12 ms·

The Bitcoin Central Bank’s perfect monetary policy

47 points by mike_esspe 13y ago

dsjoerg 13y 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 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 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.
- VienneseCPA 13y 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 13y ago
    I'm not 100% sure but I think that part in parentheses was added after the OP made his original post.
crystaln 13y 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.
- VienneseCPA 13y ago
  Would you please write an article about Bitcoin monetary policy with your understanding of money?
- dnautics 13y 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.
jacques_chester 13y 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.
- crystaln 13y 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_chester 13y 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?
moocowduckquack 13y 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 13y ago
  If we use it as a means of exchange, whatever it is, we're going to call it currency.
  - moocowduckquack 13y 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.
miles 13y 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 http://www.cnbc.com/id/101048897 [2] List of Bitcoin Heists https://bitcointalk.org/index.php?topic=83794.0 https://bitcointalk.org/index.php?topic=83794.0 [3] Coinkite FAQ https://coinkite.com/faq/money https://coinkite.com/faq/money
- wmf 13y 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 13y ago
Novice question : Can in say 10-15 years, Bitcoin's crypto standards be bypassed using raw computing power? How about 20 years?
- jacques_chester 13y 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).
  - fragsworth 13y 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.
- wmf 13y ago
  No. All modern crypto is designed to take at least hundreds of years to brute-force. Algorithmic attacks are what people worry about.
- crystaln 13y 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 13y 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.

47 points by mike_esspe 13y 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 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 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.

VienneseCPA 13y 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 13y ago
  I'm not 100% sure but I think that part in parentheses was added after the OP made his original post.

crystaln 13y 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.

VienneseCPA 13y ago
Would you please write an article about Bitcoin monetary policy with your understanding of money?
dnautics 13y 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.

jacques_chester 13y 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.

crystaln 13y 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_chester 13y 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?

moocowduckquack 13y 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 13y ago
If we use it as a means of exchange, whatever it is, we're going to call it currency.
- moocowduckquack 13y 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.

miles 13y 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 http://www.cnbc.com/id/101048897 [2] List of Bitcoin Heists https://bitcointalk.org/index.php?topic=83794.0 https://bitcointalk.org/index.php?topic=83794.0 [3] Coinkite FAQ https://coinkite.com/faq/money https://coinkite.com/faq/money

wmf 13y 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 13y ago

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

jacques_chester 13y 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).
- fragsworth 13y 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.
wmf 13y ago
No. All modern crypto is designed to take at least hundreds of years to brute-force. Algorithmic attacks are what people worry about.
crystaln 13y 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 13y 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.