What is the ideal inflation rate in an inflationary PoW currency?

[americanpegasus]

I have started up some discussion on the possibility of moving an experimental Cryptonote to an inflationary tail emission model.  This idea stems from trying to stick to the "0, 1, or infinity principle" (https://en.wikipedia.org/wiki/Zero_one_infinity_rule).
  
It would seem that we already have a 'zero' tail-emission coin: Bitcoin.  We also already have a '1' tail-emission coin: Monero.  (has a perpetual tail emission so that there will always be Monero to mine after the initial distribution, but since it's a fixed block reward that will decrease percentage-wise over time).  I have suggested we consider making Aeon the first coin to embody the 'infinity' principle by instituting inflationary tail emission.  
  
 
My core question to the economics section is this:  If there were blockchain designed to focus on spendability and consumer friendliness, it might make sense to create a perpetual inflationary emission after the initial distribution had ended.  What is the ideal inflation rate for a currency?  
  
The online research I have done has suggested that most economists agree it is likely somewhere between 0.7% and 2.4%.  I know we have some very bright minds here, so I want to know: what is your opinion on this? (if possible, please back this up with references)

[1715052832]

1715052832

[1715052832]

1715052832

[1715052832]

1715052832

[americanpegasus]

It should also be noted that users Johnny Mnemonic and GingerAle have proposed a new type of "smart inflation" that would change rates depending on the velocity of the money.  Thoughts and opinions are welcome regarding this too.  Perhaps there would be a lower and upper bound, and the rate would vary between those two extremes depending on the velocity of the network?  

It would be great to see a "smart inflation" of sorts, where the rate of debasement is determined somehow by the transaction volume. This way, the money supply grows in proportion with the network.

I imagine this can be achieved by looking at the average block size or transaction fee over a fixed period and adjusting the block reward accordingly.

yes! This! If money velocity is high (i.e., there are a lot of transactions in a block), that means money is valuable and people are using it, and therefore minimal (to zero) inflation is necessary. If money velocity is low (there are few to little transactions in a block), that means money is \too valuable so people aren't spending it. Of course we're diving headforward into economic theory.... of course here there'd be a maximum inflation rate and a minimum, being 0. Of course the magic number question is then what is the proper velocity of money.

 

[smooth]

I suspect that nothing we do now would be first. Pretty much everything has been tried.

I don't really know how to measure real usage given that sham usage can be created. I guess observed usage is always a bound on actual usage but not at all a tight bound.

[americanpegasus]

I don't really know how to measure real usage given that sham usage can be created. I guess observed usage is always a bound on actual usage but not at all a tight bound.

 
  
True, it seems to remain an unsolved problem how to measure real usage (especially in a Cryptonote) without leaving the system open to manipulation by false transaction volume.  Additionally, creating artificial transaction volume to manipulate the inflation rate would be a problem because it creates blockchain clutter.  Hmmmm.  I also suspect that perhaps the issue is irrelevant: the zero, one, or infinity principle would hold true with respect to (reasonable) inflation rates as well.  
  
Explanation: Either you agree to zero inflation, a rate of inflation, or eventual infinite inflation (what we have now with fiat imho).  
  
Perhaps there is some way to make it economically unfeasible to manipulate the inflation rate, even if it would be mathematically feasible to do so.  Still, more discussion on hard rates along with possible implementations for "smart inflation" is encouraged.  

[Johnny Mnemonic]

There was some good discussion regarding debasement on the Economic Devastation thread:

https://bitcointalk.org/index.php?topic=355212.msg11923878#msg11923878

3. Due to that squared law explained in #2 and coupled with the fact that small things and investments grow faster, the upcoming wealthy gain more from debasement than the egregiously wealthy. Thus debasement is a very natural, efficient allocator of wealth to the maximum production. The super wealthy can't focus their investments to the most productive because their wealth is too large to allocate efficiently. They are more concerned about safety, economies-of-scale, and return of capital, which retards production in the economy. [implication is that the more wealthy someone is, then the more reliant on usury for return-on-investment instead of non-guaranteed return (a.k.a. return versus risk)]

Thus raising the debasement rate redistributes capital from lower efficiency investment (and thus production) to higher efficiency investment and production. Thus benefiting everyone in society.

Thus the debasement rate should be set as high as such that a super majority of society can receive an increase in their personal purchasing power (personal priorities) that exceeds the debasement rate. This will not stop members of society from competing for higher return-on-investment, because to be in that super majority you must compete.

Whereas, if it were true that everyone had the same priorities and we could set the debasement to one monolithic purchasing power metric, then no one would have an incentive to compete. This is yet another example of why uniform distribution is lifeless. I mentioned this abstract concept in my two blogs, The Universe and Information is Alive!.

The problem is how to measure that in order to set an ideal debasement rate? The debasement rate must increase and decrease with some feedback loop which is the efficiency of production relative to a super majority of the society.

Referring back to my explanation about the Quantity Theory of Money:

1. The Quantity Theory of Money roughly posits that the GDP (price × quantity) is roughly proportional to the money supply (M) times the velocity of money (V). Any increase in GDP due to increase in M (such as fractional reserve debt), is illusory non-growth in the form monetary inflation. Any increase in GDP due to increase in V that is not a derivative effect of monetary inflation, is real growth. Real growth translates to an appreciation of buying power, thus protecting store-of-value. Thus we can conclude that friction on unit-of-exchange is undesirable.

We see that Δ real growth is somewhat related to Δ transactions × price. Thus such a feedback loop could modulate the Δ debasement rate by the Δ transactions × price.

The variables could be:

1. Debasement is a percentage of the transactions x price.
2. Relationship of this debasement to the transaction fee, if any.

Let's assume transaction fees are set by the market and are a better proxy for the real value to society of the transactions than transactions x price, because transaction fees modulate priority. Also market determined transaction fees are society's appraisal of the (equilibrium of tolerable or neglible) cost of unit-of-exchange which is a primary function of money w.r.t. commerce that demands supply of production, thus medium of flow in the power-law wealth distribution circulating pump.

Perhaps:

debasement = factor × transaction fees

Given annual velocity of money has historically been around ≈2.0, then if factor = 1, annual debasement ≈ 2 x weighted average transaction fee percentage. In other words with a factor = 1, an annual velocity of money = 2, and a weighted average transaction fee percentage of 1%, the annual debasement rate would be 2%.

[americanpegasus]

It can only be assumed that if TPTB has already mulled over this, then perhaps he already has plans for some form of smart inflation in his coin. 
 
It wouldn't be our intention to step on his toes, but perhaps he is willing to offer his opinion in this matter so that we can better align/complement/merge goals.

[n2004al]

I have started up some discussion on the possibility of moving an experimental Cryptonote to an inflationary tail emission model.  This idea stems from trying to stick to the "0, 1, or infinity principle" (https://en.wikipedia.org/wiki/Zero_one_infinity_rule).
  
It would seem that we already have a 'zero' tail-emission coin: Bitcoin.  We also already have a '1' tail-emission coin: Monero.  (has a perpetual tail emission so that there will always be Monero to mine after the initial distribution, but since it's a fixed block reward that will decrease percentage-wise over time).  I have suggested we consider making Aeon the first coin to embody the 'infinity' principle by instituting inflationary tail emission.  
  
 
My core question to the economics section is this:  If there were blockchain designed to focus on spendability and consumer friendliness, it might make sense to create a perpetual inflationary emission after the initial distribution had ended.  What is the ideal inflation rate for a currency?  
  
The online research I have done has suggested that most economists agree it is likely somewhere between 0.7% and 2.4%.  I know we have some very bright minds here, so I want to know: what is your opinion on this? (if possible, please back this up with references)

I think that cannot be an exact and unique amount. Every country has its economic characteristics and economic development plans. The various specifics and factors of the above concepts can determine the need for different amounts of inflation. Then it is important and needed to know even the way used to determine the inflation. Are country who use the targeting inflation - which mean an fixed amount to be achieved. For example 4% within a year. Are other countries which use not the targeting but a valuation and maintaining of it within two values. For example between 2% and 4%.  So no possibility to find the best rate of inflation if are different ways to measure it. I can tell that in my country for at least the last 15 years the target of inflation was between 2% and 4%. Anyhow you can have full information about it (even more than needed according to me) here: https://en.wikipedia.org/wiki/Inflation

[BobK71]

There was some good discussion regarding debasement on the Economic Devastation thread:

https://bitcointalk.org/index.php?topic=355212.msg11923878#msg11923878

3. Due to that squared law explained in #2 and coupled with the fact that small things and investments grow faster, the upcoming wealthy gain more from debasement than the egregiously wealthy. Thus debasement is a very natural, efficient allocator of wealth to the maximum production. The super wealthy can't focus their investments to the most productive because their wealth is too large to allocate efficiently. They are more concerned about safety, economies-of-scale, and return of capital, which retards production in the economy. [implication is that the more wealthy someone is, then the more reliant on usury for return-on-investment instead of non-guaranteed return (a.k.a. return versus risk)]

Thus raising the debasement rate redistributes capital from lower efficiency investment (and thus production) to higher efficiency investment and production. Thus benefiting everyone in society.

Thus the debasement rate should be set as high as such that a super majority of society can receive an increase in their personal purchasing power (personal priorities) that exceeds the debasement rate.
...

This all sounds good, until you realize that the entire debasement regime requires humans to operate.  Since these political and financial elites then benefit hugely from issuing the assets that cause and enable the debasement, the system suffers from a bias toward inflation and instability, as shown by history.

[jaberwock]

Maybe something more intelligent and without fixed inflation rates, but that could be determined beforehand so no one would be caught off guard. The rate would have something to do with the number of coins possibly with private keys lost, the mining difficulty and the usage, among other factors.

But I don't know exactly how it would be implemented

[Febo]

Maybe something more intelligent and without fixed inflation rates, but that could be determined beforehand so no one would be caught off guard. The rate would have something to do with the number of coins possibly with private keys lost, the mining difficulty and the usage, among other factors.

But I don't know exactly how it would be implemented

Yes i also think that "lost coins" will have huge impact on  number of coins in circulation.  But question is, how can you determinate that coin is lost?  Even wallets that with today's technology cant be brute forced, will be someday in future. So i would use term "lost coins" really carefully.

[funkenstein]

The ideal inflation rate is one which increases the supply LOGarithmically.  In other words, the reward follows a harmonic series. 

A fixed rate of annual inflation, regardless of who gets the new coin, will debase the currency too quickly as people will eventually flee to other more stable coins. 

On the other side, with no inflation or a time period with too little, the miners will eventually flee and security will be reduced forcing large transactions to other chains. 

A logarithmic supply curve is the middle ground between these two concerns.  Every block is worth more than the one after it, so that a "early adopter" incentive always exists.  Meanwhile, there will always be a block reward -- despite the existence of a hard cap in total supply. 

[mcplums]

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin. It would never be in our individual interest to own a coin with inflation, therefore no one will, even if society at large would benefit it we did.

[funkenstein]

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin. It would never be in our individual interest to own a coin with inflation, therefore no one will, even if society at large would benefit it we did.

Bitcoin currently inflates at 25 new coins every 10 minutes, roughly at 10% annual inflation.  Dude there's a reason the chart is stickied at the top of the economics section. 

[americanpegasus]

Since I see some of you revived this topic, it bears mentioning that Aeon has decided on an inflationary rate of 8/9%, or 0.888~%.    
  
This barely debases the currency but always gives new Aeon to mine.

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin.

 
  
What will you do if miners abandon bitcoin because transaction fees aren't covering their costs and there's not enough significant bitcoin to mine anymore?  Then the network will collapse and the token you hold will rapidly become worthless.  Miners are the most important part of the equation because without miners, there is no currency.  Inflation gives miners something to mine, forever, at a very nominal cost to you. 

[mcplums]

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin. It would never be in our individual interest to own a coin with inflation, therefore no one will, even if society at large would benefit it we did.

Bitcoin currently inflates at 25 new coins every 10 minutes, roughly at 10% annual inflation.  Dude there's a reason the chart is stickied at the top of the economics section. 

That is correct. But in the long run inflation will go to zero- it won't just tend to zero, it will actually be exactly zero. I think that is what OP meant when he referred to zero inflation coins.

[GingerAle]

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin. It would never be in our individual interest to own a coin with inflation, therefore no one will, even if society at large would benefit it we did.

Perhaps, but I think there is a delicate balance to create between the interest of the individual and the interest of society, and I think this balance is ultimately what every economy is seeking. One cannot do this in a system in which the primary means of value transfer is of a static quantity, because there is little to no incentive for hodlers to relinquish themselves of these value transfer units, because if they continue to hodl them, they will go up in value. Sure, this is overly simplistic.

And this type of balance is difficult to achieve in a system where the primary means of value transfer is overdynamic - there is no incentive for an individual to hodl in the first place, hence they don't contribute to this society because the society's value transfer system is flawed and lacks value.

So ultimately the goal is to create a value transfer unit that encourages investing in the society in which the value transfer unit is used.

I think that, ultimately, the inflation rate in a currency should be so subtle and on such massive timescales that changes in the function of the currency as a value transfer unit are imperceptible. And I think another focus of the discussion should be the time window in which the inflation is executed. Probably a 10 year smoothing would be a good start.

And indeed, the internal data that such a system would need to use (money velocity) could be gamed, but if the smoothing window is large enough, the transaction fees to game the system will be exorbitant. And even so, if some Large Actor (such as the state) implemented a plan to game this system, what would be the end game? A redistribution of wealth due to the transaction fees going to the miners (if the system remains decentralized).

And ultimately these types of metrics needed rely on a singular blockchain (so sorry small blockists), unless meta-data is also transferred from sidechains

[mcplums]

Since I see some of you revived this topic, it bears mentioning that Aeon has decided on an inflationary rate of 8/9%, or 0.888~%.    
  
This barely debases the currency but always gives new Aeon to mine.

Although I think this is very interesting to think about, in practice I think it is a non starter.

Even if we all agree that a coin with non zero inflation is better for everyone, it benefits us as individuals to hold the coin with zero inflation- bitcoin.

 
  
What will you do if miners abandon bitcoin because transaction fees aren't covering their costs and there's not enough significant bitcoin to mine anymore?  Then the network will collapse and the token you hold will rapidly become worthless.  Miners are the most important part of the equation because without miners, there is no currency.  Inflation gives miners something to mine, forever, at a very nominal cost to you. 

You make a good point. However, this would make Aeon no more than a hedge- something to wait on the wings just on the off chance transaction fees are not sufficient to cover bitcoin's mining costs a long time in the future. Is this the role you envisage for this coin?

If yes, then how about you throw in something else which I have been worrying about: broken cryptography. If ECC was broken bitcoin's value would go to zero within hours. I've always thought an ideal coin would require multiple different and independent cryptographic methods, such that if one was suddenly broken the coin would still be secure, and the community would have time to replace the broken method with something new.

This is similar to how some patients are given three antibiotics at once just in case the bacteria becomes resistant to one, it doesn't matter because it has to become resistant to all three at the same time for it to be a problem.

[David Rabahy]

How long does it take for a currency to be worth ½ of the present value?  At 0.7% annual inflation it only takes ~100 years.  At 0.1% it is ~700 years.

How much are folks willing to lose value to inflation?  Soon enough there will be no inflation in Bitcoin then no one will lose value forever.

Why do we have to tolerate *any* inflation?  Why do we have to tolerate *any* debasement?

The sooner we all begin thinking long term the better.

[americanpegasus]

Why do we have to tolerate *any* debasement?

The sooner we all begin thinking long term the better.

 
 
Because the purpose of money is to act as a language and fluid through an economy.  Like blood being pumped through a body, money serves its purpose when it moves through civilization quickly and creates lots of transactions. 
 
Money sitting stagnant does nothing and helps no one.  What does the body do when it goes into shock?  The same thing an economy does when it goes into shock. 
 
I can tell you I have a Pegasus Coin worth a billion dollars right now, but unless others believe it to be true then it's worthless.  Why would anyone want a Pegasus coin?  Only because it is being exchanged all over the world for goods and services.  Volume and liquidity are more important than price, and if Bitcoin becomes a big stagnant pool of never moving money, then it will certainly not become the currency of the world. 
 
Mild debasement is good for civilization.  If you do nothing with your money for an entire human lifetime then it should lose roughly half its value (about 1%).  This value death allows civilization to renew and revitalize itself throughout the ages, or else we would all still be slaves to the man who first invented steel. 
 
You are in bitcoin to change the world for the better, right?  Or are you just here to get rich?

[David Rabahy]

I personally will spend some and save some.  Anyone that doesn't save some is taking a very big risk of not having some when they need or want it in the future.  Why should we tolerate having our savings erode over any period of time?  Why should I be obliged to keep my savings in an investment?  A bank savings account is investment with risks.  A bank CD is too.  A government issued bond or treasury note is too.

The portion I spend contributes to velocity and fluidity.  The part I save doesn't hurt either.

The mistake made is to think savings reduces fluidity.  The divisibility of Bitcoin avoids that.  The lack of useful divisibility of gold led to big problems and our response was to let the dollar float away.

I can chose to invest some of my savings but then it is put at risk.  Investments help projects and businesses fund their activities; in return I am paid, dividends, interest, and/or capital gains distributions.  Later I might exit that investment with a capital gain or loss.

I want Bitcoin to succeed.  I own some Bitcoin.  I promote Bitcoin to my friends and family.  I participate in this forum, contributing ideas.  I run a full node without direct compensation.  Geesh, what more do you want?  If I spend all of my Bitcoin then I will want more.