2 sats \ 19 replies \ @south_korea_ln 6 Sep 2023 \ on: Tail Emissions Are A Terrible Idea bitcoin
Peter Todd has done work in that direction: https://petertodd.org/2022/surprisingly-tail-emission-is-not-inflationary
You can also check the bitcoin-dev mailing list discussion https://www.mail-archive.com/bitcoin-dev@lists.linuxfoundation.org/msg11681.html and another post on bitcointalk https://bitcointalk.org/index.php?topic=5405755.0 for further discussion of his blog post. I don't know what his current stance is on the topic.
I've never read that post but I think it's pretty dumb. He solved a differential equation to show that the supply will always be a proportion of growth divided by loss, but implicit to his calculation is the loss of Bitcoin continuing until the supply reaches 0.
It's obvious that if the loss of Bitcoin continues every year at a non-zero rate, it will eventually hit 0. The differential equation wasn't needed at all, and I feel like inserting math where it isn't needed makes someone look too desperate to seem smart lol.
The fact is, economic volatility dwarfs the effect of small amounts of inflation. Even a 0.5% inflation rate over 50 years only leads to a 22% drop. Meanwhile at the time of writing, Bitcoin has dropped 36% in the past year, and gained 993% over the past 5 years. While this discussion is a nice excuse to use some mildly interesting math, in the end it’s totally pedantic.
So because Bitcoin's fiat value has historically fluctuated mostly up, and because Bitcoiners are okay with double digit down swings today, we should be okay with whatever double digit inflation rate tail emissions might incur?
reply
It's obvious that if the loss of Bitcoin continues every year at a non-zero rate, it will eventually hit 0
What do you mean? 0.5^n approaches zero, but never reaches it.
reply
The amount of money someone holds is not continuous and in the context of losing Bitcoin, you either lose Bitcoin or you don’t. If there was only 1 utxo with 1 sat (which is impossible with today’s policy rules), and I said there was a non-zero rate of loss that year, then that means that person has lost that utxo.
When working with discrete everyday things, constant values like .5^n as n approaches infinity just doesn’t make sense as there will always be a point where subdivision is impossible.
It’s the same reason why the dichotomy paradox’s thought experiment can’t actually be done. There gets to a point where it’s impractical/impossible to move half of a set distance.
reply
Bitcoin is not gold, which you can't divide beyond the individual atom, making it discrete at the micro scale.
Bitcoin is code. By the time Bitcoin's supply is a miniscule fraction of what it is now due to lost coins, we'll have forked it to divide 1 sat into micro- or nano-sats. Also, when the supply is that small, tx fees will be denominated in nano or pico sats.
reply
Forking to make sats more divisible is almost like raising the 21 million cap and equally distributing the new coins. Either way you shouldn't assume either is the direction the project will go.
reply
More divisible is nothing like raising the 21m cap. Going from one to two pizzas is different from cutting your one pizza into 4 pieces instead of 2.
reply
Comparing pizzas to Bitcoin doesn't make much sense because pizza is not money and its economic mechanics don't work the same.
Let's take the simplest example of a world with only 2 sats, one held by you and one held by me. We both valued our respective sats at some defined value X. If both you and I decided to make a rule that allowed the doubling in value of all utxos, we would now have 2 sats respectively.
This would be an instance of inflation, but because both of us received sats according to our existing distribution it acts as a halving of value among all sats. Since each of us now have a monetary unit worth X/2 we can effectively treat that as a division of sats. In wallet interfaces, I can even halve the display value of sats such that when someone sends 1 sat, it shows .5 sats instead.
Going back to the pizza analogy, it doesn't make sense in this context because the value of pizza isn't in its usefulness as money. Pizza is typically seen as useful because it provides calories/nutrients/etc. If I were to double everyone's amount of pizza then they would double their value in calories/nutrients/etc. Applying this logic to money would mean everyone benefits from inflation, which we know is not the case.
reply
We both valued our respective sats at some defined value X
There is no single X, value is subjective, but that's an aside.
This would be an instance of inflation [...]
That's a convoluted way to look at it. It's just an increase in precision. More decimal places.
Like expressing the length of something as an integer number of metres. If you need more precision, you go to millimetres. So if your backyard was measured to be 35 m long, if you measure it in mm precision you may discover it's actually 34.983 m or 35.112 m long. Of course it doesn't inflate anything.
In implementation, if BTC amounts are expressed as 64 bit ints (in fact, I reckon only 51 bits are needed), this could be increased to 128 bits if mankind and Bitcoin survive for long enough.
FYI, sats are already 1000x more divisible on LN. This did not change anything in terms of the value attributed to each sat.
But let's agree to disagree :)
I thought it was widely accepted that's the direction the project would go. No controversy here at all.
reply
demurrage
but I think delay of halving is more sure and more robust solution, but harder to introduce
not saying that aproval for demurrage would be a piece of cake :)
reply
The delay of halving breaks the 21M hard cap though, and doesn't even guarantee a finite supply, because the block reward might stay at a certain level forever. This introduces a variable that's hard to predict and completely redefines Bitcoin.
reply
If bitcoin is dying due to lack of free market between active users and passive free-riders - what is the more important thing than to rescue Bitcoin from death?
Free market is more important than finite supply,
and yes - this is the only way to reach equilibrium between active and passive users - by step-by-step iteration to the certain level and stay there forever (just to allow Bitcoin to stay here forever)
I'm glad finally someone understood the idea. Now you need to some time to admit, unpredictable variable is free market at its finest :)
reply
It makes sense. Something akin to what the central banks do by setting interest rates, but algorithmic, leading to a 'natural' (rather than manipulated) level of inflation. But different, because it's directly tied to security.
I haven't thought whether it would be an issue if it works in one direction only (halvings, but no doublings), but it might.
When I was learning about Bitcoin, I heard "It costs you nothing to store your bitcoin (as opposed to, say, gold). You get security for free." and thought it sounded wonderful, but too good to be true. There is no free lunch and all that...
I understand a lack of inflation is aligned with Austrian economics, but the Austrians didn't know a monetary system whose security was tied to inflation. So it's a new concept to wrap one's head around.
But then, I wonder (and again, I haven't thought about it much, it's just some loose thoughts), if Bitcoin's security relies on there being no single entity large enough to thwart it (likely a superpower nation state, like China or the US) and "large" is relative to the size of the population, and if Bitcoin successfully decentralizes power making nation states and corporations weaker (which it may or may not), then that alone may make a 51% attack increasingly harder, mitigating the effect of the falling block rewards.
It's hard to predict what will happen - we're trying to make sure a system whose behaviour is modelled by exponential functions will work well over timeframes that exceed our lifespan.
It's not hard to imagine that even Satoshi may not have gotten it 100% right.
reply
"I haven't thought whether it would be an issue if it works in one direction only (halvings, but no doublings), but it might."
Yes, I'm not brave enough to propose "doublings" yet... ;)
frankly speaking - I'm unconfortable enough with undermining Holy Graal 21M
I wouldn't do that without my sureness that something must be done sooner or later with this long-term embedded problem.
Yes, there is no such thing as a free lunch - I'm a fan of Milton Friedman.
"It's not hard to imagine that even Satoshi may not have gotten it 100% right."
Very true. Satoshi forgot to implement free market between active and passive users, unfortunately.
When you realise that he started the system from one edge case - i.e. inflation starting from infinite in first block in fact, and stakeholders were able to survive this "early" phase with enormous annual inflation only because of "Numbers go up" (i.e. due to system expansion) - and system is simply going by design to the second edge case, i.e. with zero annual inflation.
I don't know if we could find any example of staying in the edge case - as a healthy state for any system. And we are trying to grass-root building of alternative financial system - maybe the biggest challenge I met in my life...
reply