# The implausible Quantity Theory

Suppose the gold industry was a government monopoly, and also suppose the demand for gold was unit elastic.  Now suppose the government monopoly suddenly doubled the supply of gold.  What would happen to the relative price of gold in terms of other goods?  It would fall in half, wouldn’t it?  And that is true regardless of whether or not gold has any role as money.  We would get a sort of “gold inflation.”  Stuff would cost more in gold terms.  But not in dollar terms.

Now suppose that at some point gold was no longer used for anything other than (full-bodied) gold coins.  No more gold teeth and no more gold jewelry.  Now what happens if the government doubles the amount of gold, say from \$100 per capita, to \$200?  Again the value of gold would fall in half in terms of all other goods.  But this time prices wouldn’t just rise in terms of gold, prices would double in nominal terms, as gold is now money.  And this is true no matter how small a fraction of our wealth is held in the form of gold.

Of course you may wonder where the government gets all this new gold.  Suppose they invent an alchemy machine, and make it a criminal offense to copy this machine.  Now they can easily increase and decrease the supply of gold, and hence the relative value of gold as much as they want.

At this point you might just throw up your hands and ask; “Why not just talk about fiat money—that’s clearly what you are thinking about.”  I’m glad you asked.  What if the government doubled the amount of per capita currency from \$100 to \$200?  And suppose people typically need only a \$100 in currency to buy a week’s worth of goods.  How would the public get rid of this extra cash?  Spend it?  Yes, but \$100 isn’t very much, surely not enough to boost AD significantly, and especially not enough to boost AD to the point where the price level doubles.

If you are thinking this way, having trouble imagining how a measly extra \$100 could cause a huge rise in the price level, then you are the member of a very large club.  Let’s further explore the implications of your views.  You will now hold \$200, despite (by assumption) little or no increase in the price level.  So your real cash balances have doubled.  And why is that?  Presumably because you cannot find a way to rid yourself of this extra cash.  You can’t find a way to time your visits to the ATM so that you hold your preferred cash balances; say enough to buy a week’s worth of petty expenditures, about \$100.

Now think very hard.  Does this fit your perception of how you act?  Do you let Ben Bernanke determine how many dollars you carry in your wallet?  Surprisingly, in nominal terms the answer is to some extent “yes.”  But in real terms?  I don’t think so.  If I want to hold enough currency to buy one week’s worth of stuff, believe me I know how to arrange my life so that it happens.

Even though the QTM is nothing more than an application of simple price theory—more currency reduces the real value (or purchasing power) of currency, it is not very widely accepted.  Sometimes the reservations are for specific reasons.  Currency is an asset, so the future expected value of currency affects its current value.  Thus increases perceived as temporary won’t matter.  That’s Krugman’s argument.  Or that more cash from the government might drive interest rates to zero, at which point other assets are nearly perfect substitutes.  That’s Keynes’s argument.  But sometimes economists seem to reject this little thought experiment for no obvious reason.  They don’t look for exceptions, but simply argue that it is wrong.  That’s when I start scratching my head.  This is from Arnold Kling:

Suppose that your net worth is \$100,100, of which \$100,000 is in various assets, such as home equity, mutual fund shares, and checking and saving accounts. You carry \$100 in cash in your wallet. One day, you wake up with \$99,900 in other assets and \$200 in your wallet. By how much does your spending go up? Well, if V is constant, then PY doubles. Instead, I think that V will fall by half.

Personally, I prefer to believe that my spending rate has almost nothing to do with my wallet cash. If I have a lot of cash, I can still resist buying stuff I do not want. If I have only a little cash, I have no problem using a credit card or writing a check.

On the surface this seems to simply violate basic price theory.  Why would doubling the supply of currency cause people to hold twice as much currency in real terms?  What am I missing?  Is this how you’d behave?

1.  Classical model:  Nominal spending doubles, real spending is unchanged.

2.  New Keynesian model:  Nominal and real spending rise modestly in the short run.  In the long run real spending is unchanged and nominal spending doubles.   Transmission mechanism, extremely complex, but the “extra cash” plays almost zero role in you buying more stuff in the short run.  Instead, the expectation that the QTM will hold in the long run changes all sorts of expected nominal aggregate trajectories.  And that has a really powerful effect.

Kling also makes this argument:

According to the expectational theory backed by Aandelen Kopen Online, not only do I adjust my spending upward when I wake up with \$100 more in my wallet and \$100 less in my brokerage account, but I adjust my spending upward when I expect to wake up next year with \$100 more in my wallet and \$100 less in my brokerage account. To me, that is absurdity piled on implausibility. A theoretical version of a CDO squared, as it were.

Yes, I believe in both this absurdity and this implausibility.  Furthermore, I believe that if a alchemy machine for gold was invented, but the government insisted that retailers wait a year before stocking the machine on their shelves, the price level in terms of gold would increase immediately.  By a whole lot.

BTW, Kling also likes adaptive expectations, which says that only past inflation goes into forecasts.  During October 2008 the trailing 12 month inflation rate was still pretty high, but 12 month forward inflation expectations were plunging into negative territory.  That’s rational expectations in action.  Or consider the week in 1997 that Britain announced the BOE would be made independent.  Long term inflation expectations in Britain fell significantly that week.  Can anyone explain that using adaptive expectations?

Tags:

32 Responses to “The implausible Quantity Theory”

1. Nick Rowe
7. December 2009 at 14:50

Good thought experiment Scott! Some economists tend to think: “well, if I spend even the whole of that extra \$100, that’s not a big deal.” They don’t understand that we can’t, in aggregate, get rid of the excess money by spending it.

But I want to give you a third thought experiment:

Suppose gold is used only for jewelry, not as money. And suppose people care only about the real value of the jewelry they wear, so the demand for gold is unit elastic.

Double the stock of gold.

In the classical case, where the price of gold is flexible in terms of goods, the result is exactly as you say. The relative price of goods to gold doubles.

Now consider a Keynesian case, where the price of goods is sticky, and the price of gold is sticky too, so the relative price of goods to gold is sticky too. Will real income increase? I say no. The only thing that happens is an excess supply of gold. People try to sell gold, but can’t.

If gold were the medium of exchange, my answer to the Keynesian case would be very different: real income would increase.

2. pct
7. December 2009 at 15:38

Well, if somwhow an extra \$100 were to show up in my wallet that I did not want, I would go to an ATM, stick the \$100 in a little envelope helpfully provided by my bamk, and slide the envelope into the “deposit” slot, thereby converting the cash into a demand deposit.
Referring to the previous post, this process is much easier than if I suddenly found an extra 32 lbs of copper at my front door. That would require me to go to Somerville and find a scrap metal dealer. The bid/asked spread on converting cash to demand deposits is also zero, unlike the case with copper.

3. Mattyoung
7. December 2009 at 15:49

What happens when both buyer and sellers of goods get the same inflationary boost? Why doesn’t the seller, feeling money illusion, have less urge to move his inventory to market.

4. StatsGuy
7. December 2009 at 15:59

In so far as Kling is arguing against the excess-cash-balance mechanism as a _direct_ transmission mechanism for increasing spending, I find myself agreeing somewhat. Darn…

But I think Kling is ignoring the role of financial arbitrage, asset wealth, and risk perceptions. The _expectation_ of long term base money increases is fairly rapidly priced into assets, which gets priced into wealth. And it turns out that many households base their spending decisions on _wealth_, not just income. This is increasingly true as households are less insulated from the financial markets (due to the prevalence of HELOCs, ARMs, and the shift from defined benefit pension plans to 401ks, for example).

But even this does not account for changes in risk premia. Let us assume that households, and even financial players, are somewhat behavioral. In a sense, this is mechanically true because financial models that estimate risk use recent historical trends (and by recent, I mean last 20 years).

In other words, even if a household owns non-currency assets, they may perceive the error range of the value of those assets to be more variable than they perceived them to be in 2006. This gets reflected in increasing demand for liquidity, but in a sense it’s also true that for _any_ given level of non-currency (i.e. asset) wealth, spending could be lower because households are more risk averse. In other words, they perceive their non-currency wealth to be less real.

Flooding the markets with currency allows households to hold more “safe” wealth and increases nominal wealth even more (via asset prices), but Kling does not discuss this – the story he tells is perhaps better suited to an agrarian consumption-based economy.

5. rob
7. December 2009 at 16:04

@Nick,

“They don’t understand that we can’t, in aggregate, get rid of the excess money by spending it.”

I think Kling’s point is that he would likely not spend the money but instead put it back into one of his “other assets”–and that things would have completely reverted back to the way they were the day before. But what he fails to mention is that the \$100 dollars in his wallet is new to the economy and that whatever “other asset” he lost, be it a short term bond or an ipod, no longer exists. So if say he puts the \$100 back into the bond market, since for every buyer there must be a seller, the new \$100 will still exist and the world will have to deal with it whether it likes it or not.

PS When we say “for every buyer there must be a seller” — I think of this as holding true except when the Fed is on the other side of the transaction. The Fed is like Brahma and Shiva, creating and destroying, not buying and selling.

7. December 2009 at 16:48

Scott,
Believe or not I’m reading your entire blog from the beginning. So far I’m through March.

Today I compared the Survey of Professional Forecasters (SPF) one year in advace predictions with the fed funds target rates from 2004-2009. My first observation was that in my estimation the fed funds target was generally too low through 2006 Q1 and too high from 2006 Q2 on. In my opinion this supports the idea that the fed was not forward looking enough throughout this period.

However, there was an anomaly that muddies things up a little. I was curious if you had an opinion. In January, after the failure to lower the fed funds rate by the amount the markets seemed to expect in the previous month, the fed corrected substantially, in fact in my opinion it seems to have overcorrected. However, by early July there was evidence of falling inflation expectations and by August (after the 8/6/2008 FOMC meeting) there was solid evidence of too tight a monetary policy from SPF (8/12) projections. But they refused to lower rates at all until 10/08/2008.

I suspect that there was a certain complacency in the FOMC generated by the perception that they had been on the easy side for the first half of the year and were assuming some sort of lag in the effect of interest rate policy. I’m curious what your interpretation the fed’s interest rate policy is for this period.

7. December 2009 at 17:06

When I say January in the second paragraph I mean January 2008. Sorry for the lack of clarity.

8. ssumner
7. December 2009 at 17:54

Nick, I think I see your point, but I am not sure. Can we agree that the price of currency in terms of other financial assets is never sticky, so as a practical matter people are never cuaght holding more cash than they want for more than about a week? Or am I drifing off track.

If I understand your point that both sorts of stickiness would create monetary disequilibrium, then I agree. But are you then trying to draw inferences for the actual economy we have. (If so, is the inference that money as a medium of exchange matters?)

pct, I agree, it’s not hard to get rid of cash you don’t want.

Mattyoung, I thought it was the opposite. I thought that money illusion meant the seller was overly impressed by the high nominal price he got, and didn’t understand the real price wasn’t that great. Or did I miss something in your question?

Statsguy, You said;

“In so far as Kling is arguing against the excess-cash-balance mechanism as a _direct_ transmission mechanism for increasing spending, I find myself agreeing somewhat. Darn…”

I think you misunderstood Kling, He isn’t arguing against the excess cash balance mechanism, he is arguing against any mechanism. He is saying that doubling the money supply has no impact on the price level. That is a much more radical position. That means he is saying that people are walking around grouchy all the time because they have \$200 in their wallets, but they’d really prefer to carry \$100, perhaps because of the risk of theft. But they just don’t know how to carry smaller cash balances. He is saying that people don’t understand that if you go to the ATM less often, and take out less each time, you can adjust your cash balances.

Actually, I’m just kidding here. He probably isn’t saying what I claimed. But he did pretty clearly say that doubling the money supply has almost no impact on prices. Didn’t he? So what is his explanation?

I’m going to ignore the meat of your argument, not because it isn’t important or correct, but because until we figure out why doubling the money supply doesn’t raise prices, I’d have no idea how to address his argument, or even what his argument is. He seems to leave a market out of equilibrium that is very easy to equilibrate.

rob, Exactly right. And of course the Fed is the God of Creation (of money.)

Mark, Thanks for your interest in my early posts. I agree with your analysis entirely. I think they believed they blew it in Janauary, and then overreacted the other way in the fall. But here is the thing. I don’t think they blew it in Janauary. NGDP wasn’t rising fast, core inflation wasn’t fast, it was merely the headline CPI due to oil. But oil shocks are merely transitory if the Fed targets NGDP or core inflation. So I don’t think they needed to worry so much, and I think later events proved I was right.

9. StatsGuy
7. December 2009 at 17:58

For a non-alarmist, non-extremist pragmatic view that contrasts with Kling’s “it’s all structural” perspective…

Tim Duy’s Fed Watch:

10. Mike Sproul
7. December 2009 at 17:58

Scott:
“the QTM is nothing more than an application of simple price theory””more currency reduces the real value (or purchasing power) of currency, it is not very widely accepted. ”

Price theory applies to commodities, not to pieces of paper or computer blips that can be instantly created and retired in infinite amounts at zero cost. That’s why price theory books, when they draw isoquant/isocost diagrams, put things like labor, capital, and corn on the axes, while paper money does not appear in the diagram. Likewise, utility functions always show actual commodities as arguments of the function–not units of paper money.

It’s a relief to see that you find the quantity theory implausible, but the backing theory of money is sitting right in front of you and you reject it. The backing theory simply says that paper money is a financial security, and like any financial security, its value is equal to the value of the assets backing it. That is a very plausible theory, and it doesn’t suffer from any of the flaws of the monetarist or Keynesian view. It also doesn’t lead you to mis-apply price theory to pieces of paper.

11. Nick Rowe
7. December 2009 at 18:15

Scott: “But are you then trying to draw inferences for the actual economy we have. (If so, is the inference that money as a medium of exchange matters?)”

Yes. I’m back on my old hobbyhorse. I’m trying to persuade you that it matters that money is the medium of exchange. And opportunistically slipped my argument into your post on something slightly different. Sorry 😉

7. December 2009 at 18:26

Scott,
Thanks very much for the response. I just dug up a monetary economics paper I wrote in the spring of 2006 that argued that based on a foreward looking monetary policy that things at that point were far too tight. My professor in that course (my thesis advisor) told me only time would tell.

13. ssumner
7. December 2009 at 18:46

Statsguy, Yes, that’s a good post. Obviously, however, I put more blame on the Fed than he does (for the original crisis.)

Mike, I think price theory does apply to currency. Just because the government has a monopoly, and has a high profit margin on cash, doesn’t mean there isn’t a well-defined demand curve for cash (as a function of the inverse of the price level.)

Nick, Then you and I are the same. For me, no matter what the question about the economy, the answer is always that targeting NGDP at 5% cures all. 🙂

BTW, I left a longer reply to your comment on my other post.

Mark, I hate to say this, but I think the spring of 2006 was too soon. I don’t think policy got too tight until somewhat later. The growth in NGDP between 2003 and 2006 was fairly rapid. Some slowdown was needed. But then of course in 2008 they overreacted.

7. December 2009 at 18:58

Scott, in retrospect I agree it was far too early. Back then I was also probably little too given to easy money. But on the other hand I do think I was right in focusing on expectations.

15. ssumner
7. December 2009 at 19:07

Mark, Yes, and that’s really what’s important.

16. David Stinson
7. December 2009 at 19:14

Hi Scott,

A couple of points.

First, I am wondering whether Arnold Kling’s presentation of the cash balance mechanism is a little misleading in that the portfolio balance is a stock at a single point in time but the average money holding of \$100/\$200 is a “recurring” average stock. For example, if someone was paid weekly and rebalanced their portfolio every week, the \$100 might be a weekly average holding. In that circumstance, one would have to get rid of the extra \$100 every week, not just once. That in turn might explain why not all of the additional \$100 would be used to purchase existing assets in an effort to replenish the investment portfolio in Arnold’s example. In addition, if some of the extra \$100 was used to purchase new stock or bond issues, then presumably the funds would be directed at some point to new capital goods and therefore to a component of current NGDP.

Second, on another point, I find this confusing:

“New Keynesian model: Nominal and real spending rise modestly in the short run. In the long run real spending is unchanged and nominal spending doubles. Transmission mechanism, extremely complex, but the “extra cash” plays almost zero role in you buying more stuff in the short run. Instead, the expectation that the QTM will hold in the long run changes all sorts of expected nominal aggregate trajectories. And that has a really powerful effect.”

If there is no cash balance or “supply in excess of the demand for money” effect, what is the structural or behavioural basis for individuals’ expectation “that the QTM will hold in the long run”?

17. Mattyoung
7. December 2009 at 20:43

Money Illusion…”numerical/face value (nominal value) of money is mistaken for its purchasing power (real value). ” Says Wiki.
This one:
Experiments have shown that people generally perceive a 2% cut in nominal income as unfair, but see a 2% rise in nominal income where there is 4% inflation as fair,

I wonder if the experiments showing the deluded subjects buying more than they should, and also selling less than they should. Money Illusion must imply a net asymmetry in trade? I think it does, but I have not explored this.

18. scott sumner
8. December 2009 at 07:28

David, That’s true, but even if cash turned over very slowly he would still be wrong. The reason that more cash makes the price level go up has nothing to do with the fact that more cash makes people feel richer, and therefore go out and spend more. But that’s what Kling seems to assume people are arguing. The reason is because more cash makes each unit of cash worth less. It’s basic economics 101.

Regarding your second question, I was trying to say that the extra cash doesn’t motivate people to go out shopping, it motivates people to hold less cash. They can also hold less cash by buying financial assets. And in the short run this is a much more powerful effect than the shopping channel. I guess if you defined “stuff” to include buying stocks and bonds and time deposits, then you would be right.

I think the reason why people do buy more stuff in the short run is that some prices are sticky. If there were no sticky wages or prices than a nominal shock would not create any excess cash balances, prices would simply rise in proportion to M immediately (ignoring distributional issues.) But in the short run (in the Keynesian model) real cash balances rise almost in parallel with nominal cash balances, as asset prices rise to restore equilibrium in the money market.

Mattyoung, I agree with that defintion. It implies that during unexpected inflation people buy more than they should, and vice versa for deflation.

19. Doc Merlin
8. December 2009 at 08:28

Hrm, that gets me thinking. So you had a 2% rise in income with 4% inflation, that Mattyoung talks about, long enough, and the value of your income didn’t match up with the expected value. Would you have a catastrophic correction as you rushed to change your spending habits after finding out too late what your real income was? How aggregate this over the whole economy.

Thanks Scott, Mattyoung, you two have made the best argument for the Hayekian trade cycle I have ever seen. Heh, an argument for the trade cycle from the standpoint of the money illusion, will wonders never cease.

20. David Stinson
8. December 2009 at 10:39

Still confused. Sorry for belabouring this.

I probably wasn’t clear. I didn’t mean to suggest that people spent excess cash necessarily because they felt richer. The point I was making about Kling’s point was that the excess supply of money was greater than his example might be interpreted as implying because it is recurring.

You said:

“The reason that more cash makes the price level go up has nothing to do with the fact that more cash makes people feel richer, and therefore go out and spend more. But that’s what Kling seems to assume people are arguing. The reason is because more cash makes each unit of cash worth less.”

And:

“I was trying to say that the extra cash doesn’t motivate people to go out shopping, it motivates people to hold less cash.”

I guess my original question with respect to your explanation of the New Keynesian approach might be restated as why, if the assumption is that the real balance effect plays no role, do people in the New Keynesian models believe that “more cash makes each unit of cash worth less”. What is the mechanism that underpins people’s expectation that QTM holds in the long run, if not an understanding of individuals’ reactions to excess cash balances? Or is the belief based simply on the equation of exchange?

21. Scott Sumner
8. December 2009 at 17:40

Doc Merlin, People should respond by demanding higher wages, and this should increase unemployment back up to the natural rate.

David, I didn’t explain myself well. I do think that in the long run the excess cash balance mechanism explains why prices rise in proportion to M. My point was that the initial effect of more money is due to its effects on asset prices, relative prices, etc. These effects drive AD and inflation higher. Once prices have risen as much as M, then there is no more excess cash. So I do think it explains the long run effect. But the initial increase in AD, in an economy with sticky prices, is driven more by changes in asset prices, than by the fact that people have a bit more money burning a hole in their pockets.

Perhaps my distinction isn’t important. Maybe at a basic level the excess cash is doing all the work. But somehow I find it more plausible to think in terms of the asset price story.

On the reoccurring cash balances issue, you’d have to ask Nick, as he is the one that emphasizes that issue.

22. Greg Ransom
10. December 2009 at 14:51

” Now suppose the government monopoly suddenly doubled the supply of gold .. ”

Who gets the stream of new gold first and what do they do with it and who gets the money next after them.

This is Hume / Cantillon / Mises / Hayek / Garrison economics 101.

Money expasions have real effects on relative prices and real goods.

23. Greg Ransom
10. December 2009 at 15:01

Who gets the money first matters.

Say Fred gets it first and buys up corn
to make gas. I don’t get the money but I need more cash in my account to cover my purchase of corn for cornbread. I have to expand my cash holdings not because I have more money but because any single purchase demands greater cash holdings for a successful transaction.

24. ssumner
11. December 2009 at 07:36

Greg, I don’t think the “who gets the gold first” question is that big a factor in its overall macro impact. Here is something to think about:

In the limit, as the MB approaches zero, or as the gold stock approaches zero, the QTM remains equally true. Thus in some countries the MB is only about 2% of GDP. But doubling the base will still double the price level. On the other hand if the base is only 2% of GDP, then the question of who first gets the money seems less important. I would add that those who get it first aren’t getting a windfall, they exchange bonds for cash, or bonds for gold. Then the extra cash may be placed in banks, where it spreads to everyone else. I understand that many Austrian’s disagree with me on this, but I think they’re wrong.

I do think the Austrians are on to something important with the “relative prices” issue. I think a key impact of expansionary monetary policy is that it lowers the relative price of labor and raises the relative price of assets like stocks, real estate and commodities. So I think there is some overlap in our views. The difference is that I believe those relative price effects would still occur if the new money was distributed equally across the population.

25. Greg Ransom
11. December 2009 at 13:16

Is it really your claim that this money is “spread” equally and with no effects on leverage or the structural shape of demand?

Scott wrote:

” Then the extra cash may be placed in banks, where it spreads to everyone else.”

26. ssumner
11. December 2009 at 17:10

Greg, Not spread equally, but I don’t think the unevenness has much impact on AD. Perhaps there is some impact on bank lending, I’m not ruling that out, but I really think the main impact works through expectations and asset prices. I think the distributional impacts would be more significant if the cash were handed out to people. But it is sold in exchange for T-securities. So no one is richer directly because of the distribution. Of course if an expansionary monetary policy changes asset prices, that might make someone richer. Thus if inflation expectations soared then gold prices might rise and gold investors would be richer. But I don’t see that effect as being due to the specific distribution of cash, rather due to its effect on asset prices. Does that distinction make sense?

27. Lifting up the inflation debate « Freethinking Economist
17. December 2009 at 13:38

[…] is where I tend to get 90% of my views.  If you have time, enjoy the discussion between Scott Sumner and Kling, and then Kling’s answer.  Both are right wing, hating fiscal policy, but Sumner is […]

28. Michael F. Martin
17. December 2009 at 20:01

As usual, for me the question is about the frequency of different economic activities. The difference between the two scenarios you begin with is nominally elasticity, but it’s not the elasticity that matters so much as the rate of diffusion of “wealth” from persons holding gold into other currencies –if the change in the supply of gold happens too much faster than the rate of diffusion, then there will no longer be enough consensus on the value of gold for it to be useful as a medium for exchange in increments of widely agreed-upon value.

In effect, the exogenous shock desynchronizes expectations about the future value of gold.

29. ssumner
18. December 2009 at 07:41

Michael, I agree that gold isn’t a very good medium of exchange. I am not sure I followed the rest of your argument.

30. Michael F. Martin
18. December 2009 at 23:11

Here’s another try.

Assuming we are not in a perfect global equilibrium, the value of every currency will change over time within a given locality. I called the percent rate of change in value per unit time “the rate of diffusion” because the value is either increasing or decreasing as economic activities either concentrate or diffuse the value of the currency within a given locality.

There are two-different sets of mechanisms at work here. One is typically short-term and the other long. But my point is that the usually long-term mechanism can occasionally affect the short-term dynamics in a very bad way.

Over the short-term of hours, days, and weeks, the price of a currency (measured against a basket of others, within a given locality) will fluctuate due to temporary shortages and gluts in the inventory needed to complete transactions. In effect, agents are letting each other borrow small increments of value to complete well-understood transactions, knowing that the value of the currency isn’t going to change much over the short-term.

Over the long-term of (usually!) months and years, the price of a currency (measured against a basket of others, within a given locality) reflects the comparative value of the resources, goods, and services of the locality relative to others’. At any given moment, every market participant is clamoring to obtain additional units of currency, but is successful only when the participant has resources, goods, or services valued more highly than the next best alternative available. Over the short-term, these dynamics may not be so obvious because the “rates of diffusion” tend to cancel to within tiny percentages. Only over longer time periods to these tiny differences in the rates of diffusion add up. It took China a decade to build up its reserve of U.S.\$ That’s not long in absolute sense, but it’s an example of what I’m calling the long-term dynamic here.

But what happens when something or somebody disrupts the long-term dynamic? Let’s say, for example, that China were to remove its currency peg. Or that immense veins of gold were discovered in the mountains of New Zealand.

Short-term dynamics break down because agents are no longer willing to make even small loans to keep transactions going. Agents have no clear expectations of how the “long-term” dynamics affect their risk. The long-term is — in these situations — short-term.

What the thought experiment reflects is that the value of any currency is based on a consensus of market actors about the stability over time of the currency. More specifically, the consensus about stability reduces latency in the flow of resources, goods, and services among buyers and sellers.

31. ssumner
19. December 2009 at 11:58

Michael, Most of that makes sense, but I am struggling a bit with the end. We know that there is no currency peg between the dollar and many other currencies. And we know that those currencies often fluctuate by one or two percent in a given day. But those fluctuations are mostly unexpected–the forward markets predict little movement day to day. So is this really a big problem for people who do business between the countries with floating currencies?

32. æ€ªã—ã„è²¨å¹£æ•°é‡èª¬ by Scott Sumner – é“è‰
8. April 2011 at 23:29

[…] by Scott Sumner //  æ˜”ã®è¨˜äº‹ã§ã™ãŒScott Sumnerã®ãƒ–ãƒ­ã‚°ã‹ã‚‰ã€”The implausivle quantity theory“ï¼ˆ7. December […]