For seven years I’ve been trying to discuss monetary policy, and commenters (mostly at MoneyIllusion) keep wanting to bring up a completely unrelated topic—banking. Monetary policy works though the hot potato effect, and banking is no more essential to the transmission mechanism than is the auto parts industry. Print more currency than people want to hold and NGDP gets forced higher—that’s monetary economics.
Over the past decade the central bank that boosted NGDP the most was in Zimbabwe. Here’s a picture showing someone shopping for pants in 2006.
Banks played no role in the hyperinflation—you could have destroyed the entire Zimbabwe banking industry with a neutron bomb, and printing that much currency would still have led to hyperinflation. You print more currency than the public wants to hold, and they’ll bid up prices. How do you inject it without banks? Simple, pay government worker salaries in cash. Or buy T-bonds for cash. Cantillon effects don’t matter, unless the central bank is doing something bizarre, like buying bananas. A budget deficit isn’t essential either. Zimbabwe did finance its deficit by printing money, but you could just as easily create hyperinflation with a budget surplus. A quadrillion Zimbabwe dollars will do wonders for NGDP, regardless of what else is going on.
Some commenters keep insisting that “banks don’t lend out reserves.” That’s only true in the sense that “customers can never leave Walmart or Target.” If banks lend someone money, and they withdraw the funds in cash, it’s no longer called “reserves.” But that’s like saying a customer can never leave Target because if they left the store they no longer be a customer—they’d be a pedestrian.
Others say banks can’t control the aggregate amount of reserves. Sure they can; they simply lower the interest rate on deposits (or if the rate is already zero then raise the service fee) and the public will choose to hold more cash. If you made the interest rate on reserves negative 3% then almost all of the reserves would pour out of the banking system and be held as cash. In the US (in 2015) that would lead to hyperinflation, whereas in the US in 2009 or in Japan today it might lead to hyperinflation, or it might lead to lots of cash being stored in safety deposit boxes.
Negative interest on reserves is certainly not a first best policy; NGDP level targeting is far better. But if central banks really are refraining from doing the appropriate amount of monetary stimulus due to fear of an excessively large balance sheet, then negative IOR is a good way of reducing that balance sheet. But don’t make it negative 1/4%, do something that matters—negative 3%.
PS. Just to be clear, cash in safety deposit boxes is considered to be cash in circulation, not bank reserves. Vault cash owned by banks is part of bank reserves.
READER COMMENTS
Britonomist
Sep 30 2015 at 1:58pm
“If you made the interest rate on reserves negative 3% then almost all of the reserves would pour out of the banking system and be held as cash. In the US (in 2015) that would lead to hyperinflation, whereas in the US in 2009 or in Japan today it might lead to hyperinflation, or it might lead to lots of cash being stored in safety deposit boxes.”
Why would it lead to hyperinflation in 2015? Why would it not lead to people storing money in safety deposit boxes? What’s so special about having money as cash which would make people want to spend it more?
Furthermore, if everyone really did rush to get their money out as cash, it would likely cause a huge bank run, which would cause a massive financial crisis leading to a huge contraction in lending, which is almost certainly deflationary. It’s ridiculous to ignore banking, because bank credit has an overwhelming effect on the quantity of broad money, so a credit contraction is a monetary contraction, period.
Dan W.
Sep 30 2015 at 2:29pm
As Britonomist said.
One of the casualties, if not contributors, of the 2008 crisis was the Reserve Fund money market fund “broke the buck”. It actually quoted at 97 cents – a 3% loss. There immediately was a concern that other money markets would face a similar fate and, for all intents and purpose, generate a negative interest rate on cash reserves.
Did this negative interest environment, real and imagined, stimulate spending? Did it cause people to cash out their money market funds and spend the money?
No.
There is nothing at all stimulating about the prospect of one’s savings losing value.
Dikran Karagueuzian
Sep 30 2015 at 2:47pm
Scott,
Your best chance of getting through to your “banking matters” commenters is by expanding the following remark:
“Cantillon effects don’t matter, unless the central bank is doing something bizarre, like buying bananas.”
Try to imagine a strange world in which the central bank’s behavior makes the commenters’ concerns valid. (Say the central bank buys only rotten bananas with its newly created dollars, and the bananas are immediately thrown away.) Then show how different this is from what’s actually going on, with a few intermediate examples.
While I agree with your main thesis, I don’t think the exposition will be convincing to skeptics.
Scott Sumner
Sep 30 2015 at 3:20pm
Britonomist, I can’t speak for others, but if someone told me I had to withdraw $20,000 from my bank account, I would not put it into a safe deposit box. I’d “spend” it on some sort of financial assets. But of course my purchase is someone else’s sale. And in net terms it’s a closed system—base money is fixed by the Fed. That’s where the hot potato effect comes, in. The only way in the long run to get Americans to hold that much extra cash is to make NGDP rise enough so that they prefer to hold that much cash. That requires hyperinflation, except under very unusual conditions where alternative investments are not very attractive and people are willing to cram the safety deposit boxes full of cash.
You could avoid a banking panic by just having the IOR apply to excess reserves.
Dan, You are confusing a fall in market interest rates, which is not expansionary, as you say, with a fall in the interest rate on reserves, which is expansionary because it reduces the demand for base money.
Dikran, Consider these two options:
1. The Fed injects a billion dollars by paying out a billion in cash to public employees, as part of their salary.
2. The Fed injects a billion dollars by purchasing T-bonds.
At first glance it looks like the second option would put more upward pressure on bond prices. But consider that the first option implies the government would be able to borrow a billion less, because the Fed is covering part of their expenditures. So the net effect in both cases is exactly the same, T-bonds held by the public fall by a billion dollars.
Britonomist
Sep 30 2015 at 3:28pm
“Britonomist, I can’t speak for others, but if someone told me I had to withdraw $20,000 from my bank account, I would not put it into a safe deposit box. I’d “spend” it on some sort of financial assets. But of course my purchase is someone else’s sale. And in net terms it’s a closed system—base money is fixed by the Fed.”
But you’re still making a distinction between money in a bank earning 0% interest and cash, which is fine, but I don’t see how you can really make any strong prediction about how people would behave with just cash vs just money in a bank with 0% interest, the hot potato effect should apply to both.
“You could avoid a banking panic by just having the IOR apply to excess reserves. ”
But then it becomes impossible to say what effect this has unless you consider bank behaviour – will they simply pass on their costs to savers? Will they try to lend more to reduce their excess? Will they just let their excess slowly erode over time and do nothing? Will something more complex happen? What the banks do is pivotal, so you can’t just ignore it.
TallDave
Sep 30 2015 at 4:43pm
Aha! More proof that Wal-Mart is evil!
TallDave
Sep 30 2015 at 4:52pm
Britonomist —
Doesn’t matter what they want to do with the money, there’s more money and the same amount of stuff, therefore stuff is worth more money.
Ah, but the Fed has unlimited ability to lend cash to the banks to cover deposits, and perhaps must do so under FDIC.
ThomasH
Sep 30 2015 at 5:21pm
This is related to the Krugman-DeLong discussion of why “finance” types, always and everywhere, favor tighter money?
Mike sproul
Sep 30 2015 at 6:12pm
“Banks don’t lend out reserves”
Hmmm…
A bank gets 100 silver dollar coins on deposit, and it issues 100 paper receipts (called paper dollars) in exchange. Then a farmer comes in asking for a loan of $200, while offering his $500 farm as collateral. The bank prints up 200 new paper dollars and lends them to the farmer, and the paper dollars circulate as money.
Wouldn’t you admit that this bank didn’t lend out reserves?
Dikran Karagueuzian
Sep 30 2015 at 6:17pm
Scott,
Your 1. 2. comment example doesn’t get at the heart of the “banking matters” comments. Of course, it’s a fine example, and I agree with the conclusion.
Here is the best version of the “banking matters” comments I can give.
Assume the Fed purchases 1 billion in bonds from LocalBank. Previously, Local had assets of 1.1B, 0.1 B in cash and 1B in bonds, and liabilities of 1 B in deposits. Now Local has assets of 1.1B in cash, and liabilities of 1B in deposits.
Assume Local elects to make no new investments, and instead keeps 1.1B in physical cash in its vaults. (This is crazy, of course, but bear with me.)
Now consider 2 cases: 1. Local destroys the 1B in cash. 2. Local decides on a policy of never again making risky investments–they will sit on 1.1B in cash. The duration of their deposits is several years, and the depositors are unaware of the new policy.
Is inflation generated in case 1? How about case 2?
Remember, I already agree with you. I’m just trying to construct a version of the “banking matters” comments with which you can meaningfully engage.
If the “banking matters” people feel I have not been fair to their position, improvements are welcomed.
James
Sep 30 2015 at 8:03pm
Scott,
You say Cantillon effects don’t matter but they certainly seem to.
When the Fed expands the money supply in the US, they generally do so by purchasing financial assets from primary dealers. These primary dealers are all profit seeking organizations and would not pay their traders six figure salaries to engage in transactions without the expectation of profit. In order to get their primary dealers to trade with them, the Fed buys financial assets at their going market value plus some increment which is equivalent to a subsidy in terms of the economic consequences.
Do you claim that this subsidy to primary dealers isn’t actually happening? If so, then why do you believe firms are willing to act as primary dealers?
MikeDC
Sep 30 2015 at 9:40pm
It seems like banking matters for two reasons.
1. Banking is unique in a way that the auto industry is not because it IS the transmission mechanism for money. If the auto industry all failed tomorrow, people could get by on their existing stock of cars while it was rebuilt. Obviously it would be a negative shock to the economy, but life would pretty much go on as normal. If all banks went out of business tomorrow, we’d wake up and be really screwed. Everyone’s savings would be in question, billions of short-term borrowing arrangements would fail leading to millions of other business arrangements failing.
2. Isn’t it instructive that central bankers themselves seem to think there’s a relationship between banking and monetary policy?
Bruce
Sep 30 2015 at 10:13pm
TallDave,
“Doesn’t matter what they want to do with the money, there’s more money and the same amount of stuff, therefore stuff is worth more money.”
By what mechanism would the prices change? If I started counterfeiting perfect $100 bills in my basement and never spent them, how would firms know to raise prices?
Dustin
Sep 30 2015 at 10:36pm
I don’t see how “everyone withdrawal $20k” doesn’t not lead to sky high demand for money. There would be a shortage of currency, and those who are able to access some wouldn’t be so quick to let it go.
Negative IOER makes sense to me, but I’m not convinced about IOR.
CMA
Sep 30 2015 at 10:42pm
“a fall in the interest rate on reserves, which is expansionary because it reduces the demand for base money. ”
It is also contractionary in the sense that it increases the supply of base money as interest pmts are made at a lower rate from the lower interest rate.
Emerich
Oct 1 2015 at 12:47am
Scott,
One reason the Fed supposedly rejected negative interest on reserves is that it would (might?) wreck the money market. Supposedly the public would be confused by sub dollar NAVs and dire consequences would follow. What do you think of their rationale?
Scott Sumner
Oct 1 2015 at 9:32am
Britonomist, You said:
“But you’re still making a distinction between money in a bank earning 0% interest and cash, which is fine, but I don’t see how you can really make any strong prediction about how people would behave with just cash vs just money in a bank with 0% interest, the hot potato effect should apply to both.”
Cash is fairly costly to store, and I just don’t think most people want to start cramming tens of thousands of dollars into safety deposit boxes. On the other hand I freely admit that it is a judgment call, as I said in the post it’s more likely the public would behave that way in Japan than in the US (where inflation expectations are higher.)
And in any of these thought experiments the “ceteris paribus” conditions are really important, which makes it hard to have confidence in the answer (mine or anyone else’s).
Thomas, I’m not so sure about that, I’ve seen many liberals use the finance perspective. Don’t some of the MMTers look at monetary policy from a finance perspective? Having said that, yes, lots of the hawks today use finance type arguments.
Mike, Yes I’d agree if you define reserves as silver. But you’d also have to agree that sometimes banks do lend out reserves, as sometimes borrowers would ask to withdraw silver dollars.
Dikran, I think you misunderstood my claim. I’m not saying it makes no difference what people do with the money created by the Fed, I said it makes almost no difference how it’s injected, within reason. In your examples the money is injected in the same way, in both cases. All that differs is what the banks do with the cash.
MikeDC. I’m not saying banking isn’t important in a productivity sense. But the same is true of farming, water companies, or electricity companies. Without food or water or electricity we’d be in big trouble. Real GDP would fall. My claim is that banking doesn’t matter for NGDP, that the Fed can control NGDP with or without banking.
#2. No, central banks have bad judgment.
Dustin, Sorry, I don’t understand your point.
CMA, No, the Fed controls the supply of base money.
Emerich, Yes, that was their worry, but that’s really bad reasoning, for several reasons.
1. It has been applied in other economies, with no obvious problems.
2. If that were true, the Fed should have changed the regulation of money market funds so that this problem would not arise. Perhaps have the price of each share fluctuate, instead of being stuck at $1. Or maybe allow the number of shares fall over time, but keep the price at $1.
Scott Sumner
Oct 1 2015 at 9:48am
James, I meant that Cantillon effects are not macroeconomically important. The primarily dealers may make a lot of money, but its not from selling government bonds to the Fed. The commissions are tiny, utterly trivial in a macroeconomic sense. It’s a competitive industry.
paul
Oct 1 2015 at 9:53am
[Comment removed. Please consult our comment policies and check your email for explanation.–Econlib Ed.]
Jose Romeu Robazzi
Oct 1 2015 at 10:53am
Prof. Sumner, I tend to agree with you when it comes do balances, but doesn’t the banking system influence money velocity ? In your model they behave like any other agent, if you create incentives for them to sit in cash, their actions reduce velocity, if you create incentives for them to borrow and lend more, their actions increase velocity, don’t you think? Am I missing something here ? Thanks.
Tom
Oct 1 2015 at 11:40am
Scott,
If the Basel Committee scaled back Basel III signficantly and the Fed and European regulators went along with it and significantly weaker coverage ratios and stress tests were applied to US and Euro banks going forward, do you think that would boost NGDP expections? I think I know what would happen to stock and bonds markets, what do you think?
Thanks, great blogs
JoeMac
Oct 1 2015 at 12:58pm
Scott,
For years you have been saying that banks don’t matter but this is the first time, I think, you’ve gone deep into banking to explain why.
I would like to kindly request a longer a deeper analysis of these issues in a different blog post or a series of ones.
In other words, it would be useful for you to discuss the actual mechanics of real world banking with flesh and blood people making decisions, in order to explain why banking doesn’t matter in macro.
This would be opposed to your usual method of stating general principles of monetary theory to make your point.
It would also be useful if you could reference good academic papers and books that explain this, so we can study on our own.
Warm regards,
Ilya
TallDave
Oct 1 2015 at 3:12pm
Bruce — Sure, and if you discovered a gigantic gold vein that doubled known reserves, but you never told anyone, gold prices would be unaffected. But then, like your counterfeit money, you could realize no value from it, either. People generally try to realize value through trade.
Unless liquidity premia rise enough to create sufficient demand for people to hold all that new cash, people will try to trade that cash for other stuff. Since there’s no way for this to happen systemically, prices must rise.
CMA
Oct 1 2015 at 7:54pm
“CMA, No, the Fed controls the supply of base money.”
It is expanding the supply of base through IOR interest pmts. Therefore IOR is in one sense contractionary and another expansionary.
Scott Sumner
Oct 1 2015 at 10:41pm
Jose, Yes, it influences V, as does the illegal drug industry. The Fed can easily offset that effect.
Tom, Let’s say it boosted RGDP growth (which seems possible.) And lets say the Fed and ECB target inflation. Then yes, it would boost NGDP growth.
JoeMac (Ilya?) I should emphasize I mean it doesn’t matter for NGDP, unless the Fed lets it matter. Obviously banking is important to the economy, as are many other large industries like agriculture and manufacturing and health care. So the “doesn’t matter” should be viewed in that context.
My argument is that the key transmission mechanism is the hot potato effect for the monetary base, and the banking system simply isn’t essential to that process. Perhaps I’ll do another post.
CMA, I don’t understand your comment.
CMA
Oct 2 2015 at 2:14am
I am saying that IOR is in part both contractionary and expansionary.
IOR increases the supply of MB as interest payments are made, but it also increases the demand for MB.
[Nonfunctioning personal url removed–Econlib Ed.]
Jose Romeu Robazzi
Oct 2 2015 at 9:45am
Prof. Sumner
I agree that if an all powerful monetary authority wants is willing to control money and its effects it will be able to do so, no mather what. But I think that what makes the banking sector somewhat special is that it does not behave as all other economic agents. Bear with me on the following. The reason the keynesian paradigm of controlling rates works with moderate levels of inflation and rates is because in this condition, ordinary agents reaction to small changes in rates are mute, they don’t change their preferences with regard to maintaining cash balances, but the banks do react to those small rate changes: rate hikes reduces incentives for banks to lend, therefore reducing money velocity, and rate cuts increases incentives for banks to lend, therefore increasing money velocity.
Away from moderate levels of rates and inflation, the banking system behaves exactly the same, but ordinary economic agents don’t. At high level of rates and inflation, you have said that money velocity increases even more. Also, at the ZLB, ordinary agents just hold more cash if rates are cut further.
Of course, this discussion only concerns velocity (i.e. agents reaction to expectation changes). The balances effect through direct control of base money mathers too. The point is: if I am right, banking reacts possibly differently to monetary policy actions and therefore that “odd behavior” must be taken into account when designing policy.
Dikran Karagueuzian
Oct 2 2015 at 10:04am
I second JoeMac/Ilya’s call for further posts explaining this issue in detail.
The “banking matters” commenters haven’t given up after 7 years–this is evidence that the current explanations are not adequate.
There is a lot of value in a compelling exposition arguing for NGDP targeting, and if the price of “compelling” is a few posts on banking, then so be it.
Dikran Karagueuzian
Oct 2 2015 at 10:14am
Scott,
When you indicate that “it matters what people do with the cash”, you are getting closer to the heart of the “banking matters” comments.
The claim of the “banking matters” model is that the Fed interacts only with banks, and what the banks do with the cash therefore matters. In your language, the banks are a bunch of people, and what they do with the cash matters.
I admit to some confusion about your precise claim, but if this is true for a regular reader, surely more explanation is called for.
Scott Sumner
Oct 2 2015 at 10:41am
CMA, I don’t agree that IOR increases the base, in net terms. The Fed determines the level of the base through OMOs.
Jose, No, higher interest rates make velocity go up, not down. The data is very clear on that point. But that is a good example of where the finance view leads one astray.
Dikran, Another way of making my point is that monetary policy would operate pretty much the same way even if the banking system did not exist.
Jose Romeu Robazzi
Oct 2 2015 at 10:57am
Prof. Sumner, you said: “higher interest rates make velocity go up, not down”. If that is true, raising rates would be expansionary. However, of course you omitted that base money decreases even more than velocity goes up and therefore raising rates is contractionary. But why does that happen? Because banks on the margin decide to deposit at the fed, as opposed to lend out money. That is exactly the effect I was trying to describe. Isn’t it? And my point was exactly that, banks are agents that don’t behave like other economic agents.
Scott Sumner
Oct 3 2015 at 11:56am
Jose, I think you are mixing up too many things at once. Velocity can rise during period where the money supply is rising, or one where it is falling. You should look at each factor separately. Then you can think about how they might fit together.
Comments are closed.