Bitcoin transaction fees are in free fall. This report explains why

Bitcoin transaction fees have collapsed to near-historic lows even as its price sits close to all time highs, creating an unusual situation that challenges conventional wisdom.

Will Owens from Galaxy Digital Research authored a comprehensive report analyzing this, examining everything from the rise of custodians to quantum computing implications for different Bitcoin address types. He joined Bitcoin Season 2’s Writer’s Room to discuss the findings and what they mean for miners struggling with the economics and users enjoying temporarily cheap transactions.

Enjoying the read? Get our newsletter directly to your inbox on all things Bitcoin equities by clicking here.

The following is an edited transcript of the conversation between hosts Charlie Spears and Colin Harper and Galaxy Digital’s Will Owens.

Charlie: Bitcoin’s had a minor decline from 124K to 113K, but what’s different this time is what’s happening on-chain. How important is it that we look at transaction fees right now?

Colin: It is very strange that Bitcoin is ripping to all-time highs and transaction fees are so low. The last time we really saw transaction fees keep up with parabolic moves in Bitcoin was probably 2017 or 2018. You had it a little bit in 2021, but it wasn’t that crazy. The fact that we’re having empty blocks or near empty blocks when Bitcoin’s at $120,000 is kind of wild.

On a longer term view, you want to make sure miners have revenue outside of the block subsidy because eventually that’s going to zero. I don’t think that’s a big problem right now. Some miners are definitely struggling, margins are thin, but the ones with cheap enough power are going to stomach it. The network will always adjust.

My takeaway from reading your report, Will, is we need more people gambling again. We need the Ordinals degens and inscription guys to pump out spam so that miners can get paid.

Charlie: Will, why did you write this report?

Will: I was trying to think of what the best thing to analyze on Bitcoin was right now. I was sitting on mempool.space looking at transaction fees, and so many blocks kept saying “1 sat/vbyte, 1 sat/vbyte.” I thought, how many blocks are there where it’s actually like that?

One of my favorite charts in the report is this percentage of free blocks. This is something that was nonexistent in 2024 because of how the fee market was last year. Once I saw that, I decided to dive in. Everybody’s talking about Bitcoin, apparently everybody’s buying Bitcoin, but nobody’s actually using the chain.

Charlie: Can you explain the difference between median and average transaction fees in your charts?

Will: There’s always going to be random outlier transactions where people overpay. You can see multiples of how much people overpaid in fees. That’s why I overlaid median and average, I feel like it’s a better way to visualize it. Especially during ordinals and runes, there were people willing to pay super high fees just to make sure their transactions went through quickly. There was one block right around when ordinals dropped, literally the halving block. I think it was something crazy like a $200 average fee.

Charlie: What exactly is a “free block”?

Will: It’s not actually free. I defined it as blocks where the average fee rate was less than or equal to one sat per vbyte. This wasn’t happening in 2024. Now we’re starting to see more and more blocks are actually free blocks. Concerning if you’re a miner. We’ve been seeing a lot of these miners pivoting to HPC type stuff for more consistent revenue streams.

As the block subsidy decreases every halving, miners are hoping transaction fees become a bigger piece of the pie. It’s just not happening. I’m not trying to be alarmist, I’m just analyzing the data as I see it.

Charlie: Colin, do miners even realize how low transaction fees are?

Colin: Oh, they’re definitely aware of it. If transaction fees were where they were during Q3 of 2024 with ordinals and inscriptions, they’d be making anywhere from 15 to 20% more from transaction fees. Hash price got really choppy, $35 per terahash per day, all-time low. A lot of miners were near break-even depending on their electricity costs.

This becomes more concerning over a longer timeframe. If we’re still at this point in 4 or 5 years after the next halving, that’s when I’d start to get more concerned. It’s also contingent on Bitcoin’s price. If Bitcoin’s price appreciates fast enough, the subsidy will still be worth quite a lot and it won’t really matter what transaction fees are doing.

Eventually I think Bitcoin mining is going to move closer to the margin where electricity is produced. The massive industrial scale mines we’re used to will take on a different form.

Charlie: Let’s talk about OP_RETURN. Can you explain your transaction composition chart?

Will: In black we have non-OP_RETURN transactions, strictly monetary transactions. In blue we have transactions that contain OP_RETURN outputs.

This could be a lot of different things. Runes transactions are considered this. Recently we’ve been seeing more people using OP_RETURN to anchor data on the blockchain.

For the normal user, it’s not super easy to find this data. You actually have to go in a block explorer and inspect it yourself.

I really like the all-time chart above it, the all-time one.

Source: Galaxy Digital

You can see with the launch of runes, that huge spike in blue where it was previously all black.

About the spike in 2019, I honestly need to look into that. I saw it and only looked into the runes. As a general rule of thumb, if you don’t know what’s happening on-chain, nine times out of ten it’s China doing something crazy.

Charlie: Tell me about Galaxy Digital’s OP_RETURN press release.

Will: Everyone in crypto is aware that a Bitcoin whale from the Satoshi era sold 80 thousand bitcoin. At Galaxy, we announced this in a press release, then thought it made sense to actually announce this on the Bitcoin blockchain. You can see in the OP_RETURN on mempool the actual official announcement is included. It cost $94 total, under $100 to issue this press release. It got a lot of circulation. I think it makes sense for Bitcoin companies to publish notable things like this on Bitcoin.

Charlie: Have you been following the relaxing of OP_RETURN standards in Bitcoin Core version 28?

Will: From what I understand, Bitcoin Core is removing this longstanding default limit of 80 bytes for OP_RETURN data in transaction relay policies. Nodes will be able to accept up to 4 megabytes of arbitrary data in OP_RETURN fields.

An important distinction: it’s not necessary for miners or node operators to relay these transactions. The choice is left up to them. We always have this discourse on Twitter about spam concerns, but perhaps it’s actually beneficial because we need fees to come back. We want to increase on-chain activity on Bitcoin.

Will: Mempool.space did a really good report on this in May. We always hear stuff about quantum computing’s impacts on Bitcoin. Different script types have different quantum implications. Pay-to-pubkey, this early legacy script type, its outputs directly contain the full pubkey. If a sufficiently powerful quantum computer could derive private keys from public keys, these outputs would be immediately vulnerable. That’s over 1.5 million bitcoins.

Pay-to-witness-pubkey-hash recently overtook all the others as having the most bitcoins. Pay-to-taproot is the newest script type, very small in terms of the amount of Bitcoin, but not as small in terms of addresses.

Charlie: Why are certain address types more vulnerable to quantum computing?

Will: It’s all about if the public key is revealed on-chain. For pay-to-taproot address types, the pubkey is safe until the address is spent. If the bitcoins are unspent, the pubkey is safe. We’re actually going to be putting out a longer quantum report on implications for Bitcoin later on, we’ll dive super deep into that.

Charlie: This brings us to “paper Bitcoin summer”, are users actually moving their Bitcoin off-chain to custodians?

Colin: The average user is not going to put up with self-custody. The learning curve is too steep. Those custodial use cases are growing much more quickly. There’s no reason to make any on-chain payments if you’re using custodians.

More trading has also moved to financial products like ETFs. All the Bitcoin treasury companies, that Bitcoin is being held between 2 or 3 custodians. Coinbase has over a million Bitcoin. When trades happen there, there’s no reason for that to take place on-chain. It’s just paper swapping hands. I think that this was kind of inevitable.

Will: When people are buying and selling the ETF, there’s no transactions at the UTXO level. That obviously dampens transaction fees. Also, more and more speculative activity has moved to faster and cheaper chains like Solana. If somebody wants to get into crypto to trade memecoins, there’s no real reason for them to be transacting on Bitcoin right now. The UX is simply not good enough compared to the alternatives. All these degen traders have kind of left the Bitcoin ecosystem.

Charlie: Any final thoughts?

Will: It’s going to be interesting to see how it evolves. Perhaps we’ll start to see some more speculative activity come on to Bitcoin. Obviously everybody’s watching Bitcoin price action, but in terms of transaction fees, it’s not looking too great at the moment. I’ll keep everyone updated. I have all the data, so I’ll be refreshing it every day.

One last thing for users: it’s a really good time to consolidate UTXOs. Very cheap right now. It’s kind of like a temporary gift, cheap and fast transactions.