading at a huge premium because it's a pretty neat invention." He also corrected my labor-theory-of-value habit: "The mining cost doesn't cause the Bitcoin price to go up or down."
Sponsor LayerTwo Labs sent a question through me, and I asked whether Lightning was a waste of time. "Definitely not a complete waste of time," Jonathan said, and it shocked him how successful it has been. He is a user. His objection is to the target. He thinks in terms of 10,000 users and "20% year on year growth over the next kind of 20, 30 years", not eight billion people next year. Sidechains got a similar verdict: the hard part, non-custodial Bitcoin on the other side, was underestimated. "If it doesn't have Bitcoin on it, or it's got custodial Bitcoin on it, is it really that useful?"
The Filter Wars Are a Blocksize Rerun
The core of the episode. Jonathan walked through how the current fight built up. Ordinals arrived in January 2023, including a four megabyte Taproot Wizards image. "I really did not care at all, if they're putting images in there," and he thinks it is "a bit immature to kind of get very pissed off and whiny about it." He still expects the people buying those images to lose money.
Meanwhile, relay policy was failing. Pools covering about 99.5% of hash rate were mining full RBF while Core filtered it. Transactions below one sat per vbyte appeared in 40% to 50% of confirmed transactions. Marathon's Slipstream let people bypass the mempool altogether. "So the filters were not working." Core then lifted the OP_RETURN relay limit from 83 bytes, where it had sat since late 2014, to 100,000 bytes, and the people who wanted filters moved to Bitcoin Knots. Jonathan sees the single big number as classic "bike shedding", the same way blocksize was.
He also built evidence. His site, bitnod.es, shows his nodes holding 2,684 peers, at a cost of roughly $300 to $400 a year. "It is one motivated person with a budget," he said, and he could scale it ten times. Filtering is "very asymmetric". I relayed Super Testnet's bandwidth argument for filters, and Jonathan compared relaying to seeding torrents: "The purpose of being a relay is to relay data to other people." Individual node runners are powerful here, he added, but "they can only relax relay rules, not impose them."
I put my own theory to him: "every generation of Bitcoiners wants to have its own crusade, its own blocksize war." He agreed and sharpened it. Filter supporters saw the underdog small blockers win via nodes and a UASF and want that feeling. They even have a Mike Hearn analogy: Hearn's Bitcoin XT proposed going from 1 megabyte to 8,000 megabytes and was accused of trying to "assert dominance", and the OP_RETURN jump is roughly 1,200x. His wider worry is a pendulum. "The small blockers were the toxic side and the toxic side won, therefore toxicity is good," and now everything is too conservative.
Counterparty, Mastercoin and a 4 a.m. Knock in Shanghai
I steered us back to 2014, the first OP_RETURN wars. Jonathan used Counterparty and liked it. "I remember betting on the 2014 World Cup semi-final and final on Counterparty on the Bitcoin blockchain." It also did rock, paper, scissors and dividends. When it used fake pubkeys to store data, Luke Dashjr and Jeff Garzik objected, and Garzik pushed a change to make OP_RETURN standard, with the number bouncing between 40 and 80 bytes before settling. Luke kept complaining and filtered some Counterparty transactions from his pool.
Was this why Vitalik built Ethereum? Jonathan is skeptical. "I think that claim by Vitalik may be exaggerated." He thinks Counterparty failed on timing ("it was just too early"), the ten-minute blocks and rising fees. One difference from now: back then developers' opinions scared people, and he doubted anyone would listen to Ava Chow or Gloria Zhao in 2025. I added the surrounding cast, from colored coins to Mastercoin, the first ICO, whose Omni layer later carried Tether. Counterparty's coin was created by burning Bitcoin, which I called reckless. Jonathan preferred it to Mastercoin anyway: "I much liked Bitcoin's fair launch approach."
His best story was Zcash. BitMEX listed Zcash futures about eight months before launch, "a very weird product because there was no underlying price," and by his recollection the future hit 3,900 bitcoin per Zcash on launch day. He called it "the first future we did of the thing that didn't exist." At Ethereum DevCon 2 in Shanghai in 2016, an attack hit the network at night and someone knocked on his door at 4 a.m. looking for developers. "No, I'm not an Ethereum developer," he said, and went back to sleep. He thinks old conferences were better because "people were there because they wanted to be there."
Nobody Controls Bitcoin Except the Nodes
I had a theory that some of the hostility toward Core developers is cultural. The loudest maximalist lifestyle is steak dinners and push-ups, and the people maintaining the software look nothing like it. Jonathan agreed there was culture in it, but added a misunderstanding about what maintainers are. "They don't control Bitcoin. They just control a piece of software." His analogy was a plane: the engineers who maintain the safety systems matter, but "they don't represent the people that use the airplane." On the complaints about Gloria Zhao's work history: "You want her to have more work experience before she becomes a maintainer?" And on Matthew Kratter's rule that a Core developer should be at least 35, he asked: "Sounds like it comes from the US Constitution, right?" Anyone demanding vetted credentials, he said, "should invest in the US dollar." Bitcoin Core "is just a piece of software."
On dilution I was gloomy. I described a bucket brigade delay pedal where each repeat of a sound gets worse, and said "we are lowering the level to the lowest common denominator." Jonathan disagreed gently. "There's always been a lot of idiots in crypto and in Bitcoin." He recalled being mesmerized by Andreas Antonopoulos saying Bitcoin was "an unstoppable freight train", and being lost in the 2014 Amsterdam RBF debate between Peter Todd and Gavin Andresen: "I have no idea who is right." His conclusion was fairer than mine: "Yes, there were idiots in the past and yes, we were idiotic ourselves in the past."
On BIP-444, which he believed was due to activate on 1 February 2026, he was blunt. By limiting Taproot script sizes it could freeze coins sent to outputs that can only be redeemed with large scripts. "I think that BIP-444 is very dangerous in and of itself." It also fails at its own job: Peter Todd put the full BIP text on chain in a compliant form, so "you can just have BIP-444 compliant spam." If the supporters fork, "that new chain will be a huge spam magnet, in my opinion." He runs a BIP-444 client on forkmonitor.info anyway. I raised the slippery slope toward OFAC lists and needing "a license, essentially, to run a node." He replied, "It's not just a slippery slope fallacy."
Who Won the Blocksize War
We spent an hour on Jonathan's own history. His view is that the small blockers were the underdogs for much of the war. The large block side had the VC-backed companies, Bitmain and the people who spoke San Francisco. They lost because "the large blockers made a lot of tactical mistakes and shot themselves in the foot." He regrets the split, though. "We lost a lot by losing the large blockers," he said, because "the large blockers were like the sales team and the small blockers were the engineers." His evidence for Roger Ver as a salesman is mixed. At a 2014 London summit, Roger shouted at bankers about the price: "I promise you the Bitcoin price is going to go up 100x." Right, Jonathan said, but the bankers thought he was a maniac. He also watched Bitcoiners try to give a London toilet attendant a paper wallet instead of the pound he expected.
I argued that Ethereum was the winner, since builders walked away from Bitcoin. Jonathan added the other side: leaving "meant that the smaller blockers became dominant in Bitcoin." He thinks Craig Wright "inadvertently had a huge positive impact for the small blockers." On what separates the camps, he refused to name a number. "The biggest single inflection point in my view is this argument over surplus capacity." Large blockers wanted blocks bigger than demand, which he calls "an economically flawed idea" now that people would fill the space with images.
He also described the November 2017 peak, when SegWit2x was cancelled and "the Bitcoin Cash price reached 51% of the Bitcoin price", largely on Korean speculation. A Telegram group of large blockers was called Chain Death. His own fear was small: "there's a 5% chance they're going to win the war."
A chat question about ASICBoost made him pull a DragonMint T1 off the shelf behind him. (I guessed WhatsMiner. Wrong.) Covert ASICBoost, he explained, exploits the last four bytes of the Merkle root, and SegWit's commitment would break it. He doubts Bitmain actually used it in live mining: "I don't think it's likely that Bitmain were really doing covert version rolling in the live network." But the press release claiming the patent made people believe it. InnoSilicon built overt version rolling in response, and he thinks 80% to 95% of blocks use it now.
Fees, Tail Emission and Treasuries
On the future, I asked whether the fee market can replace the subsidy. Jonathan thinks "the lowering of the min fee rate is a very good thing", but fees are still too low. A fee market is the better spam filter: "The highest bidder gets in there." What he wants is more real spending. Peter Todd's tail emission is worse in his view, for political reasons: "How will we ever agree on the right percentage tail emission?" and "I probably consider Bitcoin a failed experiment if we have to have a perpetual tail emission." I said he was going full Mike Hearn. He agreed he would rage quit.
On quantum, he wants a new way to spend first: "I think the first step is we create a quantum safe way of spending Bitcoin," even at four kilobytes a signature, with users paying for the space. He thinks "a hard fork is very unlikely" and that the filter movement will make consensus changes "harder to do." James O'Beirne's presentation on raising the blocksize came up. Jonathan sees spam as a good argument against raising the limit.
Then treasuries. His latest article covers the advisory agreements treasury companies sign, which he thinks favor the advisers. "One maybe should be bearish on some of these treasury company stocks." Michael Saylor is doing the right thing for MicroStrategy, he said, but the instruments are "not good for the investors". I asked how this differs from a Ponzi scheme. "The main difference is that they're very transparent about what they're doing." The similarity is that they are "dependent on a constant new inflow of money in order for them to operate the way they're operating." Investors can lose heavily without fraud. "Read the SEC filings yourself." He disclosed his Bitwise role and said he would steer people toward ETFs over treasury stocks.
Pardons, Cypherpunks and the ETF Layer Two
We ended on the cypherpunk question, with Ross Ulbricht's pardon in mind. Jonathan noticed that early Bitcoiners both believed in libertarian ideals and expected to get rich enough to buy political influence, and that happened: "We got Ross Ulbricht free." He called Arthur Hayes' pardon "a symbolic pardon". For Curtis Green, who was a guest on the show, he thinks the government's own conduct in the Silk Road investigation strengthens the case: "he deserves compensation in addition to a pardon." We compared Ulbricht to Aaron Swartz. The system was harsh because he "didn't just break the law. He potentially undermined the law."
Does cypherpunk culture scale? "The culture was never going to scale." I tried my line that the ETF is Bitcoin's scaling layer. He took it literally: "It's a custodial layer two system," and he compared BlackRock's IBIT to WBTC. "The mechanics of BlackRock's IBIT and WBTC on Ethereum, in my view, are quite similar." I argued IBIT is not an open protocol and is not as fast to send. He pointed out that you can borrow against it with a few clicks and said he was "just comparing it."
Two last answers. Asked for altcoins that are not shitcoins, he went by fair launch: "I like Bitcoin, Monero, Dogecoin, Litecoin." Asked whether Satoshi was a big blocker, he gave a split verdict. Satoshi favored "rules being set in stone" and called a second client a menace, but also compared Bitcoin to Visa volumes in the reply to James A. Donald. Jonathan's advisory and grant roles cover privacy work such as PayJoin and silent payments, and I appreciated that he did not try to sell me anything.
I signed off by telling him not to filter his nodes, which he took well.
Jonathan's account of the large blockers pairs with the view from the other camp that I later recorded with Kurt Wuckert Jr on the big block wars, BSV and the road not taken, and the Counterparty history overlaps with Adam Krellenstein on Counterparty and Bitcoin tokens.
Listen to Jonathan Bier on Spotify, Apple Podcasts, YouTube, and More
Time stamps:
- 00:01:11– Introducing Jonathan Bier and Blocksize War Book
- 00:02:44– Meeting BitMEX Founders and Starting Research
- 00:05:31– Arthur Hayes' Blog Writing and Vision
- 00:07:32– BitMEX Success: Leverage, Perpetual Swaps, Troll Box
- 00:10:43– FTX Copying BitMEX and Backend Issues
- 00:11:20– Binance's Rise and Market Dominance
- 00:13:49– BitMEX's Slow Adoption of Stablecoins
- 00:16:39– Current BitMEX Volumes and Rankings
- 00:17:39– History of Bitcoin Derivatives and Innovations
- 00:23:49– Jonathan's First Bitcoin Encounter in 2010
- 00:26:12– 2011 Bubble and Mt. Gox Scandal
- 00:28:02– Views on Bitcoin Bubbles and Valuations
- 00:30:19– Ad: Bitcoin.com News
- 00:31:00– Layer 2 Labs Question: Is Lightning a Waste?
- 00:32:08– Lightning Network Success and Limitations
- 00:35:02– Scaling Challenges and Sidechains
- 00:37:30– Current Filter Wars and Ordinals/Inscriptions
- 00:39:40– History of Relay Policies and RBF
- 00:43:49– OP_RETURN Limits and Debates
- 00:47:04– Bitnod.es Website and Node Peers
- 00:48:39– Ineffectiveness of Filters
- 00:50:00– Super TestNet's Bandwidth Argument
- 00:51:42– Generational Crusades in Bitcoin
- 01:01:55– London Meetup Stories and Fanaticism
- 01:15:00– Loss of Merchants Post-Blocksize War
- 01:18:13– Ethereum Benefiting from Bitcoin Decisions
- 01:18:48– 2014 OP_RETURN Wars and Counterparty
- 01:23:03– Reasons Counterparty Failed
- 01:27:31– Mastercoin, First ICO, and Omni/Tether
- 01:29:37– Zcash Launch and BitMEX Futures Shenanigans
- 01:33:26– Ethereum DevCon 2016 Attack Story
- 01:35:04– Conferences Then vs Now
- 01:38:00– Cultural Divide in Bitcoin Maxis
- 01:39:46– Misconceptions About Bitcoin Core Developers
- 01:43:25– Youth of Early Developers Like Tamas
- 01:46:59– Criteria for Developers and Unicorn Myth
- 01:48:41– Educators in Bitcoin History: From Satoshi to Saylor
- 01:52:05– Dumbing Down of Bitcoin Narratives
- 01:55:34– Past Idiocy and Learning Curves
- 01:56:50– BIP 444 Dangers and Ineffectiveness
- 02:00:59– Irony of Anti-Spam Fork and Spam Magnets
- 02:01:22– Parallels to Blocksize Wars, Bitcoin Cash Price Peaks
- 02:03:30– Excitement and Fear During BCH Surge
- 02:04:28– Reasons Big Blockers Couldn't Compete, Craig Wright's Role
- 02:05:16– Satoshi's Response to Scaling Concerns (2008)
- 02:06:05– Jeff Garzik's Early Block Size Proposal (2010)
- 02:07:33– Early Onchain Exchanges and Scaling Realization
- 02:07:58– Defining Small Blocker vs Big Blocker
- 02:08:56– Surplus Capacity in Blocks Debate
- 02:10:10– Cultural Biases: Technical vs Business Sides
- 02:11:32– Competent Developers on BCH Side, Inflation Bug
- 02:12:35– Block Size Numbers Not the Core Issue
- 02:13:29– Modern Block Size Increase Justification
- 02:14:15– Hard Forks Unlikely, Quantum Signatures
- 02:15:03– James O'Beirne's Block Size Presentation
- 02:15:54– Filter Wars Impact on Future Changes
- 02:17:13– Spam as Reason Against Block Size Increase
- 02:18:18– Quantum Safe Spending Proposal
- 02:19:31– Fee Market vs Crusades Against Spam
- 02:20:45– Craig Wright's Unintentional Help to Small Blockers
- 02:22:04– Jihan Wu's Dislike of Craig Wright
- 02:23:34– ASIC Boost Role in BCH Creation
- 02:24:50– Covert vs Overt ASIC Boost Explained
- 02:30:00– Treasuries and MicroStrategy Debt Structures
- 02:35:00– Bitcoin Treasuries vs Ponzi Schemes
- 02:40:00– Balancing Professional Life and Cypherpunk Roots
- 02:42:31– Advisor Roles: Bitwise, Brink, Maelstrom
- 02:53:16– Privacy Tools: Payjoin, Silent Payments, CoinJoin
- 02:54:41– Ethereum vs Bitcoin on Privacy Support
- 02:55:45– Ross Ulbricht Pardon and Government Misconduct
- 03:00:13– Curtis Green's Torture and Fake Death
- 03:03:00– Aaron Swartz Comparison
- 03:04:29– Cypherpunk Culture Not Scaling
- 03:05:53– ETFs as Bitcoin Scaling Layer
- 03:07:39– IBIT vs WBTC Similarities
- 03:09:00– Three Legit Altcoins: Monero, Dogecoin, Litecoin
- 03:10:02– Satoshi's Views on Block Sizes
- 03:11:02– Closing and Follow Jonathan Bier
Verify this episode
This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.
- SHA-256
a4ef27023e56ed3a910a1e0dd39f527962f94254f968ecb27fef5c9b91e91345- Proof
- 2025-11-24_s16_e60_jonathan_bier_on_bitcoin_civil_wars_bitmex_research.mp3.ots · all episode hashes (proof)
- Status
- Anchored in Bitcoin block 974,312
- Timestamped
- After release. The episode came out on 24 Nov 2025, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
- Download the MP3 and the proof into the same folder.
- Run
sha256sum 2025-11-24_s16_e60_jonathan_bier_on_bitcoin_civil_wars_bitmex_research.mp3(on a Mac:shasum -a 256 2025-11-24_s16_e60_jonathan_bier_on_bitcoin_civil_wars_bitmex_research.mp3). The result must match the SHA-256 above. - Install the OpenTimestamps client (
pip install opentimestamps-client) and runots verify 2025-11-24_s16_e60_jonathan_bier_on_bitcoin_civil_wars_bitmex_research.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node,ots info 2025-11-24_s16_e60_jonathan_bier_on_bitcoin_civil_wars_bitmex_research.mp3.otsshows the block height to look up in any block explorer.
Related
S15 E53: Charlie Shrem on the Block Size Wars, Life After Prison & Bitcoin's Leaderless Future
BitInstant co-founder Charlie Shrem clarifies his block size war stance, recalls prison and rebirth, defends Tether and explains…
S17 E13: Charlie Spears on Blockspace Media, the Purity Spiral & the Case for Classical Maximalism
Editor’s note: this conversation was streamed live in mid-March 2026, a week after the US switched to daylight saving time (which…
S17 E23: Hunter Beast & Mark Karpeles on Quantum-Proofing Bitcoin, Satoshi’s Coins & the Ghost of Mt. Gox
This interview took place on April 29th 2026, at the LayerTwo Labs booth at the Bitcoin Conference in Las Vegas with Hunter…
Comments
Comments run on my own server. No trackers, no tracking cookies.