Network fees as market indicator

[hatshepsut93]

Everyone who was there remembers how at the peak of the last bubble network fees have skyrocketed to $50. There's a simple correlation between those events - when people massively sell, they have to move coins to exchanges first, and this creates a huge amount of transaction that quickly clogs the mempool and leads to high fees. With this we can use the state of the mempool - to estimate how much selling pressure Bitcoin is experiencing.

https://jochen-hoenicke.de/queue/#0,all

Here you can see that the biggest amount of activity was in May 2017, and in Dec 2017, two important months of the last bull run. Compared to them, today much less people send their coin to exchanges (although the blocks are bigger now, and that contributes to how fast the mempool clears), so it's clearly not time to sell yet.

[1714801248]

1714801248

[1714801248]

1714801248

[1714801248]

1714801248

[1714801248]

1714801248

[gentlemand]

I don't doubt fees will rise but this time around we're not going to have a mystery entity AKA Bitmain spamming relentlessly. We might see what a vaguely organic fee market will look like as things properly heat up. I think it's going to look very different.

[hatshepsut93]

I don't doubt fees will rise but this time around we're not going to have a mystery entity AKA Bitmain spamming relentlessly. We might see what a vaguely organic fee market will look like as things properly heat up. I think it's going to look very different.

Well, with that and SegWit, the fee market will indeed look different, and if Lightning will get officially released, that will add too, but on a macro scale, it still should stand - at some point a huge amount of people will be sending their coins to exchanges, signifying that the market is about to burst.

You can clearly see the bear market on the fee charts, because there's so little transaction activity in that period, and then the fees started experiencing spikes again after we entered this new bull run.

[thecodebear]

Segwit and exchange transaction batching will make a bit of difference. But the main reason why fees aren't high yet is because the crowd hasn't rushed in. We're still very early in the bull run now, as opposed to 2017 the fees skyrocketed late in the bull run. Let's see what fees are like in 12 to 18 months if that is when the bull run is coming to a close, fees will probably be wayyyy higher than the $50 they were in Dec 2017. Also a lot more money is coming from institutional players (tho obviously still only a trickle compared to how much money they could and will eventually throw in), this means that the money coming in is more concentrated and therefore many fewer transactions are needed. So yeah the price will go much higher before fees start looking like they did in 2017, but ultimately because the market will be so much bigger this market cycle the fees will get much larger than $50, unless of course those fees stop the bull market in its tracks.

[pooya87]

There's a simple correlation between those events - when people massively sell, they have to move coins to exchanges first, and this creates a huge amount of transaction that quickly clogs the mempool and leads to high fees.

this statement is only correct if you change the bold word to move not sell because fees grow when people move their coins and this can be either from their cold storages to exchanges to "sell" or it can be "buying" on exchanges and moving it to their cold storages to hold. in both cases fees go up which means you can not use fees as indicator of price fall or rise. only as price movement aka market activity.
in other words whenever price is moving up OR down there is a lot more on chain transactions and consequently fees will rise. the bigger the spike or drop the bigger the number of transactions and higher the fee.

[exstasie]

The relationship is a bit more murky/unclear than that. In the last cycle, fees were correlated with price: they both rose at the same time. I don't see how you could have used fees as an indicator that the top was coming. If investors believe BTC is going to the moon, it's not like rising fees will deter them.

Like others have mentioned, it'll look different this time around too so it's difficult to make comparisons. Services are much more likely to batch transactions and optimize for fees now, Segwit is gaining in adoption, etc.

[FanEagle]

I think it is markets going up that affects the network fees not network fees changing affecting the market. What I mean is if the price of bitcoin goes up then the people who see that get more interested in bitcoin which means there are more transactions which causes the network to have higher fees, however network fees do not get higher before the price goes up or something, there needs to be a lot of transactions and that only happens with hype so network fee can't know about it beforehand. That is why its never a good "indicator" its just something you can basically confirm the market with, if the network fees are higher there is a good chance the price is higher as well, it doesn't even to be higher neither but at least that would make sense but its never "if network fee is higher it will go up too" type deal.

[Jating]

As far as I know there was some network spammer even prior to the massive spike in price around Nov-Dec 2017, that why's the network is always clogged.

I'm not sure about this as a good market indicator though, I believed that there are lot of movements as we go along, but with Segwit and exchangers batching their transactions, the network is somewhat alleviated. Fee's is still cheap at this point even though there are a lot of movements.

[buwaytress]

Was really interesting to see that for the first time in a really long time, on the weekend I saw a 70k mempool.

But what was even more interesting as pointed out by gentlemand, no trace if inflated txs. Also, no ridiculous fees even at 70k. Yes, there were those who went for above 100 sat/bytes but I tested a 10.1 sat/byte tx at near the peak (60k) and it still confirmed in 24 hours. A 16.1 sat/byte tx even confirmed in 6 hours. We're talking cents here still even at 11k valuation.

Great demonstration of how there can still be peaks especially in high market sentiment, but how these quickly come down to manageable levels. Everything is getting more efficient. Feels really good.

Not perfect but getting there.

[Naida_BR]

I don't think that it is the same situation.
Indeed, in the previous ATH the fee skyrocketed very much showing that something goes wrong. During this period however, we undergo a halving event something that didn't occur at that period.

[leowonderful]

Hopefully we will continue to have low fees even if the price of Bitcoin continues to go up to near ATH levels and potentially beyond; things are much smoother than they were near the previous ATH where fees were incredibly high.

However, I still don't see fees being a good market indicator. Lightning is also not very widely adopted at the moment AFAIK. I've seen a few services accepting payments through LN like Bitrefill, but most places still don't accept LN payments. Hopefully this changes in the future.

[BrewMaster]

this market is so much more unpredictable than that to be easily predicted only by using network fees. much better indicators that are a lot more logical have failed!
not to mention that you are making a mistake by repeating "people send their coins to exchanges" while you have no way of actually knowing that. they may be cashing out from excahnges after their buys!

[exstasie]

Hopefully we will continue to have low fees even if the price of Bitcoin continues to go up to near ATH levels and potentially beyond; things are much smoother than they were near the previous ATH where fees were incredibly high.

Things have no doubt improved. Transaction batching by exchanges has helped a lot and we're seeing larger blocks due to Segwit adoption. Plus, the big block camp ran out of money to keep spamming the network.

Still, we should expect fees to rise considerably in a new bull market. You can't have a massive increase in users/investment, leading to huge numbers of transfers to and from exchanges, and not see increased congestion on the network. Statistics also show price increases lead to more BTC spending on goods and services = more congestion.

We've only just seen the start now. We can only hope the above factors keep fees rising at a slower rate than in 2017.

[1Referee]

Hopefully we will continue to have low fees even if the price of Bitcoin continues to go up to near ATH levels and potentially beyond; things are much smoother than they were near the previous ATH where fees were incredibly high.

However, I still don't see fees being a good market indicator. Lightning is also not very widely adopted at the moment AFAIK. I've seen a few services accepting payments through LN like Bitrefill, but most places still don't accept LN payments. Hopefully this changes in the future.

The fees will go up organically due to the price going up, so we will not have the fees as low as they were earlier this year. The situation back in 2017 was completely different and shouldn't be used as metric anymore. We can only work with what we have right now, and all indicators point at an increase in fees.

Lightning has little to no effect on getting the fees to decrease. I would say that now the price is going up again, fundamentals are pushed back again which means that there is little to no attention anymore for Lightning, Segwit, etc. People cough up the network fees and seem relatively fine with it. Only when we're going back to $5-$10 fees Lightning and Segwit will pop up again because people have no option than to use them.

[timerland]

Everyone who was there remembers how at the peak of the last bubble network fees have skyrocketed to $50. There's a simple correlation between those events - when people massively sell, they have to move coins to exchanges first, and this creates a huge amount of transaction that quickly clogs the mempool and leads to high fees. With this we can use the state of the mempool - to estimate how much selling pressure Bitcoin is experiencing.

https://jochen-hoenicke.de/queue/#0,all

Here you can see that the biggest amount of activity was in May 2017, and in Dec 2017, two important months of the last bull run. Compared to them, today much less people send their coin to exchanges (although the blocks are bigger now, and that contributes to how fast the mempool clears), so it's clearly not time to sell yet.

I think that it is a good reflection of current market activity, and what the prevailing market sentiment is. I guess you could look at it from that perspective and try predict when the market becomes overheated.

But it does have its limitations. Firstly, I don't think that this is an accurate indicator to predict the future is. Since market activity generally lags behind price, this is really a retrospective indicator, since trading activities significantly increase only after a big market move.

Also given the fact that fees have dropped due to the fact that Segwit, and potentially the implementation of LN on a grander scale in the future, you can't really compare markets between different "eras".

[Distinctin]

I can still remember the last bull run, every time you transact you have to pay for a higher fee and I agree that it's a good indicator.
Most of my earning last bull run was coming from ETH and ETH tokens and I also notice a significant increase of fee, well, I would not be surprise if that happens again and at least it's not new to me.

I'll be monitoring this as indeed it's a great indicator, bull run has no commence yet but it'll be here soon.

[figmentofmyass]

I can still remember the last bull run, every time you transact you have to pay for a higher fee and I agree that it's a good indicator.
Most of my earning last bull run was coming from ETH and ETH tokens and I also notice a significant increase of fee, well, I would not be surprise if that happens again and at least it's not new to me.

I'll be monitoring this as indeed it's a great indicator, bull run has no commence yet but it'll be here soon.

the thing is, it doesn't tell you anything that useful. fees appear to follow price. so fees rise as the price rises, sure, but they'll never tell you anything predictive like when the market is going to crash. it's just a lagging indicator. we're better off looking at the price itself for market signals, ie traditional TA.

[Pursuer]

using network fees as an indicator is the same thing as using google trends as an indicator. in both cases you are using the effect to predict the cause which is impossible. it would be like wanting to predict the weather (like for example if it is going to rain tomorrow) by looking for wet streets and then predict that since streets are wet then it must rain! but in fact you are too late to predicting anything.
in this case fees go up because the price goes up or down not the other way around. so even if you used it as an indicator, by that time it would already be too late and you will be looking at the history not predicting the future.

[Distinctin]

I can still remember the last bull run, every time you transact you have to pay for a higher fee and I agree that it's a good indicator.
Most of my earning last bull run was coming from ETH and ETH tokens and I also notice a significant increase of fee, well, I would not be surprise if that happens again and at least it's not new to me.

I'll be monitoring this as indeed it's a great indicator, bull run has no commence yet but it'll be here soon.

the thing is, it doesn't tell you anything that useful. fees appear to follow price. so fees rise as the price rises, sure, but they'll never tell you anything predictive like when the market is going to crash. it's just a lagging indicator. we're better off looking at the price itself for market signals, ie traditional TA.

That was just one of the trends I'm looking, and I'm only making a judgment on the uptrend, not the down trend.
Sure there's a lot of best method to determine the price movement, particularly the TA but we don't strictly follow that all the time, we still consider some factors that could help us to correctly predict the price movement, and this indicator is certainly part of the factors I'm talking.

[hatshepsut93]

using network fees as an indicator is the same thing as using google trends as an indicator. in both cases you are using the effect to predict the cause which is impossible. it would be like wanting to predict the weather (like for example if it is going to rain tomorrow) by looking for wet streets and then predict that since streets are wet then it must rain! but in fact you are too late to predicting anything.
in this case fees go up because the price goes up or down not the other way around. so even if you used it as an indicator, by that time it would already be too late and you will be looking at the history not predicting the future.

What you are saying is that network fees are a lagging indicators, and it's true, but lagging indicators aren't useless - technical analysis is mostly built on lagging indicators. The idea is that you look at those indicators as if they were a momentum that will stretch into the future. Yes, indicators aren't perfect, no can take the unpredictable into account, but we really don't have much tools.

Looking at network fees can tell us whether the trading is mostly done with the old money, or if new money are entering the market (both Bitcoin and fiat), or if there's a lot of money moving between exchanges. And maybe this can somehow be used to predict the price?