Deflation and Bitcoin, the last word on this forum

[99Percent]

There is nothing inherently wrong in holding money instead of spending.

I'm not saying it is immoral or anything like that. If there's deflation there's an incentive to spend. I'm not judging anyone for playing by the rules. The set of rules that define different monies is what interest me and want to discuss.

Money held in the long term could accumulate to the point of reaching capital value (like mortgage payments on a hous). Deflation will be bad if people held money because there is just no way to spend it in a "worthwhile" way. But bitcoins have the wonderful characteristic that it is very divisible. So there will always be just something even very diminute where people would be willing to "let go" of some of their hard earned bitcoins.

I understand that we're not going to "run out of bitcoins" or anything similar. But deflation increases the costs of commerce and discourages investments. While trying to stop it by inflation can only postpone the problem and make it worse, deflation damages markets. Its effects, and what causes big deflations are different matters, of course.

When I was saying there is nothing inherently wrong, I wasn't talking in the moral sense, but bitcoin as currency in an economy.

Deflation doesn't increase the cost of commerce or discourage investment, that is a myth as evidenced by the computer/tech industry.

[lonelyminer (Peter Šurda)]

But deflation increases the costs of commerce and discourages investments.

I read a paper recently, http://www.mises.cz/clanky/fungovani-ekonomiky-v-podminkach-poklesu-cenove-hladiny-254.aspx. It's in Czech, I don't know how well the auto-translate works so maybe you won't be able to get the whole thing (I don't know which languages you speak). The author makes an interesting point, in that the price of producer goods adapts to that of consumer goods. So, if you have a deflationary monetary system, the price will adapt in such a way that the interest rate won't be negative. I'm too lazy to quote exactly, it will be in my paper (including translation). On the other hand, in order for this to work, the deflationary effect must be significant. If you just have one deflationary currency (e.g. Bitcoin), but the dominant is inflationary fiat, it probably wouldn't work that way (I'm not 100% sure yet how different it would be exactly).

[jtimon]

@99Percent
I don't think the computer industry evidences anything. I'm talking about general prices deflation, not about products reducing their price by innovation and efficiency.
Here's how I see it.

About increasing costs of commerce...
Whatever you trade, deflation hurts you for every item you have on stock. You must compensate those losses somewhere else. 

@lonelyminer
My native language is Spanish. I also speak English and Portuguese.
I never claimed that the interest rates could be negative with deflation. That should never happen without demurrage.
It happened recently with German bonds. That means people trust more german bonds than having money deposited at a bank or at home (really strange IMO, but not impossible) or that someone is willing to give money away to Germany. The ECB, so that Germans don't feel jealous about Greece?
If you're talking about real interest rates rather than nominal, it happens today through printing manipulation. But with deflation, real interest rates are greater than nominal rates.
Just in case, I'll put the definitions I'm using here again:

nominal interest = real interest + inflation premium = basic interest + risk premium + inflation premium

Inflation premium is negative when there's deflation.
Basic interest is sometimes called "liquidity premium".

[99Percent]

@99Percent
I don't think the computer industry evidences anything. I'm talking about general prices deflation, not about products reducing their price by innovation and efficiency.
Here's how I see it.

About increasing costs of commerce...
Whatever you trade, deflation hurts you for every item you have on stock. You must compensate those losses somewhere else. 

If you believe deflation is general, holding items which you will sell later at a lower price does not hurt you because the prices of other goods and services will also go lower.

The cost of deflation and inflation will hurt if it spirals out of control. But because bitcoins are easily divisible and the rate of new coins is known beforehand this will not happen, much like what happened in the computer world with Moore's law.

[lonelyminer (Peter Šurda)]

Hello jtimon,

My native language is Spanish. I also speak English and Portuguese.

unfortunately then you'll have to wait for my (or someone else's) translation if the automatic one is inadequate.

I never claimed that the interest rates could be negative with deflation.

You said:

Say we have 8% deflation and 5% capital yields. The nominal interest rate would need to be -3% (which is just impossible with most monetary systems).

According to the article I referenced (Krupa), this will not happen (assuming by "deflation" you mean an increase in productivity over the increase of the money supply, rather than contraction of money supply in absolute numbers).

It happened recently with German bonds. That means people trust more german bonds than having money deposited at a bank or at home (really strange IMO, but not impossible) or that someone is willing to give money away to Germany. The ECB, so that Germans don't feel jealous about Greece?

German bonds are just one of many things on the market, they are not representative. Also they are not producer/consumer goods, they are debt instruments.

If you're talking about real interest rates rather than nominal, it happens today through printing manipulation. But with deflation, real interest rates are greater than nominal rates.

My point, citing the article, was that even the nominal rate would not be zero or lower, because the relationship between producer/consumer goods would restructure.

Inflation premium is negative when there's deflation.

This means that the productivity, on average, rises faster than the money supply.

[jtimon]

If you believe deflation is general, holding items which you will sell later at a lower price does not hurt you because the prices of other goods and services will also go lower.

It hurts you nominally. Anyone will try to avoid this "loss".

The cost of deflation and inflation will hurt if it spirals out of control. But because bitcoins are easily divisible and the rate of new coins is known beforehand this will not happen, much like what happened in the computer world with Moore's law.

I claim deflation is harmful in general and specially if it is greater than basic interest.

unfortunately then you'll have to wait for my (or someone else's) translation if the automatic one is inadequate.

I don't think automatic translations are acceptable right now. Fortunately, from what you're saying, I think we agree on the conclusions of the article.

I never claimed that the interest rates could be negative with deflation.

You said:

Say we have 8% deflation and 5% capital yields. The nominal interest rate would need to be -3% (which is just impossible with most monetary systems).

According to the article I referenced (Krupa), this will not happen (assuming by "deflation" you mean an increase in productivity over the increase of the money supply, rather than contraction of money supply in absolute numbers).

By deflation I mean price deflation: a general drop in prices. This can be caused by what you say, by a shrinking money supply or by a drop in credit transactions.

My point, citing the article, was that even the nominal rate would not be zero or lower, because the relationship between producer/consumer goods would restructure.

I agree.
My point is that the market would restructure by impeding new investments. If nobody produces new capital (producer goods, that later produce consumer goods), the price of consuming goods will rise. That will increase the yields of producing goods so that they will be produced again.
This process of "not investing" is destructive for the economy. 

Inflation premium is negative when there's deflation.

This means that the productivity, on average, rises faster than the money supply.

As said, this can also mean that the money supply or the use of money competitors (barter, credit transactions) are shrinking.
One can argue that fractional reserve is not likely to happen with bitcoin and that's probably true, but there's still barter and credit transactions.
They will increase if the interest is too high and then they may drop causing price deflation.
To understand what I mean by money competitors and credit transactions you can read this:
http://www.community-exchange.org/docs/Gesell/en/neo/part3/9.htm

Here you can find more arguments about the dangers of deflation:
http://www.community-exchange.org/docs/Gesell/en/neo/part3/11.htm

[lonelyminer (Peter Šurda)]

I agree.

Jolly good. That saves a lot of time

My point is that the market would restructure by impeding new investments.

But this is precisely a misunderstanding of Krupa's (and before that Huelsmann and Mises) argument. It won't impede new investments, it will change them. That is the point of the argument. Producer goods will find employments other than those that would have been done in a non-deflationary system. But they will still find employments.

As said, this can also mean that the money supply or the use of money competitors (barter, credit transactions) are shrinking. One can argue that fractional reserve is not likely to happen with bitcoin and that's probably true, but there's still barter and credit transactions. They will increase if the interest is too high and then they may drop causing price deflation. They will increase if the interest is too high and then they may drop causing price deflation.

Mises, in the Theory of Money and Credit, argues (if I understood it correctly) that credit itself has no effect on the "objective exchange-value" of money (in current terms, we'd just say the price of money). It also shouldn't have an effect on the interest rate, but it might be a symptom of a shift in the interest rate. The Austrians define the natural rate of interest as that which would exist if there was no money. Elastic money supply can only distort this. While an inelastic supply can, in my opinion, theoretically be improved upon to match the natural rate of interest better, it probably can't be practically implemented due to the calculation problem.

To understand what I mean by money competitors and credit transactions you can read this:
http://www.community-exchange.org/docs/Gesell/en/neo/part3/9.htm

Here you can find more arguments about the dangers of deflation:
http://www.community-exchange.org/docs/Gesell/en/neo/part3/11.htm

Thanks, I will read it.

[jtimon]

My point is that the market would restructure by impeding new investments.

But this is precisely a misunderstanding of Krupa's (and before that Huelsmann and Mises) argument. It won't impede new investments, it will change them. That is the point of the argument. Producer goods will find employments other than those that would have been done in a non-deflationary system. But they will still find employments.

But investing in producing goods is less attractive under deflationary conditions. Maybe you can tell me what's wrong with my bakery example.
Although I agree that with or without deflation nominal nor real interest rates can drop below zero (aside from central banking manipulated money of course), I'm not sure I understand the theory about how investments would change with deflation in a way that's beneficial for the economy as a whole.

Other Austrians (Hayek) admit that deflation is problematic.
"It is agreed that hording money, whether in cash or in idle balances, is deflationary in its effects. No one thinks that deflation is in itself desirable."
Printing and government spending is not the answer, that's all.

Mises, in the Theory of Money and Credit, argues (if I understood it correctly) that credit itself has no effect on the "objective exchange-value" of money (in current terms, we'd just say the price of money). It also shouldn't have an effect on the interest rate, but it might be a symptom of a shift in the interest rate. The Austrians define the natural rate of interest as that which would exist if there was no money. Elastic money supply can only distort this. While an inelastic supply can, in my opinion, theoretically be improved upon to match the natural rate of interest better, it probably can't be practically implemented due to the calculation problem.

The price of money often means the interest rather than the exchange value. We can use that term but that may be confusing.
Credit, when used as a mean of exchange certainly affects the exchange value of money, just like barter or any other thing competing with money as a tool for exchange. If you have bimetalism and then you demonetize silver, the price of gold will increase.
All the products offered on the market represent the offer. All the monetary base at offer (not counting hoarded money) plus the bills of exchange (credit) plus the barter credits, plus LETS hours, etc. represent the demand. If that's true, an increase in barter (for example) should cause price inflation (even with the monetary supply untouched).

Gesell claims that the basic interest is a purely monetary phenomenon and therefore does not exist when there's no money.
The natural interest rate would be zero. Well, no, the natural interest rate would only depend on the risk.
Without money: interest rate = risk premium
What Gesell wants is to suppress the basic interest because according to him it is a unearned rent. Just like taxing food imports represents a rent for land owners.

Thanks, I will read it.

Thank you. That will probably save us some time too. You will be able to detect my "misconceptions" inherited from Silvio Gesell.

[lonelyminer (Peter Šurda)]

But investing in producing goods is less attractive under deflationary conditions.

You are missing the second half of the picture, the "unseen" as Bastiat would say. It is less attractive to invest into some production processes, but this is offset by other business processes becoming relatively more attractive.

Maybe you can tell me what's wrong with my bakery example.

The same as I said in previous paragraph. It's only valid if you disregard the compensating effects.

Although I agree that with or without deflation nominal nor real interest rates can drop below zero (aside from central banking manipulated money of course), I'm not sure I understand the theory about how investments would change with deflation in a way that's beneficial for the economy as a whole.

Well, lets first avoid the question of whether this would be beneficial. The point is that this unattractiveness of investment A will be compensated by attractiveness of investment B. I also described it somewhere on the example with apples and apple pies, maybe even in this thread.

Other Austrians (Hayek) admit that deflation is problematic.
"It is agreed that hording money, whether in cash or in idle balances, is deflationary in its effects. No one thinks that deflation is in itself desirable."
Printing and government spending is not the answer, that's all.

Deflation is not problematic per se. It is problematic when it is sharp and unexpected by most market participants. From this point of view, it is similar to inflation. But normally, circulating money does not contract on its own. People do not suddenly get the idea to sharply switch to ascetic living style en masse. But if you have a credit expansion, then a credit contraction must follow: money supply shrinks and there is downward pressure on prices. This is exacerbated by central banking and state's interference in money. Without it, credit expansion might, hypothetically, end up in an equillibrium state in the long run.

The price of money often means the interest rather than the exchange value. We can use that term but that may be confusing.
Credit, when used as a mean of exchange certainly affects the exchange value of money, just like barter or any other thing competing with money as a tool for exchange.

Yes. But if credit is performed by the creditor abstaining from consumption (i.e. no change of money supply), then this effect does not take place.

If you have bimetalism and then you demonetize silver, the price of gold will increase.

I agree (all other assumptions being the same).

All the products offered on the market represent the offer. All the monetary base at offer (not counting hoarded money) plus the bills of exchange (credit) plus the barter credits, plus LETS hours, etc. represent the demand. If that's true, an increase in barter (for example) should cause price inflation (even with the monetary supply untouched).

It is possible if there is a permanent shift between monetary and barter exchanges, for example, or credit vs. credit expansion, that this is accompanied by a change in the interest rate. But I would say that both are just symptoms of the same cause: a change in the preferences of the human actors.

Gesell claims that the basic interest is a purely monetary phenomenon and therefore does not exist when there's no money.

Austrians would heavily disagree. There is interest even in a non-monetary economy. It's just a ratio of the value of future vs. present goods, i.e. a consequence of the time preference.

[friedcat]

I can't help myself showing this pic.

[JoelKatz]

I'll repeat it, because I like to say it at least once every few months: Long-term, predictable currency-driven deflation is essentially impossible.

If I give you 10 Bitcoins today, you can hold onto them and have 10 Bitcoins next year. You also have the option of doing anything else with those 10 Bitcoins during that year if you wish. So 10 Bitcoins today must be worth at least as much as 10 Bitcoins next year -- because it's that and then some.

Long-term, predictable deflation says that 10 Bitcoins next year can be reliably and predictably more valuable than 10 Bitcoins now. That would require the option to use the Bitcoins early to have less than no value at all, which is impossible.

As for deflation's effect on loans, permit me to quote my favorite expert on economics arguing that loans and borrowing are not affected by currency-driven inflation or deflation:

The benefit to the borrower is from being able to consume earlier than he could otherwise. This value is currency-neutral and it is this surplus that is split between the lender and the borrower. There are two fallacies that lead people to the opposite conclusion:

1) Thinking that a deflationary currency is extra valuable to hold and forgetting that this means it has greater spending value too because the person you spend it with gets to hold it if they want. The value of the deflation comes from the lender and is carried through the transaction to the end.

2) Forgetting that the cost of inflation comes from the expansion of the money supply. This expansion acts as a tax on all wealth since the newly-printed money can claim any wealth in the economy. The cost of the inflation acts as a claim against the lender's wealth and is also carried through the transaction to the end.

[finway]

Like this thread.

[jtimon]

But investing in producing goods is less attractive under deflationary conditions.

You are missing the second half of the picture, the "unseen" as Bastiat would say. It is less attractive to invest into some production processes, but this is offset by other business processes becoming relatively more attractive.

Maybe you can tell me what's wrong with my bakery example.

The same as I said in previous paragraph. It's only valid if you disregard the compensating effects.

Ok, I get it. The "invetment attractiveness" shifts from certain types of capital to others rather than from real capital to money-capital.
In my example, the bakery is capital that suffers with deflation. Can you make an example with a business that would be more attractive with deflation?

Although I agree that with or without deflation nominal nor real interest rates can drop below zero (aside from central banking manipulated money of course), I'm not sure I understand the theory about how investments would change with deflation in a way that's beneficial for the economy as a whole.

Well, lets first avoid the question of whether this would be beneficial. The point is that this unattractiveness of investment A will be compensated by attractiveness of investment B. I also described it somewhere on the example with apples and apple pies, maybe even in this thread.

Ok, deflation affects capital-intensive industries first (I think I got this from Hayek). What are the industries that benefit from deflation?

Other Austrians (Hayek) admit that deflation is problematic.
"It is agreed that hording money, whether in cash or in idle balances, is deflationary in its effects. No one thinks that deflation is in itself desirable."
Printing and government spending is not the answer, that's all.

Deflation is not problematic per se. It is problematic when it is sharp and unexpected by most market participants. From this point of view, it is similar to inflation. But normally, circulating money does not contract on its own. People do not suddenly get the idea to sharply switch to ascetic living style en masse. But if you have a credit expansion, then a credit contraction must follow: money supply shrinks and there is downward pressure on prices. This is exacerbated by central banking and state's interference in money. Without it, credit expansion might, hypothetically, end up in an equillibrium state in the long run.

Then you disagree with Hayek. It seems that the austrian school is more fragmented than I first thought.
My classification was the ones that believe in so called "instrinsic value" and the ones that can love bitcoin.
But it seems there's more divisions or Hayek is out of the school.

The price of money often means the interest rather than the exchange value. We can use that term but that may be confusing.
Credit, when used as a mean of exchange certainly affects the exchange value of money, just like barter or any other thing competing with money as a tool for exchange.

Yes. But if credit is performed by the creditor abstaining from consumption (i.e. no change of money supply), then this effect does not take place.

No. What you mean is money being lent. When no new money is created with the loan (i.e. the creditor holds his consumption) no monetary inflation is created.
We agree.
What I'm saying is price inflation created by people TRADING directly with debt instruments such as promissory notes (or through barter).
That medium of exchange credit would compete with money for being that, the medium of exchange.

If you have bimetalism and then you demonetize silver, the price of gold will increase.

I agree (all other assumptions being the same).

The opposite should happen if other systems can be used for trade.

All the products offered on the market represent the offer. All the monetary base at offer (not counting hoarded money) plus the bills of exchange (credit) plus the barter credits, plus LETS hours, etc. represent the demand. If that's true, an increase in barter (for example) should cause price inflation (even with the monetary supply untouched).

It is possible if there is a permanent shift between monetary and barter exchanges, for example, or credit vs. credit expansion, that this is accompanied by a change in the interest rate. But I would say that both are just symptoms of the same cause: a change in the preferences of the human actors.

What Gesell says is that these shifts happen precisely to maintain the basic interest rate constant. That basic interest that I see as a rent and you don't.
Anyway the main point here is whether those shifts affect prices or not. If they do, we can have deflation that is not caused by "efficiency and growth".

Gesell claims that the basic interest is a purely monetary phenomenon and therefore does not exist when there's no money.

Austrians would heavily disagree. There is interest even in a non-monetary economy. It's just a ratio of the value of future vs. present goods, i.e. a consequence of the time preference.

Bernard Lietaer would heavily disagree. According to him, the time preference is a consequence of the structure of money and not the other way around.
That explains why different monetary systems produce different time preferences. While capital-money produces "short-term thinking", monies with demurrage favor long term thinking.
That's the tree metaphor I posted here before.

I'll repeat it, because I like to say it at least once every few months: Long-term, predictable currency-driven deflation is essentially impossible.

I don't think deflation can stay for long neither. I would say that is fast doing its destructive job.

If I give you 10 Bitcoins today, you can hold onto them and have 10 Bitcoins next year. You also have the option of doing anything else with those 10 Bitcoins during that year if you wish. So 10 Bitcoins today must be worth at least as much as 10 Bitcoins next year -- because it's that and then some.

Long-term, predictable deflation says that 10 Bitcoins next year can be reliably and predictably more valuable than 10 Bitcoins now. That would require the option to use the Bitcoins early to have less than no value at all, which is impossible.

I don't understand the last sentence.

As for deflation's effect on loans, permit me to quote my favorite expert on economics arguing that loans and borrowing are not affected by currency-driven inflation or deflation:

The benefit to the borrower is from being able to consume earlier than he could otherwise. This value is currency-neutral and it is this surplus that is split between the lender and the borrower. There are two fallacies that lead people to the opposite conclusion:

I think that price inflation and deflation don't affect real interest too. I call it inflation premium and I substract it from the nominal interest to obtain the real interest.
Monetary inflation reduces real interest when only when the new money is loaned into existence (which is the case today). If it's spent into existence (or found in a mine) it actually rises the rates (by increasing the inflation premium).
Deflation can reduce the rates in the same way but to a certain point. The nominal rates must be always greater than zero with capital money.

But interest is not currency neutral !!
Lenders accept lower interest rates when their money suffers demurrage than when it doesn't.
That's obvious to me. But maybe Bernard Lietaer can convince you that money is not value neutral.
Different forms of money produce different values and different societies. Monies are not value neutral.

1) Thinking that a deflationary currency is extra valuable to hold and forgetting that this means it has greater spending value too because the person you spend it with gets to hold it if they want. The value of the deflation comes from the lender and is carried through the transaction to the end.

2) Forgetting that the cost of inflation comes from the expansion of the money supply. This expansion acts as a tax on all wealth since the newly-printed money can claim any wealth in the economy. The cost of the inflation acts as a claim against the lender's wealth and is also carried through the transaction to the end.

I remember your point "deflationary currency is more valuable today because it will be more valuable later".
It just seems a conceptual cheat for me.
Again, please, try to solve this problem:

Let's say we have a unit of value called stablecoin (STC) that is not really issued and it's defined as 1 BTC today = 1 STC today, 1 BTC next year = 1STC next year - btc CPI
Let's say that BTC remains deflationary for ten years (let's forget for a while that it's impossible for a while).
What's the price in STC of 1 BTC today if the deflation rate for the next 10 years is 1%?
What's the price in STC of 1 BTC today if the deflation rate for the next 10 years is 5%?
What's the price in STC of 1 BTC today if the deflation rate for the next 5 years is 1%?
What's the price in STC of 1 BTC today if the deflation rate for the next 5 years is 5%?

In the problem we've defined that 1 BTC is 1 STC for all the questions. But we're not taking into account that btc is worth more today because it will be worth more next year.
Btc1 = btc0 + (deflation_rate * btc0)
Btc2 = btc1 + (deflation_rate * btc1) = btc0 + ((deflation_rate^2) * btc0)
But wait, you're saying that
btc0 = f(btc1)
worse, that btc0 = f(btc1, btc2...)

So if
btc0 = btc0 + a * (btc1 - btc0) ?? I don't think so.

With all due respect, your claim doesn't make any mathematical sense to me.

Sorry for the very long post...

[JoelKatz]

If I give you 10 Bitcoins today, you can hold onto them and have 10 Bitcoins next year. You also have the option of doing anything else with those 10 Bitcoins during that year if you wish. So 10 Bitcoins today must be worth at least as much as 10 Bitcoins next year -- because it's that and then some.

Long-term, predictable deflation says that 10 Bitcoins next year can be reliably and predictably more valuable than 10 Bitcoins now. That would require the option to use the Bitcoins early to have less than no value at all, which is impossible.

I don't understand the last sentence.

Tell me where in this chain of logic I lose you.

1) Long-term predictable deflation means that we can know now that 10 bitcoins next year will be worth more than 10 bitcoins are worth today.

2) If you had 10 bitcoins today, one of the things you could do is hold them and have 10 bitcoins next year.

3) If you had 10 bitcoins today, you could also do other things with them.

4) Thus 10 bitcoins next year is worth less than 10 bitcoins today. 10 bitcoins today includes the ability to have 10 bitcoins next year, but also allows you to do other things if you would prefer. So it must be at least as valuable as 10 bitcoins next year. (Less negligible 'holding costs'.)

5) Since long-term predictable deflation allows us to know something that is false, it cannot be.

The rest of your argument just starts out by assuming that Value(X)+Value(Y) < Value(X) where Value(Y) > 0, so it's not surprising you get silly results. Don't start out by assuming the impossible and you won't wind up with contradictions.

[jtimon]

The rest of your argument just starts out by assuming that Value(X)+Value(Y) < Value(X) where Value(Y) > 0, so it's not surprising you get silly results. Don't start out by assuming the impossible and you won't wind up with contradictions.

Ok, I think I get your reasoning now.
But that not only implies that predictable deflation is not possible in the long run, that implies that predictable deflation is not possible at all.
According to your reasoning, the current unit would get the future value instantly just because "everybody knows its future value" and acts rationally.
I guess that a sustained predictable increase in the prices of housing is impossible too for the same reasons. If we all "know" that houses will worth more next year, prices would adjust automatically.
If oil is going to be more scarce and demanded with time and "we all know it", oil prices won't increase steadily as the production curve goes down, they will adjust suddenly the second we realize that energy supply won't grow exponentially like demand.

Is that what you're saying?

[JoelKatz]

But that not only implies that predictable deflation is not possible in the long run, that implies that predictable deflation is not possible at all.

Not quite. These aren't quite logical proofs like the way you prove two plus two is four. The assumption behind the proof is that deflation is essentially perfectly predictable. In any real-world case, there will still be some uncertainty as to whether the currency will actually deflate. And the argument only applies to currency-driven deflation, not productivity-driven deflation.

In a short-term case, at a minimum, there will be uncertainty about when the deflation will end. That's sufficient to break the argument's "predictable" assumption.

According to your reasoning, the current unit would get the future value instantly just because "everybody knows its future value" and acts rationally.

That's correct. Of course, that assumes everyone knows its future value and everyone acts rationally. If you pick cases that come close to those assumptions, the conclusion is valid. If you pick cases that drift from those assumptions, the conclusion gets weaker.

I guess that a sustained predictable increase in the prices of housing is impossible too for the same reasons. If we all "know" that houses will worth more next year, prices would adjust automatically.

No, because that can be driven by increasing demand. That's why I'm limiting the argument to currency-driven inflation.

If oil is going to be more scarce and demanded with time and "we all know it", oil prices won't increase steadily as the production curve goes down, they will adjust suddenly the second we realize that energy supply won't grow exponentially like demand.

This is again why the argument has to be limited to long-term, predictable cases. There won't be sufficient capacity to make an instantaneous adjustment. But in a long-term, predictable curve, this is correct. Speculators cause predictable future changes in commodity value to take effect much sooner than they otherwise would. If there's a continuous stream of predictable value changes (as there more or less is with oil, measured in inflation-correct currency) than speculators will always keep the price higher than it would be without speculation. (Because they are pricing in the future increase.) Of course, oil has storage cost associated with it, so the calculation is not as simple.

But this is basically correct, oil speculators build into today's price as much of the expected future value changes as they are able to until the price today gets close to the present value of the expected future prices. This is the same way currency speculators build into today's value as much of the expected future value as they can.

[jtimon]

I think I finally understand you. Your point is that predictable deflation would be eliminated by speculators arbitraging.
Fair enough. Then my my simple example gets invalidated.
Let me give it another try.
We accept that we cannot have predictable price deflation.
We have 10% monetary deflation. That is, the monetary base is shrinking. Let's keep the why out for now.
And it is expected to keep contracting for at least the next 2 years.
Someone offers you an investment opportunity. The capital you buy will have a dropping price by deflation and you will get a return of 5% (so the nominal return is shrinking with time too).
Nobody knows what the price deflation rate will be, but I wouldn't expect price inflation. Would you prefer to lend your money or to buy the real capital that is dropping in price?

[JoelKatz]

Let me give it another try.
We accept that we cannot have predictable price deflation.
We have 10% monetary deflation. That is, the monetary base is shrinking. Let's keep the why out for now.
And it is expected to keep contracting for at least the next 2 years.
Someone offers you an investment opportunity. The capital you buy will have a dropping price by deflation and you will get a return of 5% (so the nominal return is shrinking with time too).
Nobody knows what the price deflation rate will be, but I wouldn't expect price inflation. Would you prefer to lend your money or to buy the real capital that is dropping in price?

It would depend on my risk/reward preferences and my capital needs.

The reason people want to borrow money is because they get some surplus from spending money now rather than later. The value of this surplus is independent of the currency. It is this value that is split between the lender and the borrower, and the interest is paid out of this surplus.

To oversimplify, say you want to borrow money to build a factory. You will pay the money back in ten years. The benefit to you of the loan is the surplus you get from being able to build the factory now versus having to build the factory in ten years. This surplus is roughly equal to how much more valuable the goods the factory will make over those ten years are than their cost to make. This value is currency independent.

If the currency is deflating, it is especially valuable because anyone can hold it and gain value from its deflation. Because the currency is especially valuable due to deflation, sellers will accept less of it than they would otherwise for their goods and thus the borrower doesn't need to borrow as much of it as he would otherwise.

[jtimon]

Would you prefer to lend your money or to buy the real capital that is dropping in price?

It would depend on my risk/reward preferences and my capital needs.

Yes. But assuming an equal return (I guess you won't accept that assumption), say 5%. Would you prefer to lend at 5% with 10% monetary deflation or to buy capital that yields 5% but will probably drop in price?

The reason people want to borrow money is because they get some surplus from spending money now rather than later. The value of this surplus is independent of the currency.

Take into account that I disagree with this last sentence. You think that real interest rates tend to equal capital yields, but I see it the other way around.
Capital yields tend to equal real interest rates just like costs tend to equal prices, but not the other way around.
To me your mistake is in some sense similar to the mistake in the labor theory of value.
What Gesell refers to as basic interest represents a minimum that capital yields tend to be above. When a type of capital yields less than the basic interest, that type of capital is in "a bubble" (I would say that there's still demand for that producing good while its yields is greater than zero, but it's still a bubble in a capital-money based environment). 
So with capital-money (that presents basic interest), more goes to the lender and less to the entrepreneur.
With free-money (money without basic interest), more goes to profits. Since more investments can be made in the same sector while profits are positive, more investments will be made in a free-money cased market.
That's why the basic interest maintains capital yields at an artificially high percentage. Just like taxing food imports maintains land yields artificially high.
Basic interest is an economic rent (not only benefits lenders but also the owners of real capital, producing goods) and therefore should disappear.
That's the very basis of Gesell's work and I disagree that capital yields, interests and time preference are currency independent.

It is this value that is split between the lender and the borrower, and the interest is paid out of this surplus.

The way it is split is as follows: the lender gets the basic interest plus the risk premium (real interest) and the borrower gets the rest of the surplus (profits).
So even with zero risk, if your capital yield is lower than the basic interest you're losing money and your investment is a failure.

To oversimplify, say you want to borrow money to build a factory. You will pay the money back in ten years. The benefit to you of the loan is the surplus you get from being able to build the factory now versus having to build the factory in ten years. This surplus is roughly equal to how much more valuable the goods the factory will make over those ten years are than their cost to make. This value is currency independent.

The value may be independent of the currency, but I'm talking about the price. Prices drop in a deflationary environment. Both the price of the capital (even if it takes exactly the same resources to build it 10 years from now) and the prices of the consumer goods the factory produces.
If you borrow the money to build the factory and one day later everybody knows about this monetary contraction, your business plan gets screwed.
That's my main point. The same business plan is worse with deflation. Maybe it even gets unprofitable and the investment is simply not made.

If the currency is deflating, it is especially valuable because anyone can hold it and gain value from its deflation. Because the currency is especially valuable due to deflation, sellers will accept less of it than they would otherwise for their goods and thus the borrower doesn't need to borrow as much of it as he would otherwise.

You're saying you could also buy the same factory with less money, but that doesn't impede the factory to drop in price even more later.
Speculators do a job here but they're not gods and cannot price in everything instantly. Price deflation IS possible. Maybe you can't call it predictable and be accurate but that doesn't change the dynamics of deflation.

Imagine we're gods and watch the entrepreneurs in our examples from above.
They don't know the future not the price deflation rate but we do.
My baker instead of not making his investment because of deflation, will make it. Because he doesn't know about the future deflation.
But we do, and we know that if he was right on everything else he will go bankrupt because of deflation.

[99Percent]

In order for investments to be worthwhile in a deflationary economy, profits would have to be much higher and with a quicker turnaround.  The curious thing is that higher profits in the general market contributes to inflation so there is automatic feedback. The key issue then is whether deflation would spiral out of control. I don't think this would happen with bitcoins, since the money in circulation is visible and completely knowable.

Basically, we simply need to adjust our business thinking.