Bitcoin Forum

We all know here that there's a maximum number of bitcoin which is 21 million.

The answer to "what if we need more" is : we can always divide bitcoin.

Problem: bitcoin cannot be divided infinitely. It means that there's an absolut limit on the possible number of bitcoin. This number might look huge.

But remember: 640k was enough memory for anyone.

So, if it happens that we reach the maximum possible division of bitcoin, what will happen? Economicaly speaking and technically? What will we do?

Yes it can.

Store amounts with more precise representations.

Who needs an infinite number of bitcoins anyway?  There are a finite number of potential bitcoin users.  An infinite currency is a worthless currency.

Bitcoin can be updated to allow for higher divisibility than it currently supports if deemed necessary (far in the future if ever), and is currently divisible more than anyone currently (or for the foreseeable future) needs in any practical sense.

Neither can gold. You can divide it down to the atomic level. To go past that you need fission, but you'll end up with whole isotopes of yttrium and zirconium (or something) instead of half atoms of gold.

Perhaps we'll need a bitcoin fission reactor. Feed it a bitcoin, get at least two betacoins.

Get a second "branch" (new genesis block) of bitcoins. EUcoins, USAcoins, Chinacoins, etc. ;)

20,999,999.9769 BTC actually.

There are an infinite number of potential bitcoin users, until the end of the world. Currently, the world population is a mere 6.7 billion, and there is more than enough land in Texas for every last one of us to have our own house. If BitCoin succeeds long-term, we will need much more base units. But that's so far off, it's not worth bothering with now.

What for ?

I will stay with the main branch, thank you.

2.1 quadrillion basic units should be plenty.  The US M2 money supply is (according to Wikipedia) less than 1 quadrillion pennies.

If bitcoin is ever twice as popular as dollars, that would be a very good problem to have.

Thanks for confirming my understanding.

Now, I'm playing devil's advocate here: why would I invest in a money that has a physical limit like this? See the year 2000 bug and the IPv4 story: the problem seems far far far away and, one day, we wake up and the problem is there.


Now, one possible solution I foresee would be to run a second bitcoin network. The clients would operate on both networks in a way transparent to the user. If you have 10BTC, you could, in fact, have 8 on the old network and 2 on the new one. The "only" problem would be that BTC addresses would have to differentiate from one network to another. We could imagine adding "@2" at the end of the addresses. An BTC address without "@2" (or with "@1") would be considered as part of the first network, which allows backward compatibility.

This is like saying the quantity of dollars is limited by the amount of paper in the world. 2.1 quadrillion is a huge number, the planet isn't getting any bigger and the worlds population is projected to peak at 9 billion people. We have bigger problems ;)

Haha yea that's right, the year 2000 bug totally destroyed the internet, so will the ip4 shortage.

This would be reasonable if the problem was like "It breaks completely and there are no possible solutions at any cost", but it isn't that bad at all. There are many solutions that won't be very hard to implement. And so what if it cost $1B to fix the thing, that's like 1BTC by that point. 

Second network... ? Alternative chain...?

The Bitcoin is divisible enough, and protocol supports even more divisibility, so what exactly are we discussing ? I thought this matter is already resolved.

According to my understanding of bitcoin and to gavin, you cannot infinitely divide a bitcoin. Thus, the maximum of possible BTC units is 2.1 quadrillion.

I'm thus trying to think about what we could do if we want more units. (currently, it's a pure intellectual challenge for the sake of it).

Well, i think i remember Gavin (or maybe it was somebody else ?) saying some time ago, that this is not a hard limit, and the protocol/client can be upgraded in future versions to support even 128 digits of precision.

Isn't this right, Gavin ?

There are a number of possible solutions, if more precision is ever required. One of them involves using the existing "sign bit" to represent a precision shift of the remaining bits of the 64-bit amount.

But really there's no point planning for this now. The world will have changed in unimaginable ways if ever Bitcoin gets to the point of needing more precision.

+ 1

I should print this out and put it on the wall over my desk.

On Security Now, A listener asked a very intelligent question, which fits into this discussion quite well:

With real (minted) currency, a hole in my pocket results in change dropped onto the ground, lost.

With Bitcoins, a hard-drive crash or un-backed-up wallet, results in bitcoins lost into the ether, gone.

The difference is, with minted currency, someone can come along and pick up that change. If my 50 BTC wallet is gone forever, so too is that 50 BTC. Now, obviously, that reinforces the necessity of backing up your wallet after every transaction (Which I do), but someone, somewhere, isn't going to. This will result in deflation, slowly, over time.

Is there a solution to this?

A solution to what?  Don't lose your bitcoins.

Problem is, If I lose bitcoins, the market loses bitcoins. The solution I'm looking for is to the inevitable deflation. If Bitcoins are to be a stable currency, it must be, you know... stable.

You can make as many copies of your wallet.dat as you wish. You cannot make any copy of your real coins that you can lose due to having a hole in your packet. This limits the probability of losing bitcoins greatly, unless someone is stupid enough not to do backups.

That proposition is correct, but...

Bitcoin is not a stable currency. It is better than stable. It will inevitably deflate, to the benefit of everyone who doesn't lose their wallet.

True, but if life has taught me anything, it's that if something can happen through human stupidity, it will. Probably multiple times.


You have a point, but deflation is problematic in an entirely different way than inflation. Sure, it's great that you could buy a house with only 5 btc, but no one likes seeing zeros at the front of their paycheck.

It's more of a perception thing.

Once bitcoin becomes stable, any owner of bitcoins should wish that a small number of others' bitcoins were lost every year.  In fact, I hope that many of you lose your bitcoins right now.  Not too many of you, but at least some.  What else would you ask of a scarce currency, myrkul?

I believe that's been established. I like my money supply stable, neither inflating, nor deflating. That way, I know what the cost of a loaf of bread will be tomorrow, and the next day, ad infinitum.

I will grant you that knowing that my btc will become more valuable simply by not being lost is good. I will grant that other factors besides money supply determine market value. My point is, will people abandon btc when paychecks start to look small?

Of course, we could always phase in a btc2.0 when btc 1 starts to deflate to the point of absurdity (0.00000000001 btc for a meal, for example). It's all pretty moot, anyway. These are issues for our grandkids (or theirs!)

No, you are correct. Bitcoin is a slightly deflationary currency when looking only at hard coins.

However nothing stops somebody from setting up a fractional reserve bank based on BitCoins. In fact at some point it's downright likely. At that point you'd have quite high inflation as we see now. The blurb on the front page about being free of 'arbitrary inflation' does not seem really accurate to me.

As to losing wallets. It has of course happened. I lost the coins I minted back in mid 2009 when it was possible to make lots with just a CPU, sad face. As BitCoin gets more valuable though working, automatic wallet backup solutions will be built and losing your wallet will become less common.

Thought experiment time: How do you go about loaning "spreadsheet bitcoins"? Seing as how btc are designed to avoid double-spending, which is exactly what fractional reserve banking is.

Meh. Someone will come up with a catchy name for the sub-units. Then, instead of getting a paycheck of 0.000432 BTC, you will get a paycheck of 43,200 Satoshis. Problem solved.

I'll be happy knowing that tomorrow's loaf of bread will cost no more than today's loaf.

Fine points, both of them. Especially the first. Admittedly, I'm just picking nits here, and worse, my grandson's grandson's nits, but still, the question was a good one. Thanks (all of you) for answering it to my satisfaction.

Now: can anyone figure out a way to FRB with bitcoins? Cause I'd LOVE to be able to make some out of thin air. I contend that it can't be done, however.

How do you know that MtGox isn't already doing this? Holding a number of bitcoins that is smaller than the total of the bitcoin balances of all its customers?

Technically it's doable, although I don't see any economic incentive.

If anything, the incentive is the other way. Hold fewer dollars than the total of the Gox-dollar balances of all its customers. Buy bitcoins with the rest. When someone wants their dollars back, sell some of the bitcoins (and hope like hell that the exchange rate has gone up rather than down in the meantime).

Well, yes, it's technically doable to hold less than all your customers' accounts "hold"... but risky. True fractional reserve banking requires actual inflation: loaning out the same money twice or more. No bank can do that with bitcoins. The worst they could do is loan out so much that if they get a run, someone's getting screwed. (Actually, the worst they could do is loan out so much that if anyone actually had the temerity to use their money, they'd be screwed) As you pointed out, there's incentive against that. I don't think inflation will ever be an issue with btc.

Sure you can do that with bitcoins. Imagine if MtGox went into the lending business. They lend money to someone, and suddenly that person's MtGox balance shows some bitcoins, without MtGox needing to accept any more Bitcoin deposits.

If everyone who borrowed money from MtGox withdrew their money, that would cause a run. But if borrowers just used it as a "bank account" for paying to other MtGox accountholders, there would be no run.

With any kind of fractional reserve banking, if there is a run someone loses out. If MtGox went into the fractional reserve loans business and suffered a run, the losses would be borne by MtGox or by its creditors or by its depositors. When a bank suffers a run, the losses are borne by the bank's shareholders, or by its creditors or depositors, or by the taxpayer.

But that is not inflation. It just looks like inflation, on paper. If I think I have X BTC in my account, but MtGox only has Y (where Y<X) available, any transaction of Y+1 or greater will fail.

As long as they have income >= outflow, they'll get away with it. But that first big purchase will literally break the bank.

tl;dr: You cannot spend "spreadsheet bitcoins"

I think i see few problems with running a fractional reserve bank with bitcoin:

1. In the internet era information flows very quickly, so the probability of a bank run is high if people start to suspect something.
1a. It's too easy to call the bluff with bitcoin. If people want to take their funds out of the bank, they can do it instantenously. Much easier than with gold. So it is bank run is even more likely.
2. To run a fractional reserve bank in any country of the world, you need to cooperate with local authorities or you will be closed down. And since authorities will not understand what Bitcoin is, there may be problems.
2a. The central bank would have to have reserves of Bitcoin for fractional - Bitcoin based system to work.... Nah, too complicated, not likely to happen in the closest future.

I specifically didn't use the I-word.

Only external transactions fail in this scenario. You can send your X BTC to another MtGox account-holder, and nothing breaks. And if multiple banks are doing this, provided the transactions amongst them are roughly balanced (within the limits of each bank's fractional reserve), nothing breaks.

There's no doubt that fractional reserves will come to the bitcoin economy, but I think people will come to understand that a "real" bitcoin is safer than a "spreadsheet" bitcoin.

You're right, [mike] did. You're also right that internal transactions would not fail. But the first time a bank fails to pay an amount that should be there, I think that bank would go out of business. Fast. When starting a run is only a tweet away, Banks step carefully.

One way I see to make this work is to have a small portion of your account (A minimum balance) which is effectively the Bank's until you close the account. That amount of money is used to loan out or make other investments, while the rest is yours to spend. In exchange, you are essentially a stockholder in the bank, and earn dividends based on your share. You can buy a larger share by increasing your minimum balance. This way, while the bank has technically loaned out your money, They did it with your knowledge and consent, and as long as they don't loan out more than their shares, everyone can clear out their available balance without conflict. If you also agree that closing your account has a waiting period (at least to get that minimum balance back), even the shutdown of a bank should be able to be done in a relatively civilized manner.

That system is safe from loss due to fractional reserve, but the depositor carries some risk from the loan itself. Suppose the bank lends the depositor's money to a house-buyer who defaults, and the bank sells the house for less than the amount outstanding on the loan.

If this happens only occasionally, the default can be covered from the bank's profit margin, but if it happens a lot then the bank will default.

Investors always carry risk from the company making mistakes. Just imagine what would have happened had the iPad tanked.

Of course! But depositors in a bank don't think of themselves as investors. The investors are the bank's shareholders.

Well, under the plan I outlined, the bank's shareholders are it's depositors are it's investors. This would need to be clearly outlined in the deposit agreement, and spelled out in no uncertain terms, preferably verbally, before anything is signed. This could be spun positively:

NewBank: The bank where YOU own the bank!

But would definitely need to be said, Lest the depositor wonder where their interest is. The upside of this is, While few would actually make use of it, each account holder would have a say in the decisions of the bank (Again, votes based on shares, and thus, risk).

Depositor-operated "banks" have a long history. In the United Kingdom they are known as Building Societies if they lend on property, and as Credit Unions if they lend for other things.

They flourished in the past, but nowadays it's impossible for a small group of individuals to set one up legally (http://www.bsa.org.uk/faq/setupnewsociety). You're not allowed to start with less than £1 million, and the application process takes two years.

After all, society can't allow innovative new businesses to compete with established banks.

Yeah, I had the feeling I was reinventing the wheel. Credit unions are just about the only banks you can trust, over here in the US.

What's the Bitcoin banking situation look like?

Give each depositor a vote and you have a mutual bank, no?

I don't know if it's been mentioned before but what would happen if one person like Bill Gates bought most of the bitcoins with the intention of selling them back for a higher price.

The price would rise while he's buying them, then drop while he's selling them. The bitcoin system would cope with that.

He would make a profit, but only because the bitcoin value is rising anyway. If it weren't for that, the price rise while he buys would be expected to equal the price drop while he sells.

When someone with a lot of money starts buying up bitcoins, we know that Bitcoin has "made it" and is here to stay.

Fractional reserve issued currency is inflation by any reasonable definition of the word and is where nearly all inflation in todays fiat currencies comes from. Or did you think it's possible to inflate a currency the size of the US dollar by noticeable amounts through physically running printing presses?

Now if you want to argue that this is not "inflation" feel free but all it means is you aren't using the mainstream accepted definition of the word.

I don't think it quite fits that definition, as you're using it, since each depositor would have votes proportional to the amount they have "signed over" to the bank. Since that investment is at risk, the more they have invested, the more they should have control of how the bank's funds are used. Most small investors would not likely use this capability, but it would be there.


Well, yes, of course it is. I'm not arguing that the dollar isn't bloated beyond belief by spreadsheet dollars.

What I'm saying is you can't spend spreadsheet Bitcoins, as you can Dollars. When you attempt to transfer "fake" btc from your account to the seller, you'll get a double spend error. The bank can "loan" as many spreadsheet bitcoins as they want, but as soon as someone attempts to spend more than the real balance that the bank has on-hand, FAIL.

So, Unless you can show a way that a bank can make large amounts of usable bitcoins from thin air, Inflation isn't happening at the bank.

Someone needs to come up with the names of these fractions of bitcoin now and make them entrenched. Hopefully based on the number of zeros in front of the final number. Like a two5 bitcoin or a seven1 bitcoin because its pretty much inevitable that whole bit coins will keep going up in the number of any other currency needed to buy them. And humans are soo soo bad at comparing anything with more then 7 things all together. .0000000005 as opposed to a .00000000005. Those would be nine5 bit coins and ten5 bit coins respectively.. and when you get down to that scale I would not be surprised if theres a difference of 10 of any fiat currency you care to name between the two. Which would suck if you meant to buy a loaf of bread (which sells for 1 dollar) for ten5 bitcoin and ended up paying 11 dollars or nine5 bitcoins because you didn't count the zeros. At that point you know someone would offer bread at ten4 bitcoin.. which would be... .90 cents? See, my brain isnt wired for this currency..
Already someone needs to invent the fractional bit coin translator too.
And having bit coins arn't going to take the fluctuations of supply and demand into account either. One day a loaf of bread could be ten5 bit coins, the next seven1 bit coins, the day after thirteen3 bit coins. Its the flow of information and a little bit of over production that equals smoothness now, not the ability to infinitely fracture currency. Of course instant trading and multiple other currencies could help smooth the bumps.
I am having so much fun thinking about this as a currency, more then I want to own it..

I did come up with something like this.


OR perhaps:


etc.

People also proposed the µBTC and mBTC, but for some reason i think these won't work. Perhaps because "normal" (non-geeky) people don't really know names of all SI units.

Well maybe these people ought to become really normal and learn to add a couple new words to their dictionaries once in awhile

And stop using Internet Explorer while they're at it.

Also, they could start fighting for their personal freedoms and abolish governments...
Did that fairytale have dragons in it ?

Seriously, not gonna happen. Stupid people will mostly stay stupid.

What the hell is complicated about 0.001 BTC = 1 mBTC and 0.001 mBTC = 1 uBTC?

I refuse to believe that anyone would be too stupid to get this, even most Americans. It’s damn simple.

technically to talk about bitcoins completely you need to use nano. The issue with nano is it starts at 10^-9 so you have 10 nanobitcoins, but not 1nanobitcoin.
(bitcoins only go upto 10^-8)

we should use some name that sounds as good as bitcoin to talk about these base units.. like 'bitcoin units' or BCUs?
(i know people have suggested 'satoshis', but i don't see how that can be shortened..)
if we then ran most systems with both Bitcoins/BTC and Bitcoin Units/BCUs then I think it will alleviate issues of readability with decimal places.

Well, we all could just agree that when it comes to BTC, 'nano' means 10 ^ -8, instead of the usual 10 ^ -9. But that may be a bad idea.

Of course nobody really has to worry about this, not because it's not possible to implement a fractional reserve system in BTC, but because even a monkey could arbitrage the inflationary USD against the deflationary BTC.

If you want a lot of BTC get a loan in USD, and then convert, as your payments come due liquidate BTC and receive the advantage of the additional spread in conversion rate. No BTC bank could offer loans at low enough rates to compete with this poor mans arbitrage.

That's not quite accurate.
What causes the inflation in our world today are central banks, when they inflate the monetary base.

Fractional reserves are direct responsible for inflation only when the compulsory decreases. When it increases, it causes deflation. In a period where the compulsory is fixed, fractional reserves are nothing but a constant multiplier applied to any change the central bank promotes. If the central bank doesn't inflate nor deflate, a constant-compulsory fractional reserve system won't inflate nor deflate either.
You see what I mean?

It is perfectly possible to implement fractional reserves with BTC or any other currency.

Fractional reserve banking doesn't increase base money, only credit money. We've been living in a bubble, protected from fractional banks collapsing by the FDIC and FED printing. In our world credit money is as good as base money because base money will be debased if credit money is ever worth a penny less.

In free banking world there may be lots of new credit money, but that credit money will be a truly different thing than actual money. We just live in bizzaro world where the gov/banks make base money and they make credit money so they can keep them equal in value. But it's not natural for a promise to pay a dollar to be worth exactly a dollar when it's publicly known that the dollars aren't there.

Nice job not quoting the leading negation.

Sorry, I was just reading too fast, it was not intentional.