[yummyfajitas]

So much pop Keynesianism spreading real misunderstandings. I really wish people who argued against deflation would take the time to understand the underlying theories.

Deflation means, that the most fundamental law of finance is broken. The law, that says: It is much better to be given one dollar today, than to have one dollar next year.

Deflation does not break this law. It is always better to be given one bitcoin now than one bitcoin in the future - if you get it now you have the option of spending it, which has value. (If 1 BTC is worth 4 bags of pot today and 5 next year, 1 BTC in a year is worthless to me if I want to smoke now.)

He then argues that investment vehicles with a positive rate of return reduces consumption. This is true of both BTC and SPY. Unless bitcoin has the highest rate of return of all investment vehicles, the effect on consumption is irrelevant. I.e., if bitcoin has a return rate of 1% and SPY has a return rate of 2%, bitcoin will not affect the savings rate or reduce consumption [1].

If you are worried about people investing in bitcoin, consumption is irrelevant - the worry is that people will shift investments from SPY to BTC.

The fundamental problem with deflationary currencies is nominal rigidity. If you argue against deflation without mentioning this, you don't know what you are talking about.

http://en.wikipedia.org/wiki/Nominal_rigidity

tl;dr; Go learn Keynesian economics from a textbook that does math and carefully states assumptions. Don't learn it from newspaper columns.

[1] I oversimplify slightly. Since SPY and BTC are likely to be minimally correlated with each other, you can construct portfolios of both securities which will likely have a higher risk-adjusted return SPY alone. So the portfolio rate of return might go up a few bps, and consumption will shift a little bit towards investment. This is a small effect.

[daveungerer]

> It is always better to be given one bitcoin now than one bitcoin in the future.

A counter-example: Your pot dealer says you can have 4 bags of pot now, and you have 1 year to pay the 1 BTC you owe him. Being a man of better impulse control than you, and a prudent financial manager of his drug enterprise, he values 1 BTC in a year more than 1 BTC today.

That is, of course, assuming the expected value (after applying the deflation and factoring in risk of non-payment) is greater than the value he could build by investing the 1 BTC in his business today - an open question until we have some real deflation figures.

> (If 1 BTC is worth 4 bags of pot today and 5 next year, 1 BTC in a year is worthless to me if I want to smoke now.)

As I tried to illustrate above, finance is not something that's based on your personal preferences - it's a complex inter-dependent system.

[yummyfajitas]

Your counterexample makes no sense. If I pay him 1 bitcoin today and he puts the bitcoin under his mattress, he has 1 bitcoin in 1 year with 0 probability of non-payment.

    1 x 1 > 1 x P(payment)

He also has the option to invest in his business or anything else, which has value.

Your claim amounts to the idea that giving someone a loan at 0% interest can be profitable. It can't.

[daveungerer]

You're reasoning in circles. The loan is not 0% interest when denoted in dollars, since a deflationary currency will be worth more dollars as time goes by. Your reply only makes sense in a world where bitcion is the only currency in existence - a logical absurdity. You've disproved your point by negation - giving someone a 0% interest loan should indeed not be profitable, but with a deflationary currency it is.

Disclosure: I have no stake in the success or failure of bitcoin. I'm interested in knowing if you do, just so I know to stop wasting my time?

[yummyfajitas]

Suppose 1 BTC = $X today and $Y in a year. If I pay my dealer 1 BTC today and he doesn't spend it, he has $Y in a year. If he extends me a 0% interest loan for 1 year, there is a probability P < 1 he has $Y in a year, for an expected value of P x $Y < $Y.

Loaning someone an asset at 0% is always worse than holding the asset. The value of the loan is P(repayment) x value of asset in future. The value of the asset today = value of asset in future + option value. There is no escaping this mathematical identity. Provided P(repayment) < 1, you are better off holding the asset, regardless of the rate of return on the asset.

If you were correct, then you could make a profit by loaning out shares of SPY or other security with a positive expected rate of return at 0%. You can't.

If you disagree with this, please express your disagreement in math.

I don't own any bitcoins, and have no stake in it. I think it's a good idea that might solve the current mess of payment systems, however.

[daveungerer]

> Loaning someone an asset at 0% is always worse than holding the asset.

Exactly. My previous reasoning was indeed a bit muddied, but what you're saying now clarifies it. I don't disagree with your math.

There are so many logical contradictions in the concept, which is why I end up making silly assertions that you correctly pointed out. Instead, the argument should simply be that the rate of return on bitcoin, due simply to built-in deflation, and expressed in dollars, may or may not be so high that it always makes sense to hold the asset rather than use it as currency to buy other assets. In which case it can't function as a currency: contradiction.

Of course, I do believe the BTC USD exchange rate will be the dominant factor in determining prices expressed in BTC in the long term, but I'm sure plenty of people would disagree. If that does happen, a lot of this becomes theoretical - the worsening exchange rate will take care of the deflation.

Note I'm not necessarily for or against bitcion. Thanks for the discussion.

[ahh]

> Loaning someone an asset at 0% is always worse than holding the asset.

Nit: if the carry cost of the asset is zero. This is _not_ true for dollars (at least at large corporate/pension fund/hedge fund scale) and various government securities have occasionally sold at negative nominal interest rates.

(It is, essentially, true for BTC, which is why your analysis is largely correct here, but there certainly are cases where one would reasonably lend USD or the like at 0%.)

[maxerickson]

Then the dealer doesn't value the pot at 1 tomorrows bitcoin today. His willingness to accept payment in the future points at the discount he sees.

I don't have any interest in bitcoin, I'm a spectator. Edit: I realized that's not entirely true, I got 1.5 namecoin from a fountain a couple of years ago.

[waterhouse]

> Being a man of better impulse control than you, and a prudent financial manager of his drug enterprise, he values 1 BTC in a year more than 1 BTC today. [...] expected value (after applying the deflation and factoring in risk of non-payment)

I think the status of that example as an alleged counterexample results from a subtle misuse of words. If the pot dealer's choice were between 1 BTC now and 1 BTC a year from now, the former is still of greater or equal value. If you are able to factor in "risk of non-payment", then you are talking about a decision between "attempting to get you to pay 1 BTC now" and "attempting to get you to pay 1 BTC during the succeeding year", which is by stipulation a materially different choice than "1 BTC now" vs "1 BTC a year from now".

As for the deflation--if the pot dealer believes that the increase in value from deflation is greater than what he could get by investing in his business (or in anything else), then he always has the option of sitting on the BTC for a year. The only reasons I can think of for this not to be the case are (a) if the pot dealer lacks impulse control and would spend it and regret it, or (b) if the pot dealer may be robbed of all his worth during the intervening year [so getting the BTC a year later is essentially a substitute for a safe money-storage service]. Both of these possibilities are generally assumed out of existence in this context, hypothesizing a hardheaded economic actor and a market with well-defended property rights.

[daveungerer]

> he always has the option of sitting on the BTC for a year.

Eureka! That's also known as bitcoin hoarding. If the value (in dollar terms) of holding on to bitcoin is indeed greater than spending it, people and financial institutions will hold on to it and primarily keep transacting in dollars. Which leads to a nice logical contradiction, since how can a currency be worth anything if no-one wants to spend it?

[pikachu_is_cool]

....that's called a loan.