Thursday, October 25, 2012

Reserves before the Crisis


Total Reserves relative to GDP:

Graph #1: Total Reserves as a Percent of GDP

Falling. From about 3¾ cents per dollar's worth of GDP in the late 1950s, to less than half a cent in the 2000s. After that came the big spike that was the Fed's response to the financial crisis. (Only the very beginning of that spike is shown on this graph.)

I want to point out a slight high spot in the downtrend, a double-hump that is over and done with by 1975. Nothing to do with Paul Volcker.

And I want to point out the hump after 1985 and another one, crossing 1995. These two humps I have discussed before.

(The humps are not part of the current topic.)


Same graph, multiplied by prices instead of by 100:

Graph #2: Start with Graph #1 and Multiply by the Price Level
Graph #1 is all downhill. Graph #2 is a lot more horizontal, until 1995 anyhow. Almost looks like the Fed could have been using Graph #2 as a guide to keep reserves in a target window.

But Graph #2 is of course identical to Graph #3, where instead of dividing by GDP and multiplying by prices, we reduce two steps to one and divide by so-called "real" GDP:

Graph #3: Total Reserves as a Percent of Inflation-Adjusted GDP
Because inflation remains in the numerator but is removed from the denominator, the plotted line is reluctant to fall. The inflation remaining in the numerator on Graph #3 pushes the line up. The only difference between Graph #1 and Graph #3 is that Graph #3 has inflation removed from the denominator.


If the Federal Reserve was using one of these three graphs as a guide for setting the growth of Reserves, it would have been Graph #3, where the victim is divided by a denominator that has inflation stripped away, so that the inflation which remains in the numerator distorts the resulting values by pushing them up.

In the case of Graph #3, the distortion pushes the Graph #1 values up just enough to keep the plotted line relatively flat until 1995 -- and again, from 2000 to the crisis.

But let me ask a question.

I don't know what is the "best" level of reserves. But there must be such a level. For the sake of argument let's say the "best" level is the first value shown on the graphs.

You could pick any level. For me, the first value is the easiest to think about. And it is the same amount of reserves at the start, no matter which graph we look at. The same amount at any moment, no matter which graph: The y-axis numbers are different, and the plots are different, because we're dividing by different things. But the amount of reserves is the same on all these graphs no matter what the reserves are divided by.

So here's the question: What is the right thing to divide by?

We want a level of reserves that stays more or less in proportion to the quantity of money, and a quantity of money that stays more or less in proportion to GDP. So reserves should stay more or less in proportion to GDP. But which GDP? Nominal or Real? Actual-price GDP, or inflation-adjusted?

The denominator should be actual-price GDP, like Graph #1. Why? Because the amount of money we need depends on the actual purchase price of GDP. If prices go up, then we need more money to buy the stuff. This is true, no matter the cause of inflation.

Reserves must remain in proportion to actual price GDP.

Graph #1 shows a long-term decline in reserves, as total reserves increased more slowly than actual-price GDP. For all that time the quantity of money was restricted: Reserves were restricted to fight inflation. For all that time, actual-price GDP increased faster than reserves, forcing the line down. The result was we had less and less money to buy the stuff we were producing.

If inflation was the result of too much money -- or too much reserves -- the decline of reserves would eventually have stopped inflation. Somewhere along the path from almost 4% of GDP to 2% of GDP to 1% of GDP to less than half a percent of GDP, the decline of reserves would have brought inflation to a halt.

But inflation never did come to a halt. The quantity of reserves continued to decline relative to GDP until the crisis, but the decline never brought inflation to a halt.

If "too much money chasing too few goods" was the problem -- if too much reserves was the problem -- then the decline of the line shown on Graph #1 would have fixed the problem, and eventually the line would have gone flat.

The line never went flat.


Here's another look at reserves as a percent of actual-price GDP, like Graph #1. But this time the graph also shows the Fed's response to the financial crisis:

Graph #4: Total Reserves as a Percent of GDP (thru September 1, 2012)
From the 1950s to the crisis of 2008, the Federal Reserve responded to inflation by reducing reserves relative to GDP. The Fed did this continually, repeatedly, over and over, no matter how low the level of reserves fell.

But then there was the financial crisis, and suddenly they changed their mind.

Here's the thing: If the quantity of reserves (or the quantity of money) was the true cause of inflation, then reducing reserves relative to GDP (as on Graph #4 before 2008) would have been the right thing to do. It would have worked. Sooner or later, it would have stopped inflation.

But it didn't work. And now we no longer even try to stop inflation. We just try to keep inflation stable. This is an extraordinary failure of policy. The failure shows that the quantity of money and reserves was not the true cause of inflation.

The Federal Reserve reduced reserves continuously relative to GDP for half a century. The policy did not stop inflation. It only stopped the economy.

16 comments:

Dan Lynch said...

Fed does not control or target the QUANTITY of money, only the price (interest rates).

Banks can create money regardless of reserves. Loans create deposits and do not draw down reserves. If a bank does need more serves, they can borrow them from the Fed's discount window.

The Arthurian said...

so i have heard.

okay then, suppose we leave the central bank out of my story. suppose the things that happened, happened endogenously.

the quantity of reserves still went down relative to gdp. and (assuming some stability in the ratio of reserves to money) so did the quantity of money.

so then, how come prices went up?

Dan Lynch said...

Debt went up -- money created "out of thin air" by bank loans and credit cards.

Home prices went up thanks to the debt bubble.

Commodity prices went up thanks to unregulated speculation (thanks, Clinton!) and China.

Set aside housing, commodities, and health care, and there has been little inflation, which stands to reason since there is significant slack in the economy outside those areas.

The Arthurian said...

"Debt went up"

Good answer, Dan!

Let me go back to your previous comment:

"Fed does not control or target the QUANTITY of money, only the price (interest rates)."

I think it is accurate to say that the Fed targets interest rates, but controls the quantity of money in order to do it.

Also I think this view is oversimplified:
"Banks can create money regardless of reserves."

I think the decline of the reserve requirement is one of the things that allowed private finance to overpower policymakers. I think such things should be remedied. I think the decline of Fed control is the thing that makes it true that (as you say) "Banks can create money regardless of reserves." But then, it is wrong to look at the results and pretend things were always thus.

Financial innovation be damned. :)

Dan Lynch said...

"the Fed targets interest rates, but controls the quantity of money in order to do it."

Paul Volker attempted to control the quantity of money and he failed. Or more accurately, he could control quantity, or he could control interest rates, but he could not control both at the same time.

Since then, the Fed has targeted interest rates and let quantity float.

Agree that private finance can definitely "overpower policymakers," especially when policymakers are asleep at the wheel. Throw in international markets which are beyond US jurisdiction but nonetheless impact our economy.

The Arthurian said...

"Paul Volker attempted to control the quantity of money and he failed."

Yeah, I know. The history of Fed policy fascinates me... but I don't have a lot of luck finding a lot to read!

My point was, in order to "control interest rates" the Fed buys or sells stuff -- which means they are in fact changing the Q of M in order to influence rates. So while interest rates are the target, the Quantity of Money is the method.

In Money Mischief (Chapter 8) Milton Friedman said the Fed CAN control the Quantity of Money but that it CANNOT control interest rates. He doesn't explain why. I have come to think he meant the Fed alone can decide on the number of dollars to put in or take out, but that interest rates depend on the interaction of buyers and sellers. So it isn't the Fed alone that controls interest rates, but the Fed in context of the demand for credit and the state of the economy and all of that.

It's the only way I could make sense of what Friedman said there, and fit it in with what I think.

I guess I'm talkative today. I love the topic.

Greg said...

I like this way of looking at it Art

When a bank makes a loan to you they do not look and see what THEY have in reserve, theu look and see what YOU have in reserve.

They dont look in their own vault, they look in your vault!


Loans first, reserves form the fed later.

The post you did the other day about balance sheets and snapshots is applicable here I think. Suppose a bank is in balance on Monday 0900, they,ve attained all the necessary reserves to meet the Feds requirements form last weeks lending activity.

At 1100 they make a 10 million dollar loan to a local developer. They have till the next Monday to get the reserves so on Tuesday it will "look" like they are violating their reserve requiremnts because their loan side of the balance sheet is up 10 milllion but the 10% reserves arent there yet. The snapshot is missing the pending acquisition of reserves. It doesnt happen instantaneously, its not an automatic process. At any paeticualr time a banks reseves and loans will likely not be within the perfect 10% ratio, thus the mythical multiplier might appear to be changing.

Greg said...

Regarding the control of money supply and the fed think of it this way. Controlling the quantity AND the interest rate means NO MARKET. Its simply a command economy type system. So in order to have a market the monopolist must control one variable and let the other float via the market forces.

The Arthurian said...



Yes, exactly!!! The Fed cannot control both, not and still keep the kind of economy we have (a market economy I guess). The Fed does things to the quantity of base money or to the quantity of reserves I think, and the economy responds to that, and the Fed looks at the change in interest rates, and based on interest rates decides whether to add more or not.

I say, therefore, that the Fed CONTROLS the quantity of money, using its interest rate TARGET as a guide.

This is probably a very minor detail. But if people simplify just a bit too much, they say things that I believe to be incorrect -- like: the Fed does not control the quantity of money.

And when people generalize just a bit too much, and come up with generalizations that are incorrect, they can end up being as wrong as Mankiw.

Dan Lynch said...

No, the Fed controls the price of money. The quantity is controlled largely by banks.

Sorry to tell you this, but Milt Friedman lied.

The Arthurian said...

Dan, Blogger thought your comment was spam for some reason.

Milton Friedman perpetrated the greatest scam of the 20th century, but not every little thing he said was wrong.

As it happens, I can now justify in my own mind the thing he said (CAN control quantity, can NOT control price). So am I lying too?

But Dan, since you insist that the Fed controls the price of money, tell me HOW they do that without controlling the *quantity* of reserves.

They pick an interest rate *target* but the arrows they shoot are called reserves. Or, "money".

The Arthurian said...

Greg: "Loans first, reserves from the fed later."

Why is this such a fetish with so many people?

If anything, the fact that the Fed allows banks to plus-up reserves "later" is an indication that the Fed goes out of its way to be good to the banks. Another example: the reserve requirement is almost unfailingly lowered, not raised.

But whether the Fed is too good to banks is not a substantive economic issue. One would expect banks to be always squeaking their wheel in that direction.

The Fed goes along with it because they think it might be good policy. They went along with it for so long that now people say (for example) that reserves serve no purpose. It's not true at all. Reserves serve a great purpose.

We just don't use them for that purpose. There is a difference.

Greg said...

So what purpose do reserves serve?

Dan Lynch said...

Greg, reserves are a relic of the gold standard and serve no purpose in the current fiat monetary system. Canada has zero reserve requirements and their banking system seems no more screwed up than ours.

Arthur, deficit spending creates excess reserves in the banking system. When banks buy bonds, that draws down reserves. Not coincidentally, the Treasury sells bonds equal to the amount of deficit spending.

If the Treasury stopped selling bonds, then excess reserves would pile up and the Federal Funds rate would plunge to zero.

As you know, the Fed manipulates interest rates by buying and selling treasuries on the secondary market, enough to be a "market maker."

The Fed also has the power, though it seldom uses it, to control interest rates by paying interest on reserves. In theory, the Fed could cease playing the bond market and control interest rates solely by setting the interest rate on reserves.

Greg said...

Hey Dan

Thanks for the response but the question I asked was directed at Art. I dont think he sees reserves the same way I/we do. I have a similar understanding of the purpose of reserves and would have answered the question much like you did. The only thing I would add is that I dont think reserves are something they just represent something. In other words reserves dont get lent or spent they are a reflection of lending and spending in the system. They are sort of a "dollar doppleganger" as I understand them.

The Arthurian said...

Dan: "Fed does not control or target the QUANTITY of money, only the price (interest rates)."

Sumner:
Because the demand for money is unstable, and because there are lags between changes in monetary policy and observed changes in NGDP data, most central banks rely on intermediate targets like interest rates or exchange rates. But these are still targets, monetary policy consists of control of the money supply...
Interest rate targeting (or exchange rate targeting) is simply a tool. It is changes in the money growth rate that actually drive macro nominal aggregates up and down.


Me: "I think it is accurate to say that the Fed targets interest rates, but controls the quantity of money in order to do it."

Me, again: What Sumner said.