Arnold Kling  

I Still Hate Monetary Theory

Diminishing Returns and Life... Assortative Living...

On an earlier thread, a commenter wrote:

1) What happens to the value of the dollar when new forms of money, like credit cards, are introduced? I expect you'd say that the value of the dollar would fall, but this means the credit card company is like a counterfeiter--which it isn't.

2) What happens when the Fed prints $100 and buys a $100 bond? I expect you'd say that is inflationary, but as the $100 cash is added to the aggregate spending power of the economy, the $100 bond is subtracted from it.

Start with the old quantity equation, MV = PY, where M is money, V is velocity, P is the price level, and Y is output. Define M as high-powered money, meaning the liabilities on the Fed's balance sheet.

In this framework, credit cards are not money. What credit cards do is raise V, by reducing the amount of high-powered money needed to satisfy the demand for transactions. To compensate for this higher V, the Fed might have to reduce the amount of M relative to what they would otherwise. As long as V changes gradually and predictably, this is not a problem.

In this framework, the Fed buying a $100 bond raises M. The seller of the bond does not have to feel $100 richer for the transaction to affect the price level. However, the bond purchase sets in motion a series of adjustments, the end result of which is that the price level is higher.

In this Monetary Walrasian story, it does not matter what type of bond the Fed purchases. Back in the real world, Ben Bernanke seems to think that buying mortgage-backed securities is different from buying Treasuries, and he thinks that the economy would be better served right now by having the Fed buy some MBS. That seems to harken back to the view that credit matters (it is not just money), which fits in with my Austro-Keynesian point of view.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (18 to date)
Matt writes:

I would think the fed would like to buy a basket of bonds that fairly represent the part of the economy where credit is used. The Fed assumes its banking system can do that. So, in theory, if the banking system were efficient should bank bond purchases be an accurate monetary strategy?

But, fundamentally I agree with you, credit should not be unique. Whatever shocks that cause volatility should propagate under the same rules whether it the shock applies to banking or some other industry.

Mike Sproul writes:
"What credit cards do is raise V, by reducing the amount of high-powered money needed to satisfy the demand for transactions. To compensate for this higher V, the Fed might have to reduce the amount of M relative to what they would otherwise."

So if credit cards were introduced, and the Fed did not adjust M downward, then there would be inflation? This seems to equate credit card issuers with counterfeiters. If a counterfeiter printed $20, and if that led to a $100 increase in M1, then if the Fed reacted by reducing M, there would be no inflation. But if the Fed didn't reduce M, the counterfeiting would cause inflation. So do credit card companies have basically the same effects as counterfeiters?

"the Fed buying a $100 bond raises M. The seller of the bond does not have to feel $100 richer for the transaction to affect the price level. However, the bond purchase sets in motion a series of adjustments, the end result of which is that the price level is higher. "

When you say "series of adjustments", do you mean the money multiplier? This also seems to equate the Fed with a counterfeiter: The fed prints $100 and buys a bond, and inflation results. The counterfeiter prints $100 and buys a TV, and inflation results. The Fed puts its name on the dollars it prints, and stands ready to use its bonds to buy back those dollars. How is that similar to counterfeiting?

dearieme writes:

What is the logical status of "MV = PY"? Is it just a definition of V?

Matt writes:

It is not inflation when the Fed buys a bond if over the term of the bond the Fed earns an amount that, in the limit, equals the amount of economic growth gained by whatever investment was made.

So, if it worked, the Fed, on selling the bond it bought, would collect the standard finance fee that the economy charges for monetary technology. The total value of the economy increases by the value of the investment minus the charge for taking the financial risk.

It is like renting software.

Matt writes:

In order to understand the difference between credit and money, it really does help to remove money and talk about the same situation with software.

What Microsoft and the Fed have in common is the optimum portfolio theory. For Microsoft to optimize the value of software, Microsoft has to make the software market available to every user that has software needs. The software market has to cover the software customers in proportion to their needs. If Microsoft does not do that, then its allocation of resources does not match the market needs, and the software market will begin to cycle.

The Fed has the same problem, it has to loan money, with the proper care, in proportion to everyone who uses money so that the Fed balance sheet more or less matches the aggregate balance sheet of money users.

Neither the Fed nor Microsoft can reach this goal because they are stuck in monopoly positions, and have segmented the market. Both the money industry and the software industry should cycle.

The difference between the two is that software, though a growing necessity, does not have the same widespread coverage as money, so when money has volatility all of us are somewhat affected, but only some of us are effected by software volatility.

In both cases, the Fed and Microsoft, have a hard time seeing the entire market.

Gary Rogers writes:

It seems to me that the government needs to sell a $100 bond to cover the $100 they print instead of buying a bond. Buying a bond adds money to the money supply now in exchange for a promise that it will be paid back in the future. The reason Ben Bernanke is buying bonds is to quickly add money to the supply, which compensates for all the people who voluntarily remove their money from the money supply when confidence wanes.

Gary Rogers writes:

Let me add another comment. The comment about credit cards increasing velocity was right on target. Just because credit cards do not add to the money supply does not mean they do not change the economic situation and velocity is a good way to descriibe it. In the same way, printing $100 and borrowing to cover it by selling a bond does not add to the money supply, but it does change the economic conditions. For one thing, it adds debt that must be paid sometime in the future. This is exactly what our government has been doing and it has led to export problems as well as sovereign wealth funds.

Matt writes:

Buying with a credit card is like going to the discount window, except we pay five times, at least, the interest rate.

We pay a high interest because our credit reserves are tied partially to personal guarantees, and not matched to asset value.

The reason using credit cards in small amounts is OK is that the banking system has time to readjust its portfolio according to what you bought.

In other words, the Fed cannot attempt to assertain what the investment profile of the economy until it finds out what you bought with your credit card. Like a loan done in reverse, you take the money, then later tell me what your collateral asset is.

Matt writes:

For an exampe of how credit cards go bozonkers, consider the relationship between the Fed, education investment and housing investment.

If the Fed is stuck over investing in housing, then it is likely to be under investing in education. So, the Fed would notice that the personal discount window (credit cards) tend to be used by students more than the general public (education is not fairly represented in the portfolio of the Fed)

A competitive monetary system would pick this signal up right away. Ben, even though he studied monopoly cycles as an academic, still, even he, cannot escape the lock until housing is way over produced and education under produced.

Bringing us back to, MV = PY

That equation is true only when summed over the banking monopoly cycle, about 16 years. In a perfect money market, that equation will be true when integrated over a 3 or 6 month period.

Matt writes:

So, if we look again at:

MV = PY but we add the integration limits over the cycle, then we have moderation in that we can estimate the long term value of investment even though investors are cyclic.

If we, collectively, and in particular industries, believe that the aggregate is subject to cycles, then we will estimate value over the long term and impose protections against speculators.

If we remove money from the equation then the
integral of MV = PY must have a general form for all semi-independent industries, it becomes the optimizing function for moving toward efficient markets regardless of the product sold.

For the software industry, the change in units is:

M = total software
V is the rate at which a software module is fired
P is the productive capacity of software in the aggregate.
Y is the total output, computed in units of software resources.

Bill Gates, like the Fed, ultimately wants software spread around such that it is has equal proportional utility for all of us. Bill and Ben both know they are cyclic semi-monopolies subject to speculator activity.

Physicists have a name for monopoly cycles, they are called Hamiltonian cycles, and the governing equation as the same form as the equation above, the units are force, energy, momentum properly applied.

I always have fun with the Hamiltonian physics guy and Alexander Hamilton, because Alexander was a central figure in designing the 16 year investment cycle in the USA.

Mike Sproul writes:


"What is the logical status of "MV = PY"? Is it just a definition of V?"

I don't see the use of that equation. "Money spent"="Money received". It's like saying that the amount of rain falling from the sky equals the amount hitting the earth. It's "true", but empty. For that matter, you could define M to be the quantity of GM stock, V the velocity of GM stock, P=the number of shares it takes to buy a bundle of goods, and Y=the quantity of goods bought with GM stock. The equation MV=PY would then be true of GM stock, but it has nothing to do with determining the price of GM stock.

Gary (and Arnold): In 1710, people used to say that paper pounds weren't part of the money supply. Only coins deserved to be called money. In 1845, people said that checking account pounds weren't part of the money supply, but were "money substitutes" or "economizing expedients". Only coins and paper pounds deserved to be called money. Nowadays they say that credit card dollars aren't part of the money supply, but merely economizing expedients. Only coins, paper dollars, and checking account dollars deserve to be called money.
There is a permanent float of credit card dollars that are never really paid off, and the same is true of checking account dollars and paper dollars. I predict that around the year 2100, economists will be denying that some future money deserves to be called money. Only coins, paper bills, checking accounts, and credit cards deserve to be called money.

Matt writes:

"The equation MV=PY would then be true of GM stock, but it has nothing to do with determining the price of GM stock."

Absolutly mostly correct.

That is because the auto industry has only partial coverage of the economy. If everyone used, or had a strong relationship to automobile, then one could sensibly identify the value of GM stock in proportion to the total economy, one could calculate everything in terms of proportional units of auto industry.

But, since the auto industry in not a complete covering of the economy, your price would have a large boundeed error when you tried to compute a loaf of bread in units of auto industry. A lot of people do not use cars, but they eat bread. One would be better off using unist of bread to price value in the auto industry. I would suppose that early units of money we derived from some common commodity that everyone had a relationship to.

Matt writes:

This point, by the way, of semi-monopolies, coverage of the aggregate, and credit brings up a question I should be asked.

If economic coverage is so important for stable money, then why doesn't the government handle the problem, for the state has the widest coverage?

Under this solution, prices would be in units of government, and that is precisely how it was originally done. Prior to the industrial revolution, there was no finance industry, just the king's finance ministry. The finance minister would perform a ranking of the various economic proposals before him, mercantalism. Ben does a lot of this right now.

But what Arnold and I are interested in is the how, where, and why that the government finance minister looked out his discount window and said, geez, we need an open market. Because when that concept was finally accepted, then we have a model of how industries are born and how this "growth" thing happens.

Mike Sproul writes:


The only way that MV=PY can be arithmetically correct is if Y represents the quantity of goods bought with the particular kind of money represented by M (or with GM stock). Note that this means that Y does not represent the economy's total output of goods. So if 10% of the goods in an economy are bought with M, then it could happen that as M doubles, the fraction of goods bought with M doubles to 20%, but real output of goods is unchanged. So the equation is true (but meaningless) regardless of the "coverage" of M.

Gary Rogers writes:

If Mike is still following this comment stream, I will make an attempt at answering some of his comments. First, the use of the formula MV=PY is primarily to understand what happens to price as other factors in the economy change. Problems with this formula are measurement and variable dependency, which makes it useful only for understanding relationships and direction of change. Most people have no trouble with the basic concept that more money (M) chasing a fixed amount of produced goods (Y) will cause price (P) to increase. Most people also have no problem with the concept that when more goods are produced and the money supply remains constant prices will drop. For an economy as a whole, increased prices are inflation or a weak dollar and lower prices are deflation or a strong dollar.

Velocity is needed in the equation because the same dollar is spent over and over by different people to purchase goods. Tom buys an apple from Joe, who takes the same dollar and buys a loaf of bread. If Tom buys the apple and Joe decides to save the dollar, the velocity slows down and there can be a shortage of money. This is what happens when people lose confidence in the economy and start saving instead of spending and it is the reason the Fed tries to add money when it looks like there might be a recession. Interest rates also change velocity by encouraging savings with higher rates and discouraging savings with low rates. Credit cards increase velocity by allowing us to spend money even before we receive it, but still limit us to the amount of actual money we earn, so they would not be considered part of the money supply.

The formula breaks down if you try to go the other direction. If you try to hold prices constant and increase money supply, as might happen with government price controls, the formula says that production would increase, which is not true. The problem is that production and price are dependent on each other so the results in the reverse direction are unpredictable until someone improves the model. The formula is also difficult to use because it is difficult to measure specific values like price and production for an entire economy. It is more difficult to measure something as vague as money supply and impossible to measure a concept like velocity. Mike is correct that there are many subjective decisions that determine what is included what we call money supply. However, there was a reason to only count coins back in 1700 when paper money was backed by gold or silver and paper money was basically a convenience like a credit card is to us.

This is all very subjective and imprecise, which is why I think Arnold says he hates macro economics. However, it is what drives most of our government policy and if we cannot understand it and explain the difference between good and dangerous policy, we will see the latter.

Mike Sproul writes:


I'm one of the few people who does have a problem with the assertion that if M rises, other things the same, then P will rise. If a bank issues 100 new dollars in exchange for $100 worth of goods, bonds, etc., then that bank's assets will rise in step with its issue of money, and the value of the money will not change, even if there is now "more money chasing the same amount of goods". I have a website devoted to the real bills doctrine that explains the idea in more detail than you'd probably care to read. You can find it by clicking on my name above, or by googling real bills doctrine.

Douglas Colkitt writes:

Arnold, I think you may be wrong in your assessment there. Of course the bond purchaser has to "feel $100 richer for the transaction to affect the price level."

What about the situation where the government sells the bonds to my Depression era grandparents who stuff there cash into hidden places in their house because they don't trust banks. My grandparents sell their bonds for paper money increasing the money supply, they then take that paper money and hide it in the attic, never spending it for decades.

In this case the price level wouldn't increase a single bit, because that money wouldn't be competing for anything. It still satisfies the equation because if the velocity of money entering the system is at or near zero (my grandparents cash spending isn't going anywhere, they're very cheap) then there is no effect.

The richer the bond purchaser feels the more of it is spent quickly which increases the velocity tied to it. MV is just a simplification its really a sum of the velocities of each individual monetary unit.

Mike Sproul writes:

Douglas Colkitt

In adition to what you said, there's an issue with the "Law of the Reflux"--that if a bank issues more money than the public wants to hold, it will simply reflux to the bank.

Also: You wouldn't use MV=PY to judge the value of GM stock, (M=number of shares, V=velocity of shares, etc.) because it would be a meaningless identity. On real-bills principles it is just as meangingless when applied to money.

Comments for this entry have been closed
Return to top