pull down to refresh

Have we all been overpaying for Bitcoin transactions?

559 sats \ 29 comments \ @Scoresby 27 Aug bitcoin

Most blocks now include sub 1sat/vB transactions

If you survey your mempool, or @wiz's mempool, you will see that since block 905,000 or so, there have been many more blocks with transactions that had sub 1 sat/vB fee rates. In the hundreds of thousands of blocks before this trend, 1 sat/vB was consistently the floor of the fee market. Often, you can find blocks that are not entirely full with a minimum fee-rate of 1 sat/vB.

In the first few hundred thousand blocks of Bitcoin, when blocks were often empty, sub 1sat/vB was much more common.

Average mining fees per day have dropped to lowest levels since 2011

If you look at the amount of bitcoin miners are earning in fees per day, it is clear that the recent trend of sub 1 sat/vB fee rates has reduced the amount of bitcoin miners are earning in fees. So much so that there are articles being published about how miners are seeing low levels of fee-revenue not seen since 2011 (#1194581).

If you look at it in dollar terms, miners aren't doing so poorly, but fee revenue certainly isn't at all time highs.

Bitcoin hash rate is at all-time highs

Yet, if you look at a chart of Bitcoin difficulty adjustments, it is up and to the right. Recently, we have seen all time highs in difficulty levels. To me, this implies that miners are able to make a profit mining Bitcoin -- if it wasn't, we would see hash rate leaving the network and difficulty adjusting down.

None of this takes into account the block reward, which is probably incentivizing people to mine bitcoin even though there isn't enough fee revenue to sustain so much hash rate. However, I think we can ignore this for now, because the blocks are going to get found every 10 minutes on average and the block reward is going to continue to be what it is until the next halving. If it's subsidizing cheap block space, no reason I should have to pay 1 sat/vB when I can enjoy 0.2 sats/vB.

One might also argue that mining is being subsidized by governments or investors (if people have a strong belief that the dollar-price of bitcoin will be much higher in the future, they might be willing to mine at a loss now in order to acquire bitcoin...although I don't know why they wouldn't just buy it if they were so convicted).

My question is: if the fee market settles around 0.4 sats/vB, might that imply that for some of the last few years, Bitcoiners were paying artificially inflated prices to get their transactions confirmed?

view all related items

440 sats \ 1 reply \ @freetx 27 Aug

One might also argue that mining is being subsidized by governments or investors

My nephew works for the cities utility company. They have recently gotten interested in mining to act as demand sinks....particularly for wind.

If you ever drive by the large wind farms, at any given time you may notice that only 50% of the turbines are spinning. This is because there is only enough demand at that time for X amount of electricity, so they put the brakes on the unneeded ones.

If they have mining equipment, it makes more sense to just always run the turbines and power up/down the miners to sink excess electricity. Its literally free money if the wind is blowing and demand isn't taking the full amount....

I think there is some amount of that happening. To that user, they don't care really about economics of mining since the question is: Do you want $0 or >$0?

31 sats \ 0 replies \ @Scoresby OP 27 Aug

It sounds pretty cool! There are so many cases where we have power that isn't wanted at that moment. So batteries are coming along, but batteries might not be the solution for all cases. Bitcoin mining seems like a huge win-win here. Hope your nephew's company actually gets into mining.

212 sats \ 3 replies \ @CruncherDefi 23h

AAAAAAA it annoys me so much: why anyone in those discussion ignores that fact that bitcoin gets more valuable with time? It makes those discussion so intellectually dishonest.

"Average mining fees per day have dropped to lowest levels since 2011"

NO! Bitcoin get over 100x more valuable since 2011. So NO, the average mining fees are NOT the lowest when you consider value. It makes sense that mining fee go down when denominated in bitcoin.

Note that I am not talking about fiat price of bitcoin (price != value). I am talking about value (so value accrued by network regardless of any denominator you might use to measure it). When talking about FIAT everyone here remembers that 'fiat gets less valueable over time' in their reasoning. But when it comes to bitcoin everyone seems to forget that 'bitcoin gets more valuable over time'.

My question is: if the fee market settles around 0.4 sats/vB, might that imply that for some of the last few years, Bitcoiners were paying artificially inflated prices to get their transactions confirmed?

No. The more adoption bitcoin have and the more value is being stored in the network, the more each sat is worth. Therefore todays 0.4 sat/vb is like 40 sat/vb in 2011.

175 sats \ 2 replies \ @Scoresby OP 20h

I agree that fiat-denominated fees have are not as low as btc-denominated fees. See the second chart in the OP.
When it comes to miners who have fiat-denominated expenses, this argument may make sense: if they pay for power in dollars, but earn income in bitcoin, and the bitcoin exchange rate with dollars goes up, then miners don't need to make as much in fees.
However, as far as the payer of the fee goes, I'm not sure I agree: If we imagine a 200vB transaction in 2011 and in 2025, 40 sat/vB is a higher percentage of that transaction than 0.4 sat/vB -- regardless of the bitcoin-dollar exchange rate. The transaction itself is moving bitcoin which is also affected by the change in exchange rates, and so it seems to me that they are moot.
Where it really gets confusing to me is this: bitcoin fees are paid based on amount of data in the transaction not amount of bitcoin. So a fee rate of 40 sats/vB on a 200vB transaction implies that you need to be moving a lot more than 8000 sats. If you're moving 8 bitcoin, you might not be as concerned about paying 40 sats/vB, because I think it is likely that many people would pay more to move a large amount than to move a small amount (even if this is illogical). It may be the case that in 2024, the capital needed to open a lightning channel was actually higher than it is today because lower fees mean you can have smaller amounts of bitcoin in your utxo. However, I'm not really sure how much this changes things, because fees change and you don't want your utxo to get dustified.
@murch made a very interesting point that where this change really lowers fees is in their increment. Prior to this recent wave of sub 1 sat/vB transactions, the default increment for changing a fee rate was 1 sat/vB. Now, it seems that people might do fee-bumping in much smaller increments because it is no longer the standard to bump by 1 sat/vB increments. So, in 2024 you might have needed to RBF from 1 sat/vB to 2 sats/vB to get your 200 vB transaction confirmed (a 100 sat difference), but now you can RBF by a fraction of 1 sat/vB -- potentially RBFing your whole transaction fee up by 1 sat. (There might be some rules around RBF that make me wrong on this, so take it with a grain of salt).

100 sats \ 0 replies \ @Murch 14h

As pointed out elsewhere, we think of the minRelayTxFee as a minimum cost for the network to forward a transaction. When the minRelayTxFee was reduced recently, also the incrementalRelayTxFee was reduced, as it doesn’t make much difference whether you are sending a new transaction or use bandwidth to replace an existing transaction with an updated version of it. In the upcoming Bitcoin Core 30.0 release, they will both be 0.1 s/vB which means that you will need to pay at least 20 sats for the 200 vB transaction from your example for it to be relayed, and if you were to bump it, a replacement of the same size would need to increase the fee by at least 20 sats to be forwarded by nodes.

75 sats \ 0 replies \ @CruncherDefi 19h

I agree that fiat-denominated fees have are not as low as btc-denominated fees. See the second chart in the OP.

When it comes to miners who have fiat-denominated expenses, this argument may make sense (...)

(...) affected by the change in exchange rates, and so it seems to me that they are moot.

That's why I explicitly mentioned 'value' and differentiated it from 'price'. Your points seem to focus on 'price'. My whole reasoning focuses on 'value'.

A metaphor could be Mass vs Weight. Mass is inherent property of a body. Weight is a measurement you get using something else. And on different planet you will get different weights, while mass stays the same. Similarly, if you use different denominating assets (like dollar, euro, eggs or 1 kg of beef) you will also get different prices.

Value is like mass - it's an inherent property of an asset. Price is like a weight - it's a relative measurement when you express value of an asset denominated in a value of other asset (UsdBitcoinPrice == ValueOfBitcoinUnit / ValueOfDollarUnit).

My claim is that not only 'price' of bitcoin increased, but also 'value' of bitcoin increased. And it's obvious why is that the case - adoption is bigger, there are more use-cases, the track record is longer - those increase the inherent value of bitcoin unit.

Hope that clear things up. Todays sats have more value than 2011 sats, so we need to pay less of them for the same thing.

390 sats \ 4 replies \ @Murch 27 Aug

Bitcoiners were paying artificially inflated prices to get their transactions confirmed?

For large parts of the last three years, there were transactions in the mempool bidding more than 1 s/vB or more. For those periods, the answer is clearly no, because the price was determined by the blockspace market, and not the minimum feerate. Beyond that, I was very surprised that the miners didn’t think their incentives through and went along with mining transactions bidding less than 1 s/vB. It seems to me that they permanently reduced the ceiling of feerate peaks, simply by replacements being incremented in smaller steps. If they had thought it through, and considered how little these fees brought in the big picture, they should not have deviated from the established minimum feerate. If they had not performed this unforced error, no in general as well.

Jargon nit: Block reward = block subsidy + transaction fees

0 sats \ 3 replies \ @Scoresby OP 28 Aug

Jargon point: ah, thanks! I was intentionally avoiding calling the subsidy a subsidy for some reason (which I cannot make sense of at the moment).

simply by replacements being incremented in smaller steps

This is a great point. Who's gonna fee bump by 1 sat/vB when they can fee bump by 1 sat? (Perhaps an exaggeration on my part).

102 sats \ 2 replies \ @Murch 14h

To be precise, Bitcoin Core 30.0 is lowering the minRelayTxFee and incrementalRelayFee by a factor ten to 0.1 s/vB.

21 sats \ 1 reply \ @Scoresby OP 14h

I did not know that there was such as thing as incrementalRelayFee and so I had to look it up and of course I found an excellent explanation on Stack Exchange.

Thanks for the clarification and the education!

0 sats \ 0 replies \ @Murch 14h

You’re welcome

102 sats \ 2 replies \ @Solomonsatoshi 22h

How exactly do you suggest that transaction fees were artificially inflated over the last few years?

It's a free market where supply and demand operate to decide the price, surely?

I would however suggest that the narrative of Bitcoin being a speculative commodity to stack and hold, and not use as a P2P payments protocol is reducing transaction volumes and demand and if this trend continues longer term it threatens the entire viability of funding Bitcoins security. The block subsidies decrease in Bitcoin terms predictably and if the speculative commodity narrative continues to overwhelm the P2P payments protocol narrative Bitcoin becomes both less inherently secure and increasingly hostage to institutional and speculator custody and control.

103 sats \ 1 reply \ @Scoresby OP 20h

I was suggesting that fees were artificially inflated by having an artificial floor of 1 sat/vB when it is clear that the actual floor is lower than this. Just in this year alone, any time there were empty mempools and you wanted to get a transaction confirmed, you could have been paying sub 1 sat/vB rates, but because the standard was 1 sat/vB, that's what you paid. It was an artificial (and arbitrary) standard, though.

102 sats \ 0 replies \ @Solomonsatoshi 14h

I see what you mean.

Just opened Electrum and the lowest fee rate does seem limited to 1 sat/vB and there is no obvious way to make an offer lower than 1 sat/vB. When I try to manually set a fee below 1 s/vB I get this message-

'This transaction requires a higher fee, or it will not be propagated by your current server. Try to raise your transaction fee, or use a server with a lower relay fee.'

Clearly some wallets have realised less than 1 can be sufficient currently.

Nice problem to have from a users perspective, but still concerning that transaction demand is so low and use as a MoE is in apparent free fall at least on the primary blockchain.

102 sats \ 1 reply \ @joyepzion 21h

The hashrate staying at ATHs while fees drop is the key insight here. It suggests we're seeing mining operations optimize for efficiency rather than just fee revenue. Many large miners have locked in electricity rates and are playing the long game with BTC accumulation.

0 sats \ 0 replies \ @Artilektt 17h

I was wondering too if OP's observation could also be explained by an unwelcome centralization of mining as I think your post suggests?

113 sats \ 4 replies \ @OT 27 Aug

Yes, but in the end who are you going to complain to about getting your excess fees back. I look back at some of the fees I paid when I first got into Bitcoin in 2017. They were likely from exchanges, but most were tens of thousands of sats.

As I understand 1 sat/vb became standard to stop spam filling up blocks. Now spam is filling up blocks paying as little as possible.

We can't fake it. Bitcoin needs to be used as it was meant to be. People taking custody of their sats.

202 sats \ 3 replies \ @028559d218 28 Aug

Please see my top post which I boosted... #1077925

Bitcoin is strong when it is used... and weak when it's not. 1sat/vb not from a 'cost' perspective but from a 'demand' perspective... means that the transactional component of Bitcoin is not being utilized.

1sat/vb means that blocks are barely full and it/btc is not being spent.

If someone buys their Bitcoin on Coinbase, leaves it there, never withdraws, never runs or uses a node... and decides to "sell" it by clicking a user interface...

What was even the point of Bitcoin? If companies custody it 'for you' it can be seized or censored... and all it is is 'number go up' on a screen that you decide to 'sell' without an on-chain transaction ever happening.

Things are OK now... but if they don't change eventually Bitcoin will fail.

0 sats \ 1 reply \ @OT 28 Aug

I remember this one.

What do you suggest exactly? More education? More institutions or exchanges getting rugged?

50 sats \ 0 replies \ @028559d218 1h

I'm not sure.

The exchanges 'getting rugged' isn't that helpful. Yes it encourages self-custody... but it does a lot of harm too.

We need time and education I think. Bitcoin however is on a bit of a "timer".... if in 10-12 years the fees are still 1 sat/vbyte (or less) I am not sure what will happen.

Bitcoin has a 'mission' of sorts to become transactional... it needs transaction fees eventually. It doesn't want to sit there and be static... it doesn't want to sit there and "do nothing" like gold in this 'captured' gray area...

It wants to be spent, and it wants to be exchanged despite its dis-inflationary nature.

The PTB (powers that be) don't want Bitcoin to serve as an alternative to fiat currencies because it would undermine the power of governments. As long as it's traded like a stock or store of value it's fine.

But used transactionally? To buy things? This is different... the PTB don't want that and frankly most people don't understand the need or "demand" to have something other than fiat to use.

What Bitcoin really is is non-state money, a kind of independent digital capital but the vast majority of people don't see it, don't understand it, think they have to buy a "whole Bitcoin" and just think of it as something too alien and too foreign maybe just a scam.

Or the vast majority of people don't have much in the way of disposable income... so they think "what's the point" "it's not going up" they don't know why they should be attracted to Bitcoin in the first place... if they have 20-50$ to spend. That doesn't buy them "a whole one" etc.

Especially in places like the United States where fiat works "just fine" and Americans have "always had Dollars"... Americans today don't know anything different.

In truth I don't think it's so much "the government" keeping people from using Bitcoin, not really. It's laziness, apathy, ignorance, and just a general lack of knowledge. Dollars are everywhere, so why use anything else?

Bitcoin is used more often in 3rd world/developing/hyper-inflating societies because the need for it is more obvious. It's the "better money" that the government can't provide.

In the United States there are just... Dollars. So why use anything else. How would anything else even be possible? I think these are the more important questions around Bitcoin than even technical ones about 'spam' or types of transactions.

If you can't "buy a whole Bitcoin" and what you can buy "isn't going to go up" that much (at least over a short time period) then why would you use Bitcoin at all? Better to buy a "cheaper coin" that's "going up" on an exchange... that you can sell for Dollars or whatever. I think that's the biggest issue with actual monetary adoption.

Ignorance, laziness, and some kind of NgU greed/unit fallacy that's what is keeping blocks mostly empty I think.

How do you address that?

0 sats \ 0 replies \ @oliverweiss 28 Aug

the transactional component of Bitcoin is not being utilized

Could it be that the transactional component has moved to Lightning? What do the data say?

102 sats \ 1 reply \ @justin_shocknet 28 Aug

This happened because in my trolling of fake L2 scammers it brought awareness to the fact that the minimum fee is the only ownership scaling lever remaining

You're welcome... unless you believe this ultimately destroys bitcoin- in which case then it was the scammers fault (and the scaling virtue signalers that enable them)

17 sats \ 0 replies \ @Scoresby OP 20h

Thank you. I am deeply grateful for the cheap transactions I've been able to get confirmed.

113 sats \ 1 reply \ @kepford 28 Aug

No. Short answer is time preference was high and we paid the fee to match that.

0 sats \ 0 replies \ @perscrutador 28 Aug

We overvalued our time preference.

102 sats \ 0 replies \ @LibertasBR 28 Aug

We were paying higher fees because no one tried to pay less, I think. From what I understand, people started posting transactions with this fee, and they were accepted naturally until everyone accepted them.

I don't know, and I don't have enough knowledge to pinpoint the starting point, but I understand it was natural. Miners could ignore these transactions, but why ignore them if the mempool was somewhat empty?

I understand that user pressure on their nodes led miners to accept these lower fees; it would be very bad to mine empty blocks.

102 sats \ 1 reply \ @Bell_curve 27 Aug

Fees are rarely static... what determines fees?

let's pretend fees 'settle' around 0.4 sats/vB, how many hours or days will that last?

11 sats \ 0 replies \ @Scoresby OP 27 Aug

Good point, but it still might be true that fee rates might have been lower at different points in the last few years.

0 sats \ 0 replies \ @perscrutador 28 Aug

I understand that it was a natural movement of supply and demand. Users entered with the offer of lower fees, and miners met the demand with these fees.