Transaction fees: send or wait

[NeuroticFish]

First of all, since nobody can accurately predict the future, please keep in mind that what I'll be telling, although the logic tells they should happen, there's no certainty.
Also, this is for those how use proper, non-custodian wallets.


As you have seen, the transaction fees were quite big lately and some of you know that there were accidents in Chinese coal mines, power outages and miners off grid for a while, causing this.
If you have been following websites like https://mempool.space/ you know also that the things are a bit better now, but the mempool is far from empty and high fees can still be needed.


What's the news?
In 3 days from now there should be the new difficulty adjustment. The new difficulty adjustment is supposed to ensure that the miners find an average of 6 block/hour. And it happens every 2016 blocks (~every 2 weeks). This will make the thing better.
What's even more interesting? This time almost certainly the difficulty will be adjusted downwards, while the miners are basically back.


This means that I expect that for around 2 weeks, starting Saturday (3 days from now), the miners will find a bit more than the 6 blocks/hour, helping the mempool clear up.

All in all, I expect the transaction fees finally go down. So if you have transactions that can wait, if you want to consolidate small inputs and such, the 8-9 May week-end could be a great moment for them.

[Charles-Tim]

The mining difficulty that will be adjusted will really help, but I noticed the mempool has being getting less congested, it even gotten to the extent the fee of 17 sat/vbyte were getting confirmed recently, and some blocks are mined even less than 2 to 3 minutes in such time (I noticed this yesterday and today). The mining hashes have being increasing compared to some days ago which means the miners are getting back to work or some new miners are substituting. But, you are very correct, if 20 sat/vbyte transactions can get confirmed today, if the mining difficulty is adjusted over the weekend which will be adjusted to compensate the remaining reduced mining hashes, more possibly the mempool will be less congested to the extent 10 sat/vbyte transaction (or even maybe less) could get confirmed.

[stompix]

The mining difficulty that will be adjusted will really help, but I noticed the mempool has being getting less congested, it even gotten to the extent the fee of 17 sat/vbyte were getting confirmed recently, and some blocks are mined even less than 2 to 3 minutes in such time (I noticed this yesterday and today).

It's just luck between some blocks, the pace for the last 24 hours is close to what is supposed to be

148 blocks in the last 24hours  
Avg. time between blocks: 9 mins 44 secs

you can have 4 blocks in 10 mins and then half an hour without one, just as I speak 24 mins  

It's not just that the normal pace was back up exactly before the weekend so it helped the cleaning but I can guess a lot of people postponed their transactions or switched to a different coin, unfortunately.

The real problem is that while a difficult adjustment might return things to normal the current "crisis" has shown what's the capacity of the network, you can't tell people endlessly to wait or to use another coin, in the end, they will simply go away and that's it. And exchanges need to start implementing LN, they have one of the biggest footprints in the blockchain, they profit the most from BTC it's about time to put those millions they make to make from fees to good use.

[NeuroticFish]

The mining difficulty that will be adjusted will really help, but I noticed the mempool has being getting less congested, it even gotten to the extent the fee of 17 sat/vbyte were getting confirmed recently, and some blocks are mined even less than 2 to 3 minutes in such time (I noticed this yesterday and today).

It's just luck between some blocks, the pace for the last 24 hours is close to what is supposed to be

That's correct. There were even some lucky blocks down to 11-12 sat/vByte. I expect that we may reach under 10 sat/vByte in the week-end and I hope for under 5 sat/vByte one week later.

you can't tell people endlessly to wait or to use another coin, in the end, they will simply go away and that's it. And exchanges need to start implementing LN, they have one of the biggest footprints in the blockchain, they profit the most from BTC it's about time to put those millions they make to make from fees to good use.

Sad, but true. And the exchanges are not in a hurry to implement, since it's still "advertised" as beta. And even afterwards, they may not change if the(ir) users don't pressure them a little.
Still, we're for now way better than in the worse months of 2017...

[posi]

The mining difficulty that will be adjusted will really help, but I noticed the mempool has being getting less congested, it even gotten to the extent the fee of 17 sat/vbyte were getting confirmed recently, and some blocks are mined even less than 2 to 3 minutes in such time (I noticed this yesterday and today).

It's just luck between some blocks, the pace for the last 24 hours is close to what is supposed to be

As stompix said, it just some luck cause I sent a 31vbyte transaction yesterday and it took 2-4 hours before it gets a single confirmation.
The downward of the difficulty adjustment that will take place in the next 2 days, 17hrs, 30 mins at the time of writing this message will help in term If tx but I don't know if you guys know that it will also cause dump in price of the market probably market correction.

[SquirrelJulietGarden]

Miners will turn on their rigs again and difficulty adjustment won't help the network resolves all waiting transactions in the mempool in a few hours or days. It would take one or two weeks for the network congestion. IMO, the next one or two weekends will give cheaper transaction fees but the size of total waiting transactions in the mempool will still be high.

I think waiting for next two weekends is good.

[Findingnemo]

The mining difficulty that will be adjusted will really help, but I noticed the mempool has being getting less congested, it even gotten to the extent the fee of 17 sat/vbyte were getting confirmed recently, and some blocks are mined even less than 2 to 3 minutes in such time (I noticed this yesterday and today).

It's just luck between some blocks, the pace for the last 24 hours is close to what is supposed to be

148 blocks in the last 24hours  
Avg. time between blocks: 9 mins 44 secs

you can have 4 blocks in 10 mins and then half an hour without one, just as I speak 24 mins  

And now blocks were mined for one hour as of now!

Thanks OP for letting us know about the possible price drop but yesterday I was lucky because when the fee went down to 20 sat/byte I made some of the transactions which I put on hold for more than 2 weeks due to the insane fees.

[pooya87]

To predict the fee market it is more accurate to predict the price because the main thing affecting the number of on-chain transactions hence the fee spikes seems to be the price. Whenever there is a big jump or a big drop there is a lot of people buying and selling bitcoin which requires moving a lot of bitcoin (to and from exchanges) that creates the on-chain traffic hence the fee spike.
Recently the fees came down mainly because the panic sell ended (of course partly because the hashrate went back to normal).

In a couple of days even if hashrate increases and difficulty drops so that miners find more blocks per day, if the price starts jumping up (eg. reaching $70k) we will end up seeing a fee that is higher than the fee from past week.

[pakhitheboss]

The transaction fee is incredibly high at the moment and it did go down a little as compared to what it was a few days back. One of my transaction is still stuck because I accidentally entered 9 sat/byte thinking it will go down in the coming days. A checked on mempool.sapce and up to 10 sat/byte are being confirmed. Waiting for it to go down by this weekend.

[SquirrelJulietGarden]

To predict the fee market it is more accurate to predict the price because the main thing affecting the number of on-chain transactions hence the fee spikes seems to be the price. Whenever there is a big jump or a big drop there is a lot of people buying and selling bitcoin which requires moving a lot of bitcoin (to and from exchanges) that creates the on-chain traffic hence the fee spike.
Recently the fees came down mainly because the panic sell ended (of course partly because the hashrate went back to normal).

In a couple of days even if hashrate increases and difficulty drops so that miners find more blocks per day, if the price starts jumping up (eg. reaching $70k) we will end up seeing a fee that is higher than the fee from past week.

I can not disagree. Price is key of transaction fee changes on the network. Rises or falls will cause higher transaction fees. It is not a matter of how difficulty is and hashrates.

Miners temporarily turn off some of their rigs to exploit panic from senders. Same difficulty, less miners, lower hashrates are only worse for senders. Miners don't lose Bitcoin or satoshi and they indeed eran more Bitcoin or satoshi by confirm transactions with higher or more expensive fees.

[stompix]

Miners temporarily turn off some of their rigs to exploit panic from senders. Same difficulty, less miners, lower hashrates are only worse for senders. Miners don't lose Bitcoin or satoshi and they indeed eran more Bitcoin or satoshi by confirm transactions with higher or more expensive fees.

How do you earn more from fees when you turn off your gear and you're not mining blocks?
So you're forfeiting a block every day for a week for an extra in fees of around ~1 BTC?
Does that make sense to you?

The last block fees were 0.84224237 BTC  the reward was 6.25000000 BTC, how many of those you need to do to compensate for every day you haven't mined?

Miners will turn on their rigs again and difficulty adjustment won't help the network resolves all waiting transactions in the mempool in a few hours or days. It would take one or two weeks for the network congestion.

A weekend at the normal hashrate cut 80vMB from the top, there is another one coming before the adjustment that should force some extra 25 blocks a day, under normal circumstances one week from now on you the only ones left will be 1sat/b transactions.

[NeuroticFish]

Does that make sense to you?

The last block fees were 0.84224237 BTC  the reward was 6.25000000 BTC, how many of those you need to do to compensate for every day you haven't mined?

I'll add that the bigger chunk of those extra/bigger fees will be most likely received by the remaining miners, not those turning off the gears.
I don't know who has come with this idea, I've seen it propagated in other threads too. And it indeed makes no sense. 

A weekend at the normal hashrate cut 80vMB from the top, there is another one coming before the adjustment that should force some extra 25 blocks a day, under normal circumstances one week from now on you the only ones left will be 1sat/b transactions.

We'll have about 1/4 of the week-end at this pace and the rest will be on lower difficulty. Yay!
I didn't want to be overly optimistic and say 1 sat/vByte, but I do expect to fall under 5 sat/vByte anyway.

[Charles-Tim]

I didn't want to be overly optimistic and say 1 sat/vByte, but I do expect to fall under 5 sat/vByte anyway.

This will still depends on mining hashes in relation to how congested the mempool is, normally before the whole thing of reduced mining hashes, 5 sat/vbyte were not even getting confirmed as usual as before, even during weekends, I remembered the last time 5 sat/vbyte transaction getting confirmed exactly on the Easter celebration day. After that day, the mempool was not so not less congested to confirm a transaction getting approximately below 10 sat/vbyte. Before the Easter day, the last time transaction with a feerate below 5 sat/vbyte get confirmed were in February and a time before. The mining hashes with the mempool congestion when bitcoin reached $50000 has not being getting less than 5 sat/vbyte transactions confirmed easily.

There is nothing that can not happen, and your guess can be right, but I think transaction below 5 sat/vbytes may be yet be stuck in the mempool for long as transactions around 10 sat/vbyte or slightly lower are the lowest even getting confirmed even on weekends like on Sunday before the whole mining hashes drop in China due to the blackout you mentioned above. So, the recent mining difficulty adjustment may not still necessary result into 5 sat/vbyte and more lower transaction to get confirmed easily even during weekends. Although, just an assumption.

[NeuroticFish]

I've already sent today a tx at 7.7 sat/vByte and I had to wait only about 30 minutes to get it confirmed.
I guess I was a bit lucky too, but for now the things go almost better than expected  

[stompix]

for now the things go almost better than expected  

And the 5sat/vb has been breached also, the pace is at 22% faster blocks than expected, that's close to ~30 more blocks or 178 per day as we speak.
The change is so ridiculous almost all estimators, including the one in mempool.space is showing every kind of fees, jumping extremely every single block.


Still about 1/4 of a day with little to no traffic so if we reach the 1sat/b it will probably happen only next Monday at best unless something totally out of ordinary happens. But there is a hidden and ugly aspect in this, usage is minimal during weekends, probably real life and try p2p transactions have dropped like a stone.
Don't know if the drop is actually something to celebrate for....whatever don't care anymore.

[NeuroticFish]

Still about 1/4 of a day with little to no traffic so if we reach the 1sat/b it will probably happen only next Monday at best unless something totally out of ordinary happens. But there is a hidden and ugly aspect in this, usage is minimal during weekends, probably real life and try p2p transactions have dropped like a stone.
Don't know if the drop is actually something to celebrate for....whatever don't care anymore.

It's already Tuesday and the mempool is still decreasing (we touched 4 sat/vByte). I'd say that the usage is closer to normal, but the +22-25% pace of mining (bigger than the max 17% I expected) handles all the daily load and then some.
I do celebrate the fact that the problem from previous difficulty period is getting cleared up nicely. Of course, this also means a drop in fees.

P2P transactions may have been dropping indeed, some decided to wait, as I also suggested in this thread. But I am sure that this difficulty period they will send all of those transactions and more, also consolidating the inputs and preparing for future bad days.


I do agree that this kind of situation should not happen and the network should be stable and at reasonable price (not for high priority, but still go through in 10-14 days), but nothing is perfect in life after all...

[Timelord2067]

I do agree that this kind of situation should not happen and the network should be stable and at reasonable price (not for high priority, but still go through in 10-14 days), but nothing is perfect in life after all...

I'm up to 2+ months for a TX sent via the Zap LN wallet https://blockchair.com/bitcoin/transaction/bb128bd08a78bb985a6fb40b78b822c484a58ba0bbc6fa6fd540565d52e1d022 - No CPFP no zaptransaction, nothing available... (I'd inadvertently set the TX fee to one sat, not ten sats...)

Hopefully the number of tX's will keep subsiding ... otherwise ... I'll just have to keep re-sending...  (any suggestions will be gratefully received).

[NeuroticFish]

I'm up to 2+ months for a TX sent via the Zap LN wallet https://blockchair.com/bitcoin/transaction/bb128bd08a78bb985a6fb40b78b822c484a58ba0bbc6fa6fd540565d52e1d022 - No CPFP no zaptransaction, nothing available... (I'd inadvertently set the TX fee to one sat, not ten sats...)

Hopefully the number of tX's will keep subsiding ... otherwise ... I'll just have to keep re-sending...  (any suggestions will be gratefully received).

It should get confirmed in the week-end.
And if it doesn't and you can import the seed/keys to Electrum (sorry, I'm not familiar with Zap), you can increase fee, since it has RBF enabled (just then it will become another transaction, which may matter or not)

[Timelord2067]

It should get confirmed in the week-end.

And if it doesn't and you can import the seed/keys to Electrum (sorry, I'm not familiar with Zap), you can increase fee, since it has RBF enabled (just then it will become another transaction, which may matter or not)

There doesn't seem to be anyway to export privkeys from Zap wallets.  Zap also doesn't allow you to spend unconfirmed TX's (and you can't cancel a TX already in play).

I appreciate your estimate that the TX will clear by the week-end.  Hopefully you are right.

[NeuroticFish]

Wow, very risky option of 2sat/vbyte. The mempool size is steadily decreasing right now, but  it may hardly reach the value that would be   friendly for such transactions.

Really? Today already 3 sat/vByte transactions were included, and in several blocks, including the last two. It's now less than 30 MvB from the top.
And since the tx was first seen 2 months ago, I guess that the sender has the patience to wait a few more days