Nice chart amincd.  Good observations all round.

Gold has many good properties for a currency - it is scarce, it is a bearer currency, anonymous, relatively easily authenticated, impossible to counterfeit, reasonably easily divisible.  It's value-to-weight ratio is pretty good, however storage, transport and security are a problem especially for large amounts over long distance.  Bitcoin equals or far exceeds gold on all counts, so rather than just "egold" it should probably be "e super-gold" or something. 

The reason behind the Au/Ag idea was that because bitcoin is so good in these categories, demand to hold it as money is extremely high.  Bimetallic currencies were used extensively over the ages, it is easy to read up on how they worked.  I do not know why people started using Ag in history since in theory the Au price can go as high and you can just divide it down.  Maybe it was simply a physical issue, Au coins became too small or you had to mix them with too many other metals and they became difficult to validate the gold content easily.  If the only reason was divisibility, then in theory if someone starts a 210M cap chain with a 500 BTC block nobody is going to be interested in it, they'll want only the scarcest stuff.  If it is a more fundamental issue of demand for more stable pricing, then Ag will flourish.  However, it will not make the AuBTC go away - of that I have confidence.

I think Au/Ag is worth exploring, however at the moment I am more intrigued by the ECC idea (1kWh = 1 BTC) in the other thread.  Conceivably both an Ag and ECC chain could be started, and let the market decide.  If both fall flat, then that may tend to prove the AuBTC to be the best design and will probably dampen all the arguments about deflation.  Competition makes the survivor stronger.

I suspect that regardless of all our opinions some other chains will be attempted as the market cap of all BTC continues its skyward trend.

I have reread your OP.  I better understand that you are also proposing to remove the difficulty adjustment.  I missed it the first time, I was overwhelmed by the circular nature of electricity2bitcoins that could (eventually) be used to buy the same amount of electricity.  It does take some time to grok.

To summarize, the proposal as I understand it is to start a new chain called ECC that operates the same as bitcoin except it removes the periodic adjustments of difficulty.  The difficulty would be set to a fixed value designed to attract enough miners to make the system secure.   When the system is fully operational with a reasonable supply of ECC, it would be self-regulating.  When the USD/ECC price rises above USD/kWh mining will be profitable increasing supply and driving down USD/ECC.  When it falls below that level, mining would stop and natural deflation would drive up USD/ECC.  Over time ECC would track kWh approximately.

Please correct this summary if I have it wrong.

Now some problems I see with this, and some solutions I will throw out there for comment:

(a)   There is no start point that has positive feedback when there are zero ECC and therefore no price to deflate.   Solution: I think you need to start by setting the difficulty to some easier value, say half a kWh produces 1 ECC for some period, maybe the first year.  Or to a certain number of ECC, say 1M.  Then you build in a schedule of raising the difficulty until it terminates at parity, 1 ECC for 1 kWh.  This incentivizes early miners.  The community will enforce the difficulty rise since it is in everyone’s mutual interest.
(b)   How do you calculate the amount of energy to solve a block?  As pointed out this will reduce over time with better mining technology.  Conceptually inflation will follow the mining technology curve, and will be steeper at first.   However, demand for the currency itself will counteract that if it is successful.  Solution: One idea would be to mutually agree on a standard reference rig – designate a particular algorithm, GPU model, main board, DRAM etc.  We would measure the kWh to solve a block and hard code that as a starting point.  It doesn’t matter that over time that rig will be obsolete, it just means at that point in time it was a reference value.   I can see how (b) might not be necessary, but I think doing this helps to make the value of ECC easy to understand for the casual user.

Again, kudos to the OP for a very excellent idea.

Keynes was not a monetarist.  He advocated making up for a decline in aggregate demand by public spending. If he was a moneterist he would have advocated taking steps to increase the money supply which presumably would increase private spending.  While I think the theories of Keynes, Hayek and the Austrians, Friedman and the rest all have their interesting points, I am personally skeptical that any of them have been proven to explain inflation, deflation, and the causes of recessions, and how to fix them or even if they need to be fixed.  Economics is a social science not a physical science and it is difficult to prove anything in Economics since you rarely have an adequate control group.  I was merely stating what conventional monetary theory predicts (which is just a theory) and that you are seeing an increasing 2nd derivative on MtGox.  The prices on MtGox are really the only facts we have.

Anti-Keynesians usually embrace some kind of free market or libertarian philosophy.  If that is the case, you will have no objection if SGallaecian starts a new chain named ECC that tracks kWh and leaves bitcoin alone.  Then the free market will have two currencies from which to choose and presumably more competition makes everyone better.  If people prefer deflation with a fixed supply nobody will use ECC, and if they prefer an elastic supply that generally tracks a unit of value with moderate inflation or deflation then they will switch. Or maybe they will use both, each for different purposes.  Let the market decide!

Still the same argument as getting hungry.  If they want a TV they have to spend fiat dollars, because that is the only way to get a TV whether or not it gets cheaper. But in this case they have a choice of two currencies so it is a different dynamic.  It is a choice whether to spend A or B and not a choice to spend or not spend.

But your buy and upward price action might attract more ask depth.  It depends how fast the market reacts.  A better experiment is to execute part of your order at the ask qty 170 and put some or all of the rest of it on the book and see if that attracts execution.   I get your point that trading is thin but I'd like to see some of the tape.  I'm sure there are examples of both execution strategies in the tape somewhere.  http://www.mtgox.com/trade/history last 47 hours doesn't work for me today either.  If you have some sections of the tape that illustrate please copy-paste the in.

Exchanges typically solve this by requiring market makers to post bids and asks of a minimum size always, and they have a minimum number of market makers (at least 2 or more). Let's say it is minimum qty 100 and 2 makers, that means there is always qty 200 ask and 200 bid at any time.  So you can execute that and obviously they will raise and widen spread but they compete for the next execution (as you work through the remaining 800).   In order to comply they either have to have short facility or maintain enough inventory on both sides of the trade as a % of daily volume to allow enough time to replenish inventory if one side gets drained.  Which is exactly what you want. The other solution is to go to an auction market.

That is the standard argument that in a deflationary environment after some time people will get tired of deferring consumption and will start to consume and that will end the deflation of its own accord.  But those arguments were made historically for closed economies or economies where external trade was minor compared to GDP (say <10%).  What we are talking about is the exchange rate between bitcoin and a much larger fiat economy.  That is why I said "They would rather hold them since they are rising and spend fiat currency which is falling" .  They do have a choice, they can use fiat currency to eat for now.  Human nature is they will hold BTC while it is rising, and as soon as it start to fall they will try to buy everything in sight with BTC, it will fall faster and then merchants won't want to take it anyway.  We don't have an experiment of a diversified mostly closed economy that can supply most its needs internally that runs 100% on BTC.   If we did we should see inflation of 43% - grwoth rate.

But then consumers won't want to spend bitcoins.  They would rather hold them since they are rising and spend fiat currency which is falling. Merchants won't bother to offer BTC prices since it will not get enough use.  A situation where the price of something (BTC) is far in excess of its actual utility happens all the time in markets, but then you get the inevitable downdraft.  I know it is very powerful to believe in the "always upwards if we leave it alone" idea but we need the wisdom to see both sides of it.

Just a thought: What if bitcoin replaced all the money in the world, what would its present value be in USD?  The velocity of money in the US historically has been approximately 2, meaning there is half the amount of money as GDP (the money turns over on average twice a year).  I don't know what it is in other countries, but just guessing that is reasonable worldwide, and the world economy is approaching $60T, I get ($60T/2)/21M ~= $1.43 Million per BTC present value.  That is the theoretical end value of the deflation once it hits 21M cap and replaces all the fiat currency.  After that monetary theory says it will deflate or inflate at the GDP growth rate if velocity does not change.

I know that sounds rather attractive, but it won't happen.  It will unravel before then, who knows next month or next year.

I haven't been able to get on mtgox last night, it hangs.

Do you have some section of tape showing trade history where small trades are at huge spreads and wide volatility?  Do they tape their whole order book, or just time/price/quantity?

While margin and short selling should add liquidity, it is controversial whether it decreases or increases volatility.  It would have to be tried. Narrow spreads that you talk of seems counter to a thin market.  Options require active committed market makers, if we don't have a strong market in the underlying not sure that will help.  But mtgox had an "options coming soon" note on their site.

Could you just remove the 2-week adustment of difficulty?  Rather, you adjust the difficulty periodically so that it always consumes about 1 kWh to generate 1 BTC, approximately.  Does that break the security?

This may be a way out of the one way problem.  If the value of BTC declines below 1 kWh, people will turn off their mining rigs.  Natural deflation will set in raising the ratio back to 1:1. If BTC gets abouve 1 kWh, now it is profitable to mine and BTC will gush out until it returns back to 1:1.  The feedback may be a little unstable someone needs to model it in matlab :-(

Possible (theoretically) or totally wrong??

Agreed, varying the block reward would be a problem.

But what intrigues me about the OP idea is this:  For any arbitrage-based index tracking to work, you need an exchange mechanism that can close the loop - think of it as the feedback loop on an op-amp.  Normally that would be very tough to think up for an e-currency.  But ironically, electricity can exchange into BTC. That is what intrigues me.

It is hard to wrap my mind around it.  You consume electricity to solve a block to create BTC. BTC then is exchanged for USD which is exchanged for electricity.  If you are trying to counterract the USD being exchanged for (buying) BTC, we seem to have the reverse path. But it is weird because the electricity is just "thrown away" as heat, after completing the computation.   Typically in arbitrage you either grow or shrink your inventory of the underlying to close the arbitrage loop.  This seems one way (thank the 2nd law of thermodynamics I guess).

I'm still intrigued.  Someone smarter than me please crack this problem!

Do you think credit default swap derivatives have an "if people burn money for heat" clause as a trigger for a default event?  It would be fun to watch it all unwind.

This is the paradox.  The rising price is gathering attention to bitcoin.  This is good. I've argued in several posts that due to any random shock to the system, prices my fall (even if temporary). While things could work out well if prices rise steadily for a long time, and that at some point a real merchant economy emerges as the BTC total quantity tails off.  That is the ideal scenario.  But I also can see that any bump in the road could cause a drop in prices, followed by a move away from BTC.  At that point merchants may decide this was just a fad and the whole project will fizzle out.  It will be very difficult to restart.

I hope for the first outcome, but prefer to plan for the second since I've not encountered many smooth roads in life.  That is why I argue for considering ways to modify the system to encourage merchant participation soon.  I'm not the only one, I see many people exploring and considering different ideas along these lines, stemming from the same concern.

The reason for the OP is not to "back" the currency for its own sake.  He is trying to address the problem by allowing the bitcoin supply to be more elastic which is precisely what is called for by conventional monetary theory.  The current system has an ever decreasing rate of increase in the supply of BTC.   The first time I heard of bitcoin I immediately thought "monetary theory would predict an exponentially increasing rate of deflation" and lo and behold that is exactly what you see on MtGox. 

You're right it is very difficult to think of a good way to add supply elasticity P2P.  But the idea isn't bad, it might take some refinement.

>Aren't the price of precious metals and energy also fully market driven?

Yes but in an importantly different way.  Take precious metals.  There are multiple exchange points: Labor and capital in the form of mining equipment are exchanged for metal (extracted from the ground), which is later exchanged for dollars, which establishes the price.  There is an indirect link between labor and capital and the metal price.  This tends to create a loose link between labor prices and the gold price which over long periods of time (50-100 years) tends to have a stabilizing effect.  Since the labor component of BTC mining is negligable, there is no such linkage for bitcoin.

It is these exchange points in and out of BTC that I was exploring in my first response. If you desire to stabilize prices some clever mechanism has to be found.  I am interested in this problem as an intellectual exercise, whether or not people think price stability is desirable and whether or not it will be adopted.

I'm half kidding.  Miners are too smart for that, but it's worth thinking about:

In order to "replace" all the paper currency, it has to be traded ie. exchanged.  You have to have a willing recipient. Who would that be?  Yes there will be some stupid diehards but at some critical mass everyone will run for the exit .... together.

Maybe I already answered my own question, once the value of BTC rises to the point where it can buy a significant portion of the goods and services offered in a given year, then by definition you have inflation on the paper because it is the same amount of paper chasing an ever shrinking pool of goods and services (the leftover that weren't bought by BTC).  At some point the price of the paper falls below its thermal value and people literally burn it.  Maybe they run a generator to .... mine more BTC.

I feel like I'm trapped in an Asimov novel.....

Sure, agreed. But it is a small market, compared to the USD/EUR/JPY liquidity out there, the BTC price would have to rise 100s or 1000s of times for the market cap to get big enough to give it enough scale to have even a slim chance at maturity.

Interesting brain teaser in the OP.  I like these.

Cryonic suspension at warp speed aside, you would not actually replace the paper.  If you exchange $4.5T for 21M BTC, the miners end up holding $4.5T of paper currency after unloading their BTC.  That's a pretty large briefcase.   

As you point out by that time the paper is pretty much worthless, except to burn as heat.  So ironically the miners are happy just to get their kWh back.  The buyers of BTC don't make out, they just get a 1:1 exchange in buying power.  Their only win is they remove the threat of inflation.

In reality there is rougly 10x that amount of money, the rest being in the form of demand deposits or equivalent, which in theory is exchangeble for paper currency in both directions 1:1 (at least in small quantities).   What BTC will do to that, who knows.

Humm, head hurting already.

I explained in depth how commercial trade usage would create sell side pressure in other posts that I thought you were monitoring. I can dig back and repost.

Do you have tape showing the price movement on 20k BTC trades you mentioned? Can you post some piece of it?  I have looked at mtgox before but assumed that is only a fraction of the market since the depth is rediculously small.  You can't tell the true depth anyway because obviously large liquidity is not going to be posted in the order book.  I also assumed that there are other markets and the largest volume is OTC.  Bitcoin encourages OTC by its very nature, more so than most other trade I can think of. So some ticker examples of lack of elasticity would support the point since it should draw liquidity away from OTC and other venues if only temporarily.

Does mtgox publish its complete historical tape?

I can think of all the usual methods for promoting liquidity but they are controversial in the real (old?) world as well. Such as steps to encourage more market makers, rules for minimum liquidity, facilities for margin and short selling, rules for pricer discovery, etc etc etc I'm sure you're familiar with the litany.  Need to think about how to do some of that p2p.

I didn't fixate on the 1% inflation.  I took the OP as a general question of should we have a price stability target and how to accomplish that.  Both are big questions, the second part is really vexing. My intuition says a better target is very mild deflation, say maybe 1-3% per year.  I am on the same page of you that a stable medium of exchange would encourage commerce.  But I have taken some flak on that point from some who think it is heresy to talk about adjusting the money supply. I don't know how else to accomplish it.  I do know that current system has proven extremely volatile.

Since I'm intrigued by the AuBTC/AgBTC idea, I will be your first reply to your first post.

It is an interesting idea.  The question, as with Au/Ag is how to determine the exchange rate?  I gather you are proposing to tune the mining rate to approximate consuming 1 kWh to generate 1 ECC.  As per a number of previous posts on Au/Ag you are proposing using market arbitrage to keep the ECC/kWh exchange rate close to 1 (similar in general concept to how Index ETFs work today).

The issue I see is the same for ETFs:  You need a fundamental conversion back into the fundamental (kWh in this case). ETFs accomplish this by allowing professionals to deliver or receive the underlying stocks into/from the fund in exchange for shares in the fund.  The issue I can see with not having such a 1:1 exchange mechanism is it could fall into a lack-of-demand trap.  In the beginning while ECC are scarce the USD/ECC price could get bid up.  When it exceeds USD/kWh miners get busy.  I'm with you so far.   But what if they over-mine and rush to sell, and then some false rumor spreads and temporarily drives the USD/ECC price to 1/2 of USD/kWh. Sure all mining stops but you still have an overhang of sell demand.  The price could languish there for a very long time.  It's more likely the bigger the ECC supply gets and you need significant demand depth to pull it up.

But it's a good idea, maybe you need some elastic RC time constant type of mechanism that somehow buys up ECC if that happens (or incentivizes the same).  Can't think of one right now....

Nice thoughtful post, Tsudico. As Vladimir points out there won't be any home loan because the lender can just hold onto BTC at zero risk.  There always has to be some interest to account for repayment risk and profit.  Say the interest is 5%, then with 10% deflation that makes an effective 15% interest rate - expensive for the borrower and all the more reason to pay off the loan quickly.

That is the conventional wisdom on deflation.  It makes borrowing very expensive, and discourages lending as well.  Another risk for the borrower: What if deflation accelerates half way through the loan?  Then you are faced with a rising effective interest rate and ever-appreciating principal to repay.  In an inflationary environment generally borrowers benefit from some inflation, or worst case it is neutral.  Lenders bear most of the inflationary risk. This additional risk for the borrower in deflation is another reason not to borrow, or to pay it off very quickly.

While you may correctly conclude "don't borrow" in fact there are short term "effective debts" incurred all the time in any economy.  You get billed at the end of the month, paid at the end of the month, things take time to ship, to store in inventory, to get consumed.  Inventories need to be sold quickly or you could lose money.  You can come up with hundreds of examples.

Things to ponder.