Topic: A Better Coin - page 7. (Read 12467 times)

Also, how would a demurraging currency account for growth in GDP?  Wouldn't price deflation hit if more people are after the same number of coins?

So the demurrage wouldn't necessarily be recorded in each block (else you'd have millions of transactions per block), but it could be easily calculated to show an accurate wallet balance at any given time.  Makes sense.

Lost coins would still take a long time to recover though... I mean, 0.97^X=0.000000004 where X = what?  634 years before a lost coin would completely recovered if rounded out to 8 decimals.  Still, it's better than not recovering them at all, and most of it would be recovered early on.

Sounds like a good idea.  I like it.  I think I like it better than my own idea, but will have to give it some more thought.  There's definitely some psychological battles to acceptance for such a currency...

Good point with the pools.

Demurrage destruction takes place in all outputs (accounts to simplify) whether the private key to access those funds is lost or not.
So the generation rate is constant [(target_base * demurrage) / 52560 = reward]. Say we have 3% annual demurrage then

(21000000 * 0.03) / 52560 = 11.9863014

The recipient can verify how much he received, but once the transaction is in the chain, he's taking the losses from the demurrage.
The payer must send more coins than he wants the recipient to receive, because in the time between he sends the tx and it gets into the block, demurrage also applies.
More technical details here:

https://bitcointalksearch.org/topic/73-btc-bounty-implement-demurrage-in-an-alternative-chain-with-merged-mining-36190
https://bitcointalksearch.org/topic/implementing-demurrage-how-many-functions-to-change-35705

Relying on outside metrics would become even more of a problem with the mining pools we have now.  If 3 of the major pools were convinced to alter the actual metrics, they could control the blockchain and force those metrics on everyone else.  There's the argument that people would switch pools, but not everyone would pay attention or care about it.  It would just get messy and cause a lot of headaches.

How would lost coins be destroyed with demurrage though?  Would the system calculate that any coins beyond a particular age (where age = the point at which the amount of demurrage would equal 100%) would be automatically destroyed by such a system?  Wouldn't that lead to a particularly unstable reward rate?  Some points, you might suddenly get 9000 coins because someone lost them a long time ago.  It'd be like a lottery system instead of a steady payout a miner could count on...

Also, how would a merchant know what the actual payment is, if the demurrage is coming out of the payment?  What if the buyer uses coins that would be 95% demurrage?  Or would the buyer have to send 2000% as many coins in order to pay said merchant?

It's an interesting idea...

The metric should be inside the system from external exchange prices.
Miners are supposed to reject blocks with wrong reports about external exchange prices, but yes, those are my fears about stable coins.
Probably more orphan blocks would be a problem for that system too.

That's why I stick with freicoin for now. The demurrage rate can be set exactly and will eventually be equal to the reward, letting you have a stable monetary base. You don't have to worry about lost coins because they will get destroyed.
Price deflation concerns are diminished too, because demurrage (unlike inflation) lowers interest rates (by discouraging hoarding while not inflating other capitals value).

0.285% was the number of coins lost in the last year (around 20k) vs the number of coins in existence (around 7M).  It's not an exact number, and it would be difficult to say that it would be an accurate extrapolation to be distributed across a large number of people, but at the moment, it's the best number I have to work with.

Again, with trying to tie the number of coins to a CPI, or similar metric, you'll have the problem of people attempting to manipulate said metric to their own favor.  If it's going to be tied to anything, it must be indisputable, easily measured, and automatically calculated.

@SgtSpike
I guess 0.285% (no matter the size of the economy) should be enough. Where did you get that number from?


%ΔV is highly unpredictable, but it would be closer to zero (more stable) if the currency had demurrage (also V would be higher).
But then, you still have to set %ΔM as a function of %ΔQ (or %ΔP as a measure).
There's a proposal named stablecoin for that. I think it could be better implemented with demurrage but I still have many doubts about if the potential problems for the network derived from the %ΔP input.
Since you're creating coins (and destroying) even when ΔP = 0, you also solve the "reward miners forever" problem with that solution.
In that thread most people prefer destructive tx fees instead demurrage, but that makes V even more unstable.

I thought I explained it sufficiently in the OP, but I'll re-explain it here.

If I expect all of the 300M people in the US to adopt "A Better Coin," and use it in place of the dollar, then the goal would be to have an available supply of $14T, which is roughly the current M3 supply, and have each coin be worth what $1 is worth today.  With regards to digital currencies in the past (AKA, Bitcoin), the loss rate has been around 0.285%/year (18K lost in a year / 7M).  0.285% of $14T = $39.9B.  If the block reward is equal to the number of lost coins, then $39.9B/year is a heck of a lot of money for miners to use to secure the network.

I agree it is a concern, but I also believe I have already addressed said concern with my initial proposal.  That said, this is exactly one of the reasons that I believe Bitcoin will fail - it will not have sufficient protection against double-spend attackers once the block reward is effectively eliminated.  If the transaction fees remain below 1 BTC/block, then the mining capacity will be reduced to 1/50th of what it is today, unless the value goes up.  But if the value goes up, then so does the incentive to initiate an attack in equal proportions.  There would be no way to "get ahead" of an attacker unless transaction fees drastically increased.

Sorry, I didn't read all the threads.


Sorry again. What I said is "If coin losses can sustain mining, then you've solved that problem."
That is, I don't know what annual losses do you expect, but if it's say 0.1% of M then maybe that's not enough to sustain miners.

I believe you're basically saying what JohnDoe explained to me, in that there is more to deflation or inflation of prices besides just the amount of money available.  I don't believe that absolute stability could (or would) be achieved.  But the goal would be to reach an equilibrium where, on average, the number of lost coins would equal the number of coins rewarded.  Of course there might be some mild deflation one year, and mild inflation the next, but in the < 1% range to where it *should* be inconsequential.  If a better plan to achieve absolute stability is presented, then I would support it accordingly.

As JohnDoe pointed out, GDP growth must be accounted for in order to achieve stability as well.  Can GDP grow indefinitely?  I doubt it... but some equation must be used to at least attempt to accurately predict average GDP growth in the coming years.  In this way, my initial plan as laid out in the OP is already outdated, as it does not account for GDP growth.

Also, you're confusing me by starting a paragraph with "You don't solve this problem neither," then ending it with "then you've solved that problem."  You said my plan wouldn't work, outlined the plan, then said it would work.  Make up your mind! 

I think it's important to think about the currency AFTER it has been fully adopted (to whatever point full adoption will come to).  You can't count on deflation happening when people flock to the currency, and have a high inflation rate to attempt to counter it, because then you will still have a high inflation rate after the increased adoption level has ceased.  However, I still agree that, at the very least, GDP should be estimated and accounted for in the equation.

I believe we can agree that currently reported GDP numbers for countries worldwide are generally accurate and indisputed.  If that is the case, could we extrapolate it into the future, and come up with an equation of growth in block reward based on historical worldwide average GDP growth?  Regardless of what the end equation is, it shouldn't be adjusted based on GDP growth numbers in the future (since those numbers could potentially be manipulated).  It would need to be a set-it-in-stone-forever equation of growth, even if it doesn't match up with the real-world number perfectly.

I don't think it is fair to look at a specific currency in the world today and assume growth based on it - we should be looking at economic growth as a whole, and assume growth based on that.

Linking %ΔM to difficulty doesn't do anything to change the fact that we are unable to predict %ΔQ and %ΔV far in the future. The only thing you are doing with this is making %ΔM unpredictable as well. Maybe I'm not seeing it so could you provide me with a formula for nSubsidy to see what you mean?


I acknowledge that we can't predict economic growth, it's a weakness of having a decentralized monetary policy, but I can't see how it can be improved upon without adding centralization. You are wrong about not having a mechanism to combat volatility though. Since %ΔM and %ΔQ can be estimated with fairly good accuracy in the short to medium range then the unknown variable %ΔV can act as a buffer when inflation is expected, as people move their cash on hand to financial assets to avoid the loss in purchasing power, thus decreasing velocity.


Yeah, my 8~10% %ΔM range is arbitrary and completely up for debate. I want to achieve as close to 0% inflation as possible but preferring mild inflation to mild deflation so I based it on emerging markets growth rate rather than the world's average. Also I considered the M2 growth of the Swiss Franc, which was 7.8% from March '10 to March '11 but still managed to gain purchasing power compared with other currencies which had lower growth, meaning that even at 10% annual expansion it's still very possible that we could experience deflation if people flock to it because of the stable monetary policy.

You don't solve the deflation problem. If you mean monetary deflation caused by lost of coins, say 0.1% of the supply is lost annually, then eventually 500,000 coins per block will be that annual 0.1% of the total money supply, it's ok.
But what matters for the financial market is only price deflation. That can be obtained with monetary stability, monetary inflation or with monetary deflation. You're only achieving fixed monetary base and only because you assume that the lost coins are somehow a constant.


You don't solve this problem neither. After reaching 500k coins, the reward for miners is (proportionally to the monetary base) lower and lower until it is equal to the lost coins. Only then the proportional reward becomes constant. If coin losses can sustain mining, then you've solved that problem.

Freicoin, on the other hand, reaches a fixed monetary base, reduces the undesired effects of price deflation and rewards miners forever.
More at first, but when the maximum freicoins are issued, they are rewarded forever with the same proportion of the total supply (the chosen demurrage rate).

Excellent post there. However, I have to say this would not be the arch nemesis of Bitcoin. This project would be par for the course for bitcoin....and that is people reinventing the wheel. There is a scene in the Matrix, where a character says 'You already know where this road goes'...and I hope people can see this same symptom in this community as well. You'll end up with the US government and the Federal Reserve....which by the way, not bashing the posters here, I consider a lot more savy. If thats the case, the point of bitcoin is to bide time and not be a permanent solution....get people past the supposed gov/dollar collapse and onto a new dollar and change in politics.

Im cynical on bitcoin's ability to accomplish this, because for one, bitcoin is moving contra gold.  That tells me that people are just chasing yield, which means its an improving economy bet, which certainly is not a bearish strategy. Its not tracking 'safe haven' allocations...at least while Ive been watching it.

As far as a better coin, I imagine it exists, but dont know what it is. Bitcoin took me by surprise. I do know it wont work if people then try to use it in a way it wasnt designed to be used. And then blame Satoshi(aka the devil) , or whoever, or whatever. And it will be an innovation, not reinventing the wheel. I do suspect it will involve a confidential social network.

JohnDoe, thanks for the explanation - that makes a good deal of sense.  I suppose the biggest concern is how would one accurately predict the growth of GDP?  Or are you just aiming for potential growth in GDP based on historical data in order to get as close to effective 0% inflation as possible?  Has the economy really grown 10% every year?  And, likewise, has the money supply increased by (roughly) 13% every year, to meet the fed's supposed goal of 3% inflation?

I think this is an absurd assumption.  No one is going to increase poor investments just because we move to a deflationary currency (thus making those poor investments even poorer).  Anyone can do any of those "community beneficial" investments they want today with an inflationary currency already.  A deflationary currency would discourage those types of investments just the same as they would discourage investments for monetary gains.  I mean, why would I build a solar farm with 10 panels today when I could build a solar farm with 12 panels tomorrow?

Linking the money supply to difficulty accomplishes what, exactly?  Why not just make it a flat-percentage increase if you're trying to keep it at a constant rate?

Link %ΔM to difficulty. Computing speed changes are orders of magnitude more predictable than growth in BTC-denominated trade. If you don't believe me, then make a bet. I will bet on the difficulty-price ratio x months or years from now. You will bet on the USD-BTC price ratio x months or years from now. Whoever comes closer in % terms wins. Would you really want to take that bet??? Would anyone??? If you would, raise your keyboard and be counted!!!
 
Arbitrarily specifying a %ΔM as 10% or 8% or 0% or -5% is completely useless. There is no way you can predict the growth rate of the BTC economy a priori. You need blockchain feedback in the system, otherwise the modification will have no affect on price volatility whatsoever. Completely pointless.

This is exactly how I rationalized it before. The problem is that in investment nothing is certain, there are way too many unknown variables to be able to say for sure that the investment will pay off.

For example, if you need a stapler for the office and it costs 1 BTC, how could you be able to asses that the stapler will produce 1 BTC or more in value? Or if you discover a molecule that you think could be the next aspirin and you need research money, how will you convince your creditors that it will be a success, that everything will go smoothly on the 5 to 10 years of research and that you won't go overbudget? Or if you need 1000 BTC to pay for college, even if you are smart as hell and you are certain to land a high paying job after, how could you justify that loan if the purchasing power of BTC is increasing at an astonishing rate and by the time you graduate those 1000+ BTC you owe will pay for a small country? You would end up being richer if you just started working as a garbage man for BTC right after highschool.

Things get further complicated in that debtors would need very low interest rates to make the venture worth it but creditors need to ask for high interest rates to make lending worth it.

Ok let's discuss what we believe is the cause of (price) inflation. It seems you have the Rothbardian view that inflation is caused simply by an increase in the money supply. I'm part of the Mises branch which asserts that inflation only results when the supply of money outpaces demand for money:

    "In theoretical investigation there is only one meaning that can rationally be attached to the expression Inflation: an increase in the quantity of money (in the broader sense of the term, so as to include fiduciary media as well), that is not offset by a corresponding increase in the need for money (again in the broader sense of the term), so that a fall in the objective exchange-value of money must occur."

The correlation between inflation and the supply and demand of money can be seen with the equation of exchange:

MV = PQ

where, for a given period,

    M is the total amount of money in circulation on average in an economy.
    V is the velocity of money, the number of times per period each unit of money is spent.
    P is the average price of all the goods and services sold during the period.
    Q is the quantity of assets, goods and services sold during the period.

In terms of percentage changes (to a close approximation under small growth rates, the percentage change in a product, say XY, is equal to the sum of the percentage changes %ΔX + %ΔY). So:

    %ΔM + %ΔV = %ΔP + %ΔQ

which can be rearranged to the following to obtain inflation:

%ΔP = %ΔM + %ΔV – %ΔQ

where, for a given period,

    %ΔP is inflation. Negative percentage means deflation.
    %ΔM is the money supply growth rate.
    %ΔV is the change in velocity.
    %ΔQ is the GDP growth rate.

With your model %ΔM will eventually become 0% on average. Let's say that the GDP growth of your economy in a given year was 10%. The equation then becomes:

%ΔP = 0% + %ΔV – 10%

Then the only way to maintain the purchasing power of the currency is for the velocity to increase by 10%. That's doable for some time but obviously unsustainable as the velocity can't be increased forever. Not to mention that deflation would encourage a reduction in the velocity instead as people shift towards saving.

My proposal is that price stability would be more likely if we try to fix %ΔM to something like 8%~10%. Of course, changes in velocity and GDP growth can't be control or predicted but I believe this model would be more suitable to achieve the behavior that we want.

(Monetary) Inflation and demurrage are not equivalent.  Inflation creates liabilities for the central bank.  With demurrage the currency can remain fully backed, and avoid any risk of a need to suspend redeemability.

I didn't say it discourages a run to the exit, and I also wouldn't want to hold onto these coins as long as hypothetical steadily deflating bitcoins are available as an alternative.  But if my above description of lending deflationary currencies is wrong, then the market might end up requiring such a currency to be available for the purposes of lending.

This does bring up the point that the inflating currency would require redemption fees to deter people from just cashing in to bitcoins as a safe store of value, and making the lending-currency effectively useless.

I agree, it's an ugly solution (to a problem I doubt is even real).  But it's the best that I can imagine.

Why would demurrage discourage a run for the exit? Do I care if prices increase by 10% and my coin balance remains fixed (inflation) or if prices remain fixed while my coin balances decreases by 10%?

They are equivalent scenarios. If I expect either of scenario to happen, I should sell my coins for USD.

There could be a (irrational) psychological effect, but such effects are notoriously difficult predict and sensitive to small changes in context.   

+1 to everything d'aniel said - extremely well put and exactly my views!

Also I'd like to add that a mildly deflationary currency while discouraging lending for pure monetary gain would encourage people to for example do ethical investments or other investments which have additional non-monetary gain for them or for society.

I might for example very well invest in the building of a wind or solar farm, even if the expected gain is below the inflation rate or below 0 in a deflationary currency - just because I gain from it indirectly in non-monetary ways.

Another example would be investing in the building of local businesses because I might think that my neighborhood profits from having more local stores or job opportunities.

I don't condemn investing out of pure monetary interests but to say that an economy could not work if the focus would be shifted a bit towards investments which have additional gain is a bit extreme.