The Bitcoin Central Bank’s perfect monetary policy

http://themisescircle.org/blog/2013/12/15/the-bitcoin-central-banks-perfect-monetary-policy/

by mike_esspe • 13 years ago

47 50 13 years ago

dsjoerg• 13 years ago

The paragraph that starts "First, it is rational for economic agents to hold as many bitcoins as they can afford to lose" is so poorly argued that I decided the rest of the article wasn't worth reading.

There are several steps missing between the statement "Bitcoin will go up in value" and "I should invest all my investable assets in it". One big missing step is, what is the optimal way for one to allocate one's portfolio, given some set of expectations, uncertainties and correlations?

There's been a lot of work in this area (optimal portfolio construction), some of it very good. My favorite starting point for this field is the Kelly Criterion. https://en.wikipedia.org/wiki/Kelly_criterion

While I sympathize with the article's enthusiasm for Bitcoin, I wish people would shut up when they have no idea what they're talking about.

EDIT: if you want to get deep in the weeds you can read more about Modern Portfolio Theory here: https://en.wikipedia.org/wiki/Modern_portfolio_theory. I don't endorse every idea you'll find there, but they are all better than the notion that you should invest 100% of your investable portfolio in whatever asset has the strongest forecast, as the OP apparently believes one should.

VienneseCPA2• 13 years ago

You really should have read what's in the parentheses immediately after that: "without materially impairing their ability to consume or invest".

In other words, don't invest all your money in it.

I wish people would continue reading instead of stopping at arbitrary points to rant.

Tycho• 13 years ago

I'm not 100% sure but I think that part in parentheses was added after the OP made his original post.

crystaln12• 13 years ago

I'm a fan of bitcoin in many ways, but this is fanboy fiction with no understanding of money.

Fractional reserve banking entails the creation of new money that is fungible with already preexisting money, i.e. it can be used interchangeably within the currency’s payment systems. This is impossible with Bitcoin.

Not true. Currency can easily be issued against bitcoin. When USD was backed by gold, there was plenty of fractional reserve banking and no-one could issue more gold. In fact, there is already debt-based money backed by BTC issued on the Ripple network.

EDIT: I should add, that fractional reserve banking does not mean infinite money supply. Depending on the amount of reserves required (by customers, government, etc.), as long is it's greater than 0%, there is a limit on the supply. E.g. if the reserve is 20%, only 80 BTC can be loaned against 100, and only 64 against wherever the 80 is deposited, and on-and-on until it approaches 0. Fractional reserve is not necessarily a bad method to allow for risk assessed expansion and contraction of money supply.

...blah blah... make it certain that ... Bitcoin will be adopted as the global currency

Oh really? It's certain that the globe will adopt a currency because of it's asymptotic money supply and proof of work algorithm? That makes no sense at all.

The world will adopt bitcoin because it's an efficient means of exchange and safe as a value store / investment from dilution by central banks.

Also, the "security" row in the currency comparison chart is misguided. Bitcoin is far less secure than bank maintained fiat currency networks, because in the latter, fraudulent transactions and electronic theft are generally reversible.

VienneseCPA2• 13 years ago

Would you please write an article about Bitcoin monetary policy with your understanding of money?

dnautics3• 13 years ago

I'm a fan of bitcoin, as well, and I totally agree with you on their #3 (Fractional reserve banking). I was going to come here with exactly the same statement about gold.

There is, however, a key difference in that unlike contemporary FRB, any FRB with bitcoin will necessarily have all of the inflationary effects of monetary expansion through debt exactly counteracted by deflationary effects of debt amortization and bankruptcies.

fragsworth2• 13 years ago

There's a major difference. With Fiat, you're basically forced to use a bank, because that's the safest way to hold your cash. Every bank, however, is participating in fractional reserve lending, so you're also forced to take part - and the government is forced to insure everyone from potential losses.

Fractional reserve banking can happen with Bitcoin, but as a user, you do not have to participate. You can hold your coins in your own wallet and not have to worry about what your bank is doing with it. This means you can never lose your coins from a run on the bank.

Now, if you want to, you can loan your coins out with interest. Or you can give them to a bank and collect an interest rate. But when you do this, you're taking a risk, and you should assume any losses from this risk on your own, and nobody needs to be insured for it. If you don't want to lose your coins, keep them in your own wallet.

eof• 13 years ago

You've actually got it wrong.

Debt-based "ripple" bitcoin (or any other similar system) is not fully interchangeable; you can use BTC IOUs within ripple and pass them around, but you cannot send an IOU across the bitcoin network. This is different than with fractional reserve; where 'created' money can be withdrawn, spent, and intermingled without any differentiation.. I can spend a fractionally-reserve created dollar at walgreens as easily as a bona-fide fed printed dollar bill.

You cannot send a Ripple BTC to my bitcoin-qt wallet, nor will bitpay/coinbase/blockchain or any other blockchain-audited wallet accept a btc-iou.

edit: i basically agree with your other point that the the stated reasons in OP are not the reasons bitcoin will likely undergo s-curve adoption

wmf3• 13 years ago

Unless blockchain transaction fees become so high that BTC-denominated IOUs become the standard.

crystaln12• 13 years ago

You're right that BTC IOUs on Ripple are not fully interchangeable with BTC. However, neither was gold interchangeable with USD, yet each gold note was redeemable for gold. This did in may ways restrict the level of fractional reserve banking, and certainly the direct issuance of money. (Many suspect that the US did not have nearly enough gold to redeem all its currency, and that this was a primary motivation for leaving the gold standard.)

The process of redemption of BTC IOUs for actual BTCs on the Ripple network resembles the process of redemption of gold-backed USD for gold. Similar to gold being transmitted over a different network than gold-backed USD (the former being physical,) BTC backed notes would be transmitted over a different network - something like Ripple.

Fractional reserve banking can also be practiced directly with BTC, simply by lending against deposits.

(explained here: http://www.cnbc.com/id/101048897)

jacques_chester11• 13 years ago

I think that Bitcoin is ultimately less secure because it has two single points of failure: the design and the implementation.

jacques_chester11• 13 years ago

The von Mises Institute interpretation of economics, which is actually a Rothbardian interpretation, dislikes inflation. A lot. They miss gold.

The problem is that gold is not like bitcoin, and bitcoin not like gold.

Suppose we all switch immediately to a gold standard. When the value of the currency rises too high, more people will go and mine it. When the value of the currency falls, fewer people will mine it. So in actual fact a gold standard resembles the behaviour of an Australian-style (yes, Australian) monetary policy, which is having money supply governed by a central bank who have an explicit inflation target.

Meanwhile, the supply of bitcoins is fixed. For all time. And they are destructible in a way that gold is not.

Bitcoins will always be irreversibly leaking from global supply. People throw their harddrives away. Delete the files accidentally. A high energy particle strikes some cheap RAM and flips a bit, rendering a bitcoin worthless, and so on.

While in theory the supply of bitcoin is meant to reach a fixed limit, over the long run it will peak and thereafter always be shrinking.

Not expanding. Not even holding constant. Shrinking.

Nobody's economics says that a shrinking monetary base is a good idea.

crystaln12• 13 years ago

Bitcoin is what it is. It doesn't have to become the only currency. It's a useful, fungible, easily transferrable commodity.

The argument that bitcoin should not to succeed because it's deflationary is silly. Lots of deflationary assets exist and people are free to put their money into them, e.g. gold, stocks, and real estate. That doesn't stop people from transaction.

jacques_chester11• 13 years ago

It doesn't have to become the only currency.

No, but arguing from the analogy of a gold standard hints at doing so.

The argument that bitcoin should not to succeed because it's deflationary is silly.

Deflation and inflation are both, at the extremes, destructive. Bitcoin, by intention, is deflationary. However I have already demonstrated that it can easily turn inflationary.

Not through the dynamism of an open market, and not through the governance of a central bank. Simply because it is a closed system. Entropy will eventually destroy all bitcoins. And then what?

lingben• 13 years ago

useful? not really, unlike money it requires digital equipment, bitcoin is NOT fungible (each bitcoin is uniquely identifiable and can be separated, isolated and even removed from circulation and it is not 'easily transferrable' (a bitcoin transaction takes several hours until it is confirmed)

miles3• 13 years ago

Meanwhile, the supply of bitcoins is fixed. For all time. And they are destructible in a way that gold is not.

But they are divisible in a way which gold (or even dollars) are not: https://en.bitcoin.it/wiki/FAQ#How_divisible_are_bitcoins.3F

"A bitcoin can be divided down to 8 decimal places. Therefore, 0.00000001 BTC is the smallest amount that can be handled in a transaction. If necessary, the protocol and related software can be modified to handle even smaller amounts."

jacques_chester11• 13 years ago

I'm not arguing that bitcoin is very fungible; that is by design. I'm arguing that bitcoins, unlike gold, can be irreversibly destroyed.

Even the "leaks" from the gold supply are predictable and reversible. A lot of gold is used in electronics these days. It winds up either in recycling centres, where it can be recovered, or in landfill, where if the price of gold rises high enough it becomes commercially viable to mine the landfills.

Bitcoins cannot be un-destroyed. They are information without physical laws to guarantee their continued existence. Their manufacture and destruction are both one-way functions.

The supply of bitcoin will peak and begin an irreversible process of shrinkage.

Indeed, those who cannot steal bitcoin will be incentivised to cause the destruction of other's bitcoin. Call it the Bitfinger Scenario.

jellicle• 13 years ago

You think gold can't be divided into arbitrarily small units? Really? The smallest unit of gold is (179/(6*10^23)) grams. That's pretty small.

Dividing bitcoins into small micro-denominations does not solve the problems of deflation in the same way that adding zeros to Zimbabwean currency doesn't solve the problems of inflation.

dualogy• 13 years ago

Suppose we all switch immediately to a gold standard. When the value of the currency rises too high, more people will go and mine it.

An interesting twist on the basic problem with fixed-price gold standards but you might be right for countries that have gold in the ground outside of controlled mines. But a fixed-price gold standard is a big fail no matter what price is set and subsequently has to be defended by the central bank. It worked "somewhat, sub-optimally, for a while" in the old days of a handful of developed slower-non-realtime economies that could coordinate policy and had the rest of the world colonized, today, no chance. Central bankers will never again go for it because they have learned that politics and private banking will always screw with the money supply for short-term objectives.. and no matter what fixed gold price you set, before long there'll always be a run on your gold. Better to mark reserves to market (as France has pioneered in the 70s and the ECB and others do today), so you don't need to constantly defend or redefine your fixed price, can manage the currency supply based on both long-term mandates and short-term needs and have its performance evaluated in gold and/or other currencies by the market forces, too.

"Any nation/state can put its economy/currency on a gold standard. They only have two requirements. Own a stockpile of gold and raise the price very high" ;)

It's funny that the "free-market advocates" these days favour a government-controlled fixed gold price, which they call "the return to an honest gold standard". Right now we have an "almost free-market gold price" in "almost free-FX-market floating (competing!) currencies". Not quite there yet according to some pundits but that's where the world is heading and has ever since the Nixon shock. Definitely a pretty impressive move that has emerged toward a "free market in moneys", what's not to like! ;)

moocowduckquack2• 13 years ago

All of this makes me wonder how long currencies themselves will last. Effectively a monetised market is using currency exchange to compute a function where the currency serves as a variable to store the current state of the computation.

From this perspective, money is one of the first distributed computation networks, however given the mass of networked computation available there may be other ways of solving the distribution problem that have even less friction than a currency.

gnaritas• 13 years ago

If we use it as a means of exchange, whatever it is, we're going to call it currency.

moocowduckquack2• 13 years ago

There are markets and systems of debt arrangements that predate currencies. Exchange is not predicated on the use of currency as a technology. Currencies have held sway for a very long time however because they are an exceedingly simple and elegant solution for a fiendishly complex problem.

These days however we have a global network of universal turing machines attached to an increasingly large array of robots, and so things that are not algorithmically possible to do in a monetary system are now completely feasable and some of them may have considerably less friction.

miles3• 13 years ago

The full reserve nature of Bitcoin is often overlooked (or misunderstood - more below):

"The [Bitcoin network] enforces the strictest deposit regulations in the world by requiring full reserves for all accounts. This is the digital equivalent of the Chicago Plan or the Austrian 100% reserve gold standard. Under this regulatory regime, money is not destroyed when bank debts are repaid, so increased money hoarding does not cause liquidity traps, instead it increases real interest rates and lowers consumer prices. This is a self-stabilizing cycle as higher interest rates incentivize hoarders to invest, while deflation increases consumption due to the wealth-effect on hoarders. The BCB prevents lending out of deposits so that it can properly target money supply and avoid the destabilizing effects of commingling the credit and payment systems."

Others have argued[1] that fractional reserve Bitcoin banks can exist, but the point is that the Bitcoin protocol itself is full reserve. Yes, intermediaries may try to convince consumers to entrust their Bitcoin wallets with them, but that hasn't gone so well[2]; Bitcoin "banks" that fool around with fractional reserves are playing with fire.

UPDATE: Coinkite, which describes itself as a "Cryptobank", addresses[3] the need to have verifiable, full reserves: "You should also be wary of bitcoin startups which claim some percentage of your coins will be in cold storage. If you do not know the details of the public keys used, it will be impossible to know if this is true, or if they are using those amounts as play money. With Coinkite, you can audit where all your funds are at any time."

[1] Of course you can have fractional reserve Bitcoin banks http://www.cnbc.com/id/101048897

[2] List of Bitcoin Heists https://bitcointalk.org/index.php?topic=83794.0

[3] Coinkite FAQ https://coinkite.com/faq/money

wmf3• 13 years ago

Maybe it's to Bitcoin's benefit that people don't appreciate the ramifications of full reserve. I suspect most people who have grown up with Keynesianism, credit cards, and "free checking" will have no desire to adopt full-reserve banking with its resultant lower velocity of money. Looking at Coinkite specifically, their fees seem crazy high and it seems inevitable (within the market for lemons that is the Bitcoin ecosystem) that a scammy fractional-reserve competitor (perhaps developed by a 17-year-old) will come along with a slick Web site and lower fees.

negamax• 13 years ago

Novice question : Can in say 10-15 years, Bitcoin's crypto standards be bypassed using raw computing power? How about 20 years?

jacques_chester11• 13 years ago

Without meaningful breakthroughs in attacks on the underlying cryptographic primitives, probably not.

If it's to be broken, it will most likely be due to someone finding a flaw in the way those primitives are assembled. In the nearer term, the main risk is people finding attacks on the implementation (in practice nobody uses 3rd party implementations).

fragsworth2• 13 years ago

Even if a serious flaw in the current encryption algorithm is found, it's feasible to create a new encryption algorithm and support both, simultaneously, allowing people to transfer their coins over to the new one.

If some quantum computing mechanism is developed that negates all cryptography entirely, then we will probably be living in a world with unlimited computing power and the robots will have taken over anyway.

wmf3• 13 years ago

No. All modern crypto is designed to take at least hundreds of years to brute-force. Algorithmic attacks are what people worry about.

crystaln12• 13 years ago

Possibly, but there is likely to be enough time to upgrade the network between the time it takes to crack a private to drops from 50 universes to 1 universe.

That is of course precluding the discovery of a major flaw.

tptacek• 13 years ago

Probably not, in the sense of computers getting better and scaling up say Pollard-Rho. "Probably" because who knows what computers will look like in 20 years.

But nobody knows what new attacks people will come up with, either against 256 bit ECC or SHA2.

The Bitcoin Central Bank’s perfect monetary policy

Related Stories