How bad banks create money (Reply to Nick Rowe, pt 2.)
I thought I would take another shot at Nick Rowe’s recent post on bad banks. I was motivated to do this by his comments in response to some of my recent posts, which made me take a second look at what he wrote. Banks are not my speciality, so take these comments as provisional, as a part of the ongoing conversation.
Nick tries to simplify things by imagining a world with no currency held by the public. I am not too fond of that assumption, nor do I quite agree with his view that (in the modern world) currency is much less important than checking account balances . I agree that checking balances are bigger, especially if broadly defined to include MMMFs, but until last year most of the base was currency held by the public, and I still think that is important. Nevertheless, for the sake of argument I am going to go with Nick’s assumption of M1 being 100% DDs.
Now let’s suppose the Fed does an open market purchase of $1000. They buy a bond from the public. The Fed gives the bond seller a $1000 check, which is essentially base money, and the seller deposits that check in her bank account. So the money supply goes up by $1000. And this is even true if the traditional money multiplier process is broken. It is true even if banks make no more loans, because of financial system problems associated with the financial crisis. (These might include a lack of bank capital, or poor loan prospects due to recession.) But this interpretation seems to contradict Nick’s view expressed here:
Central banks got rid of their fleet of helicopters when currency got replaced by demand deposits as the primary medium of exchange (again, I admit, I’m unclear on where currency fits into my picture). They contracted out helicopter operations to the commercial banks. They wanted the commercial banks to create the excess supply of money at current levels of income, leading people to want to spend that excess so we can escape any general glut. But now the banks’ helicopters won’t fly, because banks lack the capital to expand loans.
It ain’t the loans what matter; it’s the money creation that goes along with loan creation what matters. That’s why (bad) banks matter. Bad banks won’t create money.
I’m confused, because you don’t need more loans to get more money, you just need more reserves. If the Fed creates new money, only one of the following three things can happen:
1. Currency held by the public increases.
2. Bank deposits increase.
3. Reserves increase but deposits are unchanged. The reserve increase is offset by a fall in the money multiplier.
In the first two cases the money supply increases. In the third it might not. But even then the “worst case” is that the broad money multiplier falls to one. At that point all bank assets are reserves, and bank liabilities must rise one for one with central bank injections of reserves. (Of course the narrow M1 multiplier could fall below one.)
So do bad banks matter? Well I suppose bad banks matter in the obvious sense that they can impact the money multiplier. They can affect the amount of base money the central bank must inject to hit its target for expected inflation or the exchange rate or NGDP or whatever. But other than that obvious factor, I don’t see why bad banks matter. They certainly don’t stop central banks from increasing bank liabilities, even in a world with no currency. And this action will depress the value of bank liabilities relative to other assets like stocks. And this boosts AD. If the Fed is targeting the forecast of its goal variable then the amount of reserves in the banking system will simply be one more endogenous variable in the economy. So I don’t see why bad banks are important. But maybe I am missing something.
PS: If you are worried that the central bank might have to inject a whole lot of reserves, I can put your mind at ease. Reserves are profitable for central banks. And even if you accept Krugman’s implausible argument that high reserve levels expose the central bank to excessive risk, the problem can be easily eliminated with a small interest penalty on excess reserves.
Tags:
8. December 2009 at 09:03
The IMF is talking exit strategies too, but at least says we should only be Planning them, rather than Enacting them at present.
8. December 2009 at 10:09
The central question with bad banks is whether they are capacity constrained (the justification for recapitalization) or demand constrained (they can’t find good opportunities for loans because the economy is operating at a sub-par equillibrium). Another version of capacity constraints is the argument that banks will make more loans because they just feel better with more money in their pockets (e.g. reserves).
If they truly were capacity constrained, then TARP should have fixed that. If they truly were demand constrained, then monetary action should have been effective. So, we can just look at the market’s response to TARP vs. the market’s response to QE and other broad liquidity programs. It’s pretty definitive.
But Rowe’s arguments about the money injection mechanism are still material – there is still the equillibrium problem, which is to say, why should banks lend? One way of framing this is as an increase in risk premia or a demand for collateral for loans (both of which decrease the money multiplier). In that sense, banks may still be capacity constrained because of new lending practices (which should have always been in place), but the proper response to this is not to force more loans with the same currency base, but to directly increase the currency base to permit more stable loans that meet strict lending standards. In other words, the economy needs more currency (or cash – I don’t follow you and BIll W on that distinction, really).
This is why when Kling makes the argument (recently) that when the government prints money and spends it on goods/services, that is FISCAL action, he’s plain wrong. It’s both fiscal AND monetary action. It’s directly impacting the level of permanent money in the economy.
Almost certainly, there is an optimal balance between injection of money through banks and through fiscal deficits that are financed by printed money. If banks are sitting on high excess reserves (however you define that) then there is a certain point at which printed-money-financed govt spending will get them to start using those reserves (either by creating new opportunities, or by creating an inflation penalty, or both).
Cramming banks with liquidity and then penalizing them for holding liquidity is a purely non-fiscal mechanism to accomplish this, and if you truly distrust govt. usage of money and are not concerned about systemic stability, it’s obviously tempting.
So who do you distrust more, big finance, or big government? Hmmm….
8. December 2009 at 10:23
Scott: I’m very pleased to see you have another crack at this topic. I confess banks are not my specialty either! I’m trying to get my head straight on it all.
Your case 3 (“3. Reserves increase but deposits are unchanged. The reserve increase is offset by a fall in the money multiplier.”) is the one I’m worried about. It’s the worst case scenario. (And therefore is the one we ought to be worrying about).
Funnily enough, I had an earlier crack at this topic back in April 2009. The conclusion I came to there was sort of similar to the conclusion you come to here. Look at the last few paragraphs:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/04/bad-banks-and-the-effectiveness-of-fiscal-and-monetary-policies-on-ad.html
This is the thing you say that I’m not sure on:
“But even then the “worst case” is that the broad money multiplier falls to one. At that point all bank assets are reserves, and bank liabilities must rise one for one with central bank injections of reserves.”
But the bank liabilities might not be demand deposits. They might for example be liabilities to the central bank.
If the central bank buys a $100 bond (or buys $100 of anything) directly from the public, the public then deposits that $100 in a bank, and deposits expand. Multiplier of one.
But if the central bank buys a $100 bond from a bank, or just lends the bank $100 more reserves, the bank could just let the reserves sit there (if it is capital-constrained, for example). Multiplier of zero.
I can’t figure it out.
8. December 2009 at 10:41
Just found my May post, on the same subject: http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/05/quantitative-easing-circumvents-banks-capital-constraints-to-increase-m1.html
This is embarrassing. It shows how I am going round and round in circles attacking this same question.
8. December 2009 at 11:12
Off topic – but:
If employment does not improve soon here is where we will be headed.
http://www.ft.com/cms/s/0/c6683864-e398-11de-9f4f-
00144feab49a.html
Tariffs can persuade Beijing to free the renminbi
By Robert Aliber
Published: December 8 2009 02:00 | Last updated: December 8 2009 02:00
And this man is from the U of C
8. December 2009 at 11:14
Link didn’t run – here it is again
http://www.ft.com/cms/s/0/c6683864-e398-11de-9f4f-00144feab49a.html
8. December 2009 at 12:23
It seems to me that bad banks do matter (but I don’t claim to know much about this).
Let us assume a sudden decline in V because all the banks go bad. The central bank wants to increase M to compensate, but if it attempts to do that by injecting money into the banks that does nothing to increase the money in the hands of the public. In order to increase M in this situation the central bank would have to either by-pass the banks or fix them.
8. December 2009 at 12:33
Statsguy- Your analysis fits my view, fiscal stimulus with an accomodative Fed (up to a point, of course). Part of the problem in doing this is the inherent uselessness of the ‘dual mandate,’ which means both virtually anything and nothing. Therefore, the Fed is paralyzed to act because even a hint of monetization strikes the fear of Zimbabwe into some while banks starve small businesses for loans out of fear of deflation. A clearer mandate, such as Sumner’s NGDP target, should actually allow the Fed greater freedom of action.
I’d also gladly concede that a clearer mandate and policy expectations should make both less fiscal and monetary easing necessary.
Somewhat on topic, Sumner, I don’t know if you noticed this piece by Moritz Schularick and Alan Taylor:
The credit system seems all too capable of creating its very own shocks, judged by how successful past credit growth performs as a predictor of financial crises. Past credit growth spectacularly improves the forecasting power of an early warning banking crisis model in our data. Using long-run historical data, the growth rate of lending emerges as the single best predictor of future financial instability, a result which is robust to the inclusion of various other nominal and real variables. Moreover, credit outperforms other possible measures such as broad money by some margin.
http://www.voxeu.org/index.php?q=node/4349
8. December 2009 at 13:44
After reading this blog I’m confused…
The entire premise is that a fall in the price level has led to a drop in output due to sticky wages.
Yet the core CPI is at a record high. And the total CPI is only 1.7% below its peak. If expectations are correct we’ll be back to peak in a year. And I’m sure consumers and businesses are smart enough to smooth food/energy prices so it is probably core that really matters.
All this talk of restoring the price level strikes me as ridiculous. We are at record price levels and market expectations are for continued 1-2% inflation.
Such a small drop off the peak cannot even come close to explaining 11% unemployment. The entire premise is flawed.
Where am i wrong?
8. December 2009 at 15:03
Scott and Nick:
Break out the trusty T-accounts. The central bank CB starts with $100 in coins deposited, for which it issues 100 checkable dollars that exist only as bookkeeping entries. Call these central bank dollars CB$. This is shown in line (1). Then a private bank (Wells Fargo) buys a $200 bond from the public, for which it issues 200 of its own checkable dollars. Call these WF$, for “Wells Fargo dollars”. This happens in line (a).
Now, as I understand your case (3), the central bank issues 200 CB$ and buys the $200 bond from WF. This is shown in line (2) for the central bank and lines (b) and (c) for Wells Fargo. The 200CB$ are just soaked up by Wells Fargo and the money supply is unchanged. That’s the problem you’re talking about, right?
Now the central bank wants to inject another $50 into the economy. But Wells Fargo has no more bonds. Wells Fargo has nothing with which to buy the central bank’s 50CB$. Wells Fargo is now a ‘bad bank’. If the economy is full of bad banks then the central bank can’t inject any money.
But thankfully, a farmer comes to Wells Fargo asking for a $50 loan. Wells Fargo issues 50 WF$ in exchange for the farmer’s $50 IOU. See line (d). The money supply increased, but only because the farmer wanted money, and not because either bank ‘injected’ money.
Now the central bank thinks it sees its chance to inject $50. It buys the $50 IOU from Wells Fargo for 50 newly-issued CB$. (line 3) Wells Fargo sits on the CB$, without issuing any new money. (lines e and f) Wells Fargo once again has nothing to buy CB$ with, so it’s a bad bank again. Note that the farmer could just as well have borrowed the $50 from the central bank, and $50 would have been injected without involving Wells Fargo.
Moral of the story: $50 only gets injected if someone in the economy wants it badly enough to hand over $50 of their own assets.
Central bank balance sheet:
…ASSETS…………….LIABILITIES……..
1).$100 coins…………100 CB$
2).+$200 bond…………+200CB$
3).+$50 farmer’s IOU…..+50 CB$
Private bank (Wells Fargo) balance sheet
…ASSETS…………….LIABILITIES……..
a).$200 bond………….200WF$
b).-$200 bond
c).+$200 CB$
d).+$50 farmer’s IOU…..+50WF$
e).-$50 farmer’s IOU
f).+$50 CB$
8. December 2009 at 15:03
I’m with Nick — I don’t see why banks can’t indefinitely continue selling bonds to the central bank, and sitting on the payments in their reserve account.
8. December 2009 at 16:07
My non-economist intuition is as follows:
You said: “… this action will depress the value of bank liabilities relative to other assets like stocks. And this boosts AD.”
Partially due to the above, stock markets are rising. However banks are not lending to small businesses. This makes perfect sense if people expect the economy to be very risky in the near term but less so in the longer term, because small businesses have a much greater survival risk.
– So stocks this year have made for an attractive long-term bet relative to money, small businesses have not.
By this same logic, if the money supply were to continue to grow in the form of bad bank reserves, at some point the relative value of small businesses will compensate for their higher risk and they too should receive capital. So, to use a popular term: we should blow the new stock market bubbles as big as possible so that small businesses – firms that tend to actually hire people – will be relatively attractive compared to money and stocks. Then the new employees will finally, at long last, put the money back into banks as demand deposits.
8. December 2009 at 17:30
D Watson, That’s a small consolation.
Statsguy, My sense is that the problem was finding good loan candidates, not capital. Borrowing always drops off sharply in a deep recession, even if there are no banking problems. I don’t think we should focus on lending. Take care of NGDP and the banking system will take care of itself. The problems in the banking system today are mostly due to falling NGDP.
Nick, You said:
If the central bank buys a $100 bond (or buys $100 of anything) directly from the public, the public then deposits that $100 in a bank, and deposits expand. Multiplier of one.
But if the central bank buys a $100 bond from a bank, or just lends the bank $100 more reserves, the bank could just let the reserves sit there (if it is capital-constrained, for example). Multiplier of zero.
I can’t figure it out.”
I’m not sure what the puzzle is here. The central bank can buy all of the bonds that banks have. At that point all bank assets are reserves. So any additional OMOs are with the public, and thus increase the money supply. That is the third case I considered. The only issue facing the central bank is whether they consider a big monetary base to be a good thing, or a bad thing (in which case they simply put an interest penalty on excess reserves.) I still don’t see how bad banks create any sort of problems for monetary policy.
JimP, Tariffs in Chinese exports? And this is a University of Chicago professor? I hang my head in shame. Things are getting ridiculous.
Winton, No, as I pointed out, once all bank assets are reserves, then any further increase in bank assets will increase the money supply. See my response to Nick.
OGT, My problem with that sort of analysis is that they have misdiagnosed the problem. Until people realize that most of the financial crisis is caused by tight money, we aren’t going to be able to prevent repeat episodes of this. They seem to assume that this problem just sort of happened out of the blue, because of too much lending.
mlb, The standard view is that prices are also sticky in the short run, so that when NGDP falls sharply, RGDP also falls sharply, and inflation only drops by a little bit. This is a relatively flat SRAS curve.
In my view we don’t have accurate measures of the price level, which is why I focus on the fall in NGDP, not prices. NGDP has fallen about 8% below trend since mid-2008, and wages are nowhere near flexible enough to accommodate that. Hence unemployment rises sharply. The same thing happened in 1982, and inflation was 4% that year. But in 1981 it had been 10%.
Mike, You said;
“Moral of the story: $50 only gets injected if someone in the economy wants it badly enough to hand over $50 of their own assets.”
I’ve never known a central bank that had trouble finding people who wanted to sell it something. They can just walk into the bond market and start buying bonds.
JJ, Because eventually they run out of bonds. See my response to Nick.
8. December 2009 at 17:30
To put my above scenario in the context of inflation or NGDP expectations: with greater bank reserves, the value of small firms (as well as large and everything else) would inflate relative to money.
8. December 2009 at 18:19
Mike:
I think I’m following you. Except I might disagree with you here:
“But thankfully, a farmer comes to Wells Fargo asking for a $50 loan. Wells Fargo issues 50 WF$ in exchange for the farmer’s $50 IOU. See line (d). The money supply increased, but only because the farmer wanted money, and not because either bank ‘injected’ money.”
What you say sounds correct. But what if Wells Fargo is capital constrained, and so can’t issue a risky loan to the farmer? That was my worst case scenario.
8. December 2009 at 18:26
Scott: “I’m not sure what the puzzle is here.”
My puzzle is that I seem to be saying the equilibrium is path-dependent: whether or not the money supply (DD) increases depends on whether the central bank initially buys the bond from the banks or from the public. And that just seems weird. I know that path-dependent equilibria can exist, but I can’t see why the equilibrium here should be path-dependent.
8. December 2009 at 18:34
Scott
Well I think it is either going to be the Sumner way or else tariffs. Either way – people are simply not going to stand for the Bernanke brave new world of 10% unemployment and gigantic trade deficits to keep prices down. That just wont fly. I wonder why Obama does not realize this?
8. December 2009 at 20:20
Scott:
“I’ve never known a central bank that had trouble finding people who wanted to sell it something. They can just walk into the bond market and start buying bonds.”
That’s because central banks usually offer $100 cash for bonds worth $100. If the bank offered $99 for a $100 bond, they’d have lots of trouble finding people. If they offered $101, customers would be breaking down the door.
Money-issue is a two-way street. The central bank doesn’t just shove money into peoples’ hands.
Nick:
“whether or not the money supply (DD) increases depends on whether the central bank initially buys the bond from the banks or from the public.”
If the central bank bought the $50 bond directly from the farmer, the money supply rose by $50. If they bought the farmer’s $50 bond indirectly, through the bank, the money supply also rose by $50. But if not for the farmer, the bank would have been unable to ‘buy’ the $50, and if the central bank had offered the $50 directly to the public, there would have been no farmer willing to offer the $50 bond for the $50 cash. In this case it makes no difference whether the central bank buys the bond from the bank or from the public.
“But what if Wells Fargo is capital constrained, and so can’t issue a risky loan to the farmer?”
You mean that WF needs $10 of capital in order for anyone to want to do business with it, but it only has $6? My knee-jerk answer is that if the public wants the $50, then the prospect of a profitable issue of money will attract the needed $4 of capital to the bank.
8. December 2009 at 20:21
i think mike’s post demonstrates the cardinal importance of expectations/credibility. it seems key that the fed’s balance sheet is viewed not as a balance sheet but as a black hole. i agree with mike that any bank can be thought of to create money, if you think about it that way. but the fed is more powerful and thus can show more leadership. so the fed creates and destroys, but only so long as the perception of godlike stoicism exists.
8. December 2009 at 20:30
mike, i’ll argue that offering to pay market prices for anything, if the sum is large, is the equivalent of paying $101 for $100. and this is what the fed does. just like the treasury did with the AIG bailout…
8. December 2009 at 21:21
mlb – Your numbers are wrong. Peak CPI this cycle was 5.6% in mid-2008, where the last reported value is 0.2%. Peak core CPI was almost 3% in mid-2006. Last reported value is 1.7%. And this recession started with an increase in food and energy prices, so the core rate might be omitting important information. I can’t speak for Scott, but I also think its mis-stating his thesis to say that the current crisis is due to a fall in the price level and sticky wages.
8. December 2009 at 22:04
Mike:
You are confusing money and credit.
The money is the central bank liabilities. The people the central bank buys bonds from will accept the money even if they have no desire to hold it. They accept it because they can spend it on what they want.
The credit end of things–yes, the central bank will generally need to bid up the price of what it is buying so that it gets it rather than someone else.
The liabilitity side of the bank balance sheet is money and people will accept it even if they don’t want to hold any more of it. They take it because they spend it.
Now, for the central bank (or any bank) to lend more (which would include buying more securies) the may well need to accept lower yields.
Again, the horrible mistake is to assume that the Fed must change the yield on the bonds so that people are willing to hold more currency. No. People will take the currency because they can spend it. Maybe the central bank must pay more for bonds to get them instead of someone else.
The liquidity effect on interest rates is an unintended consequence of people spending excess money holdings on bonds, and lowering their yeilds, and lowering the opportunity cost of holding money, and so increasing money dmeand.
But it isn’t just the Fed has to bid up the price of bonds to get peopel to take the money, because there must be an addtional demand for money for people to take it.
People take money becuase it is the medium of exchagne. The will hold more money than they want, planning to get rid of it by spending it. And their efforts to spend it, through a vareity of avenues, result in people being willing to hold the additional money.
8. December 2009 at 22:27
Scott – I’ve been reading your posts, and Nick’s on this with great interest. And Arnold Kling’s with great puzzlement. I’m starting to wish I didn’t have job completely unrelated to monetary policy to do so I can sit and think about this for a few days.
I keep coming back to my incredulity at the idea that the actual amount of cash in the economy is so significant. It doesn’t match my intuition about the way the modern world works, so I’m happy to see this post where you assume your way into Nick’s world (which is rather more like the one in my head).
But I’m still stuck on something – real banks aren’t actually very much constrained by the amount of cash on hand in their lending activity. A bank can create $1000 by lending it to a customer by putting in their checking account. Providing the system as a whole is liquid they can do this without having to fastidiously check that their reserves are sufficient, because that same $1000 will soon be deposited in another bank somewhere, and if they don’t get enough of it to act as the reserve on their loan, they can easily borrow it overnight from another bank.
The whole system can operate without needing very much cash or even very much in reserves. So where does the fed’s influence really come from? It can put money into circulation and take it back out again, but (as above) any bank can do this. Why is the fed special? Because it doesn’t need to worry about its reserves or capitalisation at all, I suppose, and because its very big. But apparently the world is now full of big banks that don’t have to worry about their reserves or capitalisation at all …
Sarcasm aside, I’m beginning to think that the quantity theory doesn’t really capture what happens properly, or rather than the fed and M0 aren’t really all that much more important that other banks and other money. Rather actors (including banks) generally have a preferred range of holdings at different levels of liquidity eg. I’d like to have $50 in cash, two week’s salary in my checking account, 3 months salary in short term deposits and everything else invested invested in longer term assets. The fed’s role is to change the level of liquidity in different parts of the system (eg. by reducing the interest rate on my savings), and thereby force actors (especially banks) to move their holdings between different asset classes.
This may amount to the saying the same thing as the quantity theory, but its easier for me to match with the financial world in my head where cash really doesn’t do much except act as an intermediary between ATMs. It certainly doesn’t seem to contradict your general thesis that the fed should buy whatever it needs to buy in order to keep NGDP on trend.
9. December 2009 at 01:10
@Simon K. “So where does the fed’s influence really come from? It can put money into circulation and take it back out again, but (as above) any bank can do this. Why is the fed special?”
Let me explore this question more thoroughly, by looking how the fed is different than a normal bank and how the debt they issue is different than a normal bank. As an aside, do not confuse federal reserve liabilities (dollars) with treasury liabilities. Even though the federal reserve is the top owner of treasury debt, they work differently, as you can’t pay your taxes directly in t-bills at face value.
Where fiat currency is the same as bank debt:
Bank debt isn’t that much different from currency, except in a few critical aspects. As Greenspan explained, a state fiat currency is an IOU from the government against unspecified future taxes/services/fines/etc. For example: remember, recently, when California was giving out IOUs for tax refunds? If California had also accepted those in lieu of tax payments it would have created its own de-facto currency.
Where our fiat currency is different from bank debt:
1. Banks DD must pay in dollars, on demand, which the banks cannot create. (Yes I realize that term debt, doesn’t pay on demand, but one can still demand dollars for the debt, when it matures.) This is like when the federal reserve was on the gold standard and had to pay gold for dollars. Unlike free banks or national banks, retail banks cannot denominate their liabilities in terms of their liabilities.
I guess a hedge fund could do something like this but only for ‘qualified investors’; it wouldn’t be useful though as it would be hard to spend.
2. The federal reserve gets the currency first. This means any money it spends is un-inflated. (before someone starts screaming about how EMH means that the inflation is taken into account by those who sell to the bank, I would like to point out the name of this blog.)
3. The federal reserve is not a for profit entity, but a government. It doesn’t have to function by the rules of profit/loss accounting. This is why they, for example, bought debt valued on the market at less than 20% face value from Goldman for full face value. Being a part of the state, they can take losses every year (not that they do) and still not go under so long as they have political support.
4. The dollar market is the most liquid market on earth. It is prohibitively expensive for anyone who isn’t the government of a large wealthy nation to stage dollar runs or to manipulate the currency.
5. The federal reserve can tell any other US bank what to do and have the full backing of government coercion to back them up. They don’t need to use merely their own power.
9. December 2009 at 01:12
Oh, yah I’d like to add, in a fiat world, central bank debt is money. In gold standard world, gold is money and central bank debt is more like a bank demand deposit.
9. December 2009 at 05:20
Sometimes Sumner tells implausible stories about an excess supply of currency where people spend it on goods and services, bidding up their prices. These stories apply to an imaginary world where there is just goods, bonds, and “money” which takes the form of currency.
In reality, the biggest effect of an excess supply of currency is deposits into transactions accounts at banks, which, certeris paribus, generates an excess supply of deposits. Which people will actually spend on goods, services, and other financial assets.
Here is how it works. Those with excess currency go to a shop and buy stuff. The shop deposits the currency with all the checks and things they received. The excess supply of currency is now an increase in deposits, an excess supply of deposits, and excess reserves at the bank.
Of course, in a sense, the shops are spending the currency on deposits. If the shop owners had purchased restaurant meals, then the restaurant owner has excess currency balances and spends them. Well, if they “buy” bank deposits, then the bank has excess currency balances (called excess reserves.) The bank spends them by purchasing securities, or perhaps makes “loans.” This process creates more checkable deposits. Which are, certeris paribus, in excess supply. And people really spend those on a variety of goods, services, and financial assets.
If people take their excess currency and just deposit it before spending it, there is very little difference. It is like saying the Fed should use currency to buy bonds. Because then the bond dealers will take that currency out and buy coffee. No. They would deposit it. So, the Fed can buy bonds by crediting the reserve balances of the banks of the security dealers. That creates an excess supply of deposits and an excess supply of reserves more or less equivalent to sending the armored car full of currency to the offices of the securities dealers.
From Scott’s perspective, it doesn’t really matter what they do with the currency, any more than it is essential to go through the steps of the restaurant gets the money, and they pay a worker. The worker buys an apple. The grocery store buys bread… and on and on. The banks are combined with the other financial markets. There is nothing special about the banks.
If there is an excess supply of currency or money, and people decide they just want to hold increased money balances, then there is no further expenditure. But that is what must happen. An increase in the quantity of money must generate an increase in the demand to hold it for there to be a return to equilibrium. A higher price level does that. But if we just say, people decide to hold more. Well, that stops the process. Or, maybe we can come up with a reason they choose to hold more. With a currency focus, then it can be banks that decide to demand more currency–choose to hold more reserves.
Scott’s approach isn’t wrong. But the reality is that most of the purchases of goods and services and nonmonetary financial assets will actually be done by deposit–check or electronic payment.
Imaginging that all the payments are done with currency–drug deals, purchases of coffee from starbucks.. well, that is unrealistic. And so, I think the currency focus must be used with care.
On the other side of the coin, if there is not an excess supply of currency, the banking system is contrained in its ability to create deposits and lend them out. People will accept the deposits in payment without wanting to hold them, but if an excess supply of deposits is created, (an increase in the quantity beyond desired balances) people will spend them on something. If they “spend” them on currency, that is what we call a withdrawal of currency. When they spend on other things, the prices of those other things might rise. But with currency, the price of currency in deposits doesn’t rise. (Or, perhaps we might say, the price of deposits in terms of currency doesn’t fall.) Redeemability keeps the price of deposits tied to the price of currency. And so, the banks must reduce the quantity of deposits to match the demand to hold them, which limits the amount of loans banks may make.
If the central bank just issues currency passively, then this doesn’t work. If it is all part of a scheme to target short term nominal rates, then the quanttiy of deposits and currency all change so that they are consistent with short term interest rates being on target. But the banks are always controlled by the convertibility between deposits and currency.
With private currency, then it is only bank reserves that create this limit. These are reserves that banks use for interbank settlements. Clearing balances. The way it works is that an excess supply of currency and deposits (both issued by banks) geneates more spending, more gross clearings, a higher variance of net clearings, and a greater demand for clearing balances.
Again, if the quantity of reserves varies to target some short term interest rate, then the amount of reserves will be at a point necessary to hit that target. But it is still redeemability into reserves (each bank has to settle net claims with clearing balances) that limits each bank and the banking system.
All of this discussion was about possible capital contraints on bank lending. Well, that can be a problem with bank lending. But it isn’t a problem with increasing the quantity of money. As long as the central bank can buy something, and there is enough of all the things it can buy to be equal to the amount of the medium of exchange, deposits and currency, that it wants to create, then it can create it. If it owns all the assets already, and it wants to create more money, then there is a problem. If only some assets are counted as appropriate to hold, and it owns all of them already, then there is a problem. But that is it.
But half of the world seems to be more worried about the supply of bank loans rather than the quantity of money. And when they read, increase in the quantity of money, they undestand, more loans. Well, that isn’t what monetary equilibrium/disequilibrium theorists are talking about.
9. December 2009 at 05:27
Simon K:
“Providing the system as a whole is liquid they can do this without having to fastidiously check that their reserves are sufficient, because that same $1000 will soon be deposited in another bank somewhere, and if they don’t get enough of it to act as the reserve on their loan, they can easily borrow it overnight from another bank.”
Recognize that in Sept. 08, the system as a whole saw liquidity plummet due to fears about losses in short term money markets. This is why I’ve maintained the failure of WaMu was worse than the failure of an investment bank. Libor spiked immediately thereafter, interbank lending stopped, and reserves actually became important again – we saw real runs on banks. Banks started to actually compete for reserves again after years of reserves being rendered nearly worthless due to practices like sweeping. (read the story of E-trade’s virtual bank run)
The universality of this phenomena guaranteed a general liquidity crisis, but the speed (and mechanisms) by which this became a full fledged cyclical crisis caught the Fed off-guard.
9. December 2009 at 06:45
Hi Bill
“In reality, the biggest effect of an excess supply of currency is deposits into transactions accounts at banks, which, certeris paribus, generates an excess supply of deposits. Which people will actually spend on goods, services, and other financial assets.”
What we seem to have seen in the UK is them spending it 1% on goods, 1% on services, and 98% on financial assets. And this, to me, seems to make sense. In some sort of heretical way, I don’t get how quite radical changes in my liquidity position – or, really, the liquidity position of my fund manager – is going to change my decision to spend money on real concrete stuff, that I can’t get back if things go wrong. I will only change that decision if I think real variables related to ME will change soon – you know, my own wealth, my own income, security, etc.
I am a very big fan of Scott’s emphasis on expectations, and how monetary policy is meant to shift them. But this mechanical means by which we robotically spend extra liquidity is not something I can get a grip with: and it we don’t believe it will be true, doesn’t it undermine the expectation channel?
In which case, we need something else out there to make the new cash go into real spending – which is why I go along with Krugman (or Nick Rowe?) and ask for a bit of fiscal-monetary action – helicopters, or monetizing fresh govt spending.
9. December 2009 at 12:18
At last –
Someone wants Bernanke out – for the right reasons.
Yglesias – he wants Obama do DO something – what a good idea.
http://www.thedailybeast.com/blogs-and-stories/2009-12-08/why-obama-cant-create-jobs/?cid=bs:archive16
10. December 2009 at 11:21
Let’s hope the technical snafus are over!
rob, I agree.
Nick, My initial reaction is that if the Fed bought a T-bond from a bank, the bank could replace it by buying one from the public. I can’t imagine how the final distribution of T-bonds between the banks and the public could be affected by whether the Fed initially bought it from the public or banks.
Mike, I agree, but what does your reponse here have to do with your previous point?
Simon. Yes, that’s right. I’d be happy with either of these statements:
1. The recession is caused by a fall in actual prices, combined with sticky wages.
2. The recession is caused by a fall in NGDP, combined with sticky wages.
By actual prices I mean prices of things like housing, not pseudo-prices like monthly rents on long term contracts. The CPI measures pseudo-prices.
Simon#2, I understand your confusion, and it is because you are confusing nominal and real variables, and also because you assume that cash is just another fiancial asset. Here is an analogy:
Suppose there is a huge crop of kiwi fruit. I think everyone will agree that kiwi fruit doesn’t seem particularly important, right? But here’s what would happen as a result. NGDP in terms of kiwi fruit would soar. For instance, if the dollar price of kiwi fruit doubled, the NGDP measure in terms of kiwi fruit would also double immediately.
Now if you believe my opponents then this should not happen. After all, lots of other fruits are close substitutes for kiwi fruiits. Indeed far closer substitutes that DDs are for currency. And yet, we know it does happen. We observe it all the time. And I would wager that every single economist in the country agrees with me. I’d bet every single economist agrees that a huge increase in the supply of kiwi fruit will cause NGDP in kiwi terms to soar.
So in order to claim that Fed control of the base doesn’t give it this power, you have to claim that other financial assets are closer substitutes to currency than pears are for kiwi fruit. But that seems nuts. I don’t much care what kind of fruit I eat. But I do care a lot whether I have cash or DDs. Cash is far more convenient than checks for many small transactions. The biggest demand for cash is from tax evaders and foreigners. And bank accounts in the US are not at all close substitutes for those fellows. So just as more kiwi lowers its value, more currency lowers its value.
I better post this before it crashes again, and then do the rest.
10. December 2009 at 11:53
Doc Merlin. Those are very good points. Your second shorter comment is something that many people overlook. Under the gold standard, currency is a liability. Under a fiat regime currency is like gold. It is real wealth. (I’m guessing Mike won’t agree with our view here.)
Bill, You said;
“Sometimes Sumner tells implausible stories about an excess supply of currency where people spend it on goods and services, bidding up their prices. These stories apply to an imaginary world where there is just goods, bonds, and “money” which takes the form of currency.”
and
“Scott’s approach isn’t wrong. But the reality is that most of the purchases of goods and services and nonmonetary financial assets will actually be done by deposit-check or electronic payment.”
I agree with the thrust of your comment, but let me put a slightly different spin on it. I do think that in the long run the QTM works only because the extra cash bids up prices. But in the short run other (financial) factors are relatively more important. But here’s what I think most people get wrong. They realize the excess cash balance mechanism doesn’t seem to be what drives short run changes in the economy, and then they assume they can ignore it. But if it does cause long run changes, then the expectations of those long run changes can drive short run changes. So it is still very important.
Regarding you second point, I agree from a sort of transactions perspective that currency doesn’t seem that important. But again I think that is slightly misleading. What matters isn’t what share of NGDP is conducted with currency, but whether that share is stable over time. So if currency is used for only 15% of NGDP, and yet the ratio is stable, then when you double the currency supply then NGDP will also double, despite the fact that currency is only used for 15% of purchases. My argument here is sort of a Cambridge K argument. As long as k is stable, no matter how small, the QTM holds.
Giles, There is no simple answer to your question. First of all the central bank would have to insure that the extra base money goes into currency in circulation, not excess reserves. That can be easily done with a penalty rate on ERs, as I’ve discussed. So that eliminates the 98/1/1 problem you mentioned.
As for the second problem, you don’t see how extra cash would make you spend more, because you are thinking in real terms, and suffering from the fallacy of composition. We all think we can control how much money is in our wallets, and we all think that going from $100 currency in our wallets to $200 in our wallets won’t sufddenly make us double our spending on goods and services.
But it turns out both of those things “that we all know” cannot simultaneously be true. Suppose you customarily like to hold 100 pounds. Presumably this is not due to there being something special about that number. Rather it might be enough so that you only need restock at an ATM once a week. Now the BOE doubles the supply of currency in circulation. The average person will HAVE TO hold twice as many pounds as before. But by assumption they don’t WANT TO, they prefer to hold smaller balances. So they try to get rid of it. They might try to deposit the currency in their bank account, but with a penalty rate on ERs, the banks don’t want it either. The “hot potato” problem gradually forces up the price level, until things cost twice as much. Then you are again happy holding that much currency. If currency didn’t matter, then the BOE could force you to hold more real currency than you want to. But it can’t.
JimP, Yes, that’s a good Yglesias post. BTW, one of his links is to another Yglesias post that discusses my views.
10. December 2009 at 12:34
Many of the thought-experiments discussed here–for example, those involving the consequences of a big change in the supply of currency–seem unrealistic, almost as if they had been conjured up by *philosophers*. The reason for preferring an extreme thought-experiment is that it is more vivid: the consequences of a big change in the value of the monetary unit should be easier to see than if the change is small. But the irrealism of the extreme case may have the perverse effect of baffling intuition.
One of the virtues we require from the monetary unit is (relatively) stable value. This isn’t true of an ordinary commodity, such as copper or, even more obviously, kiwi fruit. Everybody knows there might one year be a bumper crop of kiwi fruit, or a “disastrous” crop failure. There is nothing unrealistic about asking: what if the supply of kiwi fruit suddenly tripled, or declined by two-thirds? Perhaps the price of kiwi fruit in the former scenario might fall, say, 25% (*ceteris paribus*; there are so many close substitutes for kiwi fruit!); equivalently, the KGDP (GDP measured in kiwi fruit units) would rise 33.3%. But if something like this were expected to happen with dollars, people would choose a different unit of account. (This is true in the long run; in the short run there are collective-action difficulties in selecting a new unit of account.)
This makes for a *rhetorical* problem: how wild should you make your thought-experiments?
10. December 2009 at 14:02
Well, I wrote a long comment, then the hamster powering Scott’s blog had a heart attack, and my comment never got posted, and I lost it. Glad to see it running again!
1. I agree with Bill Woolsey.
2. Mike Sproul: “You mean that WF needs $10 of capital in order for anyone to want to do business with it, but it only has $6? My knee-jerk answer is that if the public wants the $50, then the prospect of a profitable issue of money will attract the needed $4 of capital to the bank.”
In normal times, if the central bank increased the supply (curve) of reserves, so it became profitable for banks to increase loans, they would indeed raise additional capital, if it were needed, by issuing new shares, for example.
But if banks’ capital has taken a big hit, due to bad loans, and if their share prices have taken a big hit too, they may be reluctant to issue new shares, if they think that the price of shares is below the fundamental value. Maybe what I have in mind is some sort of asymmetric information model, where the market values new bank shares at less than the banks think they are worth. So banks are capital constrained.
3. Scott: “Nick, My initial reaction is that if the Fed bought a T-bond from a bank, the bank could replace it by buying one from the public. I can’t imagine how the final distribution of T-bonds between the banks and the public could be affected by whether the Fed initially bought it from the public or banks.”
Yes, half of my mind thinks that way too. But maybe Tbills and reserves at the central bank are close to perfect substitutes in the eyes of banks right now.
10. December 2009 at 14:05
Bad banks don’t create money, they destroy it. So huge expansion of central bank balance sheet is needed to restore NGDP growth to trend.
10. December 2009 at 14:26
Scott, the fact that only kiwi prices fall due to big kiwi crops proves that other fruits are NOT close substitutes for kiwi. If they were, then a huge kiwi crop would cause all fruit prices to fall. And how will your claim hold up N years from now when there is no cash, only electronic transfers? Today I already operate 90% with direct deposits and withdrawls.
I think the important distinction is between cash/DD’s, and Bonds/stocks/etc. As Simon said: “actors generally have a preferred range of holdings at different levels of liquidity”. If the central bank buys my bonds, I’ll want to reinvest the cash in more bonds, thereby 1) bringing bond rates down, and 2) giving the cash to somebody who wants to spend it.
10. December 2009 at 18:36
Bill Woolsey:
“You are confusing money and credit.
The money is the central bank liabilities. The people the central bank buys bonds from will accept the money even if they have no desire to hold it. ”
There is no difference. The central bank issues 100 CB$ and spends them on $100 of bonds. Wells Fargo issues 100 WF$ and spends them on $100 of bonds. Both the CB$ and the WF$ might exist as checking account dollars, or as paper dollars, or as credit card dollars, etc. WF$ might be claims to a CB$, in which case CB$ are base money and WF$ are derivative money, or they both might be claims to a silver dollar coin, in which case they are both derivative money.
“The credit end of things-yes, the central bank will generally need to bid up the price of what it is buying so that it gets it rather than someone else.”
The same is true of Wells Fargo.
“the horrible mistake is to assume that the Fed must change the yield on the bonds so that people are willing to hold more currency. No. People will take the currency because they can spend it.”
Again, same for Wells Fargo. But people only get 100 CB$ or 100 WF$ if they want it badly enough to give up a $100 bond for it.
“People take money becuase it is the medium of exchagne. The will hold more money than they want, planning to get rid of it by spending it. And their efforts to spend it, through a vareity of avenues, result in people being willing to hold the additional money.”
Nobody takes either CB$ or WF$ unless they want it more than they want their bonds. Unwanted money refluxes to the issuing bank and never gets spent. The only way the value of the money drops is if the issuing bank reduces its backing.
10. December 2009 at 19:08
Nick Rowe:
“if banks’ capital has taken a big hit, due to bad loans, and if their share prices have taken a big hit too, they may be reluctant to issue new shares, if they think that the price of shares is below the fundamental value.”
I think a capital constrained bank is like a capital constrained farmer. The bank can’t issue money and the farmer can’t grow wheat. But of course other banks and farmers move in and take over, and the invisible hand takes care of it. My conclusion is that bad banks are no more of a problem than bad farmers.
10. December 2009 at 19:33
Philo, You said;
“Perhaps the price of kiwi fruit in the former scenario might fall, say, 25% (*ceteris paribus*; there are so many close substitutes for kiwi fruit!); equivalently, the KGDP (GDP measured in kiwi fruit units) would rise 33.3%.But if something like this were expected to happen with dollars, people would choose a different unit of account.”
Actually this isn’t the case. There are very serious economies (not just Zimbabwe) that have had high and variable inflation for many years and people still keep using their currency, at least to some extent. I am talking about places like Brazil, Turkey, Israel, and many others.
You need these thought experiments because monetary theory is so counterintuitive. It doesn’t seem like inflation is nothing more than dollar bills losing value. It seems like inflation is everything else getting more expensive. But its an illusion (a money illusion). Inflation is nothing more than currency losing purchasing power.
Nick, You said;
“Well, I wrote a long comment, then the hamster powering Scott’s blog had a heart attack, and my comment never got posted, and I lost it. Glad to see it running again!”
Sorry. You wouldn’t believe how many super long comments I have lost. In January we plan to replace the hamster with a gerbil, and my tech guys insist it will work much better. (seriously, we are moving to a for-profit service, and off the Bentley server thing-a-ma-jig.)
Your last comment about liquidity traps probably does change things. But I think it is useful to first work out how things work in normal cases. I still think the liquidity trap can be dealt with by negative interest on ERs if necessary.
123, Not necessarily as much as you’d think, it depends on how much ERs are hoarded. If the central bank prevents that with a penalty rate, the base need not increase that much.
jj, I see your point, but my point was that things that seem like close substitutes actually don’t behave that way. It isn’t enough for someone to say “well DDs and cash seem like close substitutes to me.”
I could say the same about fruit. The point is to examine how people act. And they don’t act like currency and DDs are close substitutes. For 90% of the demand for cash (hoarding and many small transactions), DDs and currency are very different. When 100% electronic money arrives then I agree monetary policy will probably be done with interest rates, not OMOs. I’ve discussed that before in one of my posts.
11. December 2009 at 01:46
May I thank everyoine for the phenomenal tutorial you are all granting, for free, to occasional contributors? I suspect that for each of us there are a hundred silently enjoying.
If Banks were heavily penalised for their reserves (say, minus 2%), expect more letters from the likes of Ros Altmann, a consistent advocate for savers (the deflation lobbyists, you may call them). These will begin to tell after a while, because, frankly, a lot of people benefit from the current regime – they have fixed nominal incomes, and a high propensity to vote, and don’t really care about unemployment.
http://www.ft.com/cms/s/0/b3bf224a-e5f5-11de-b5d7-00144feab49a.html
And if the Banks don’t see an adequate, risk-adjusted return on lending the money out, won’t they either stick it in vaults as cash or buy inert items like gold etc – or back into Tbonds? So I still wonder what happens once people re-deposit cash. Personally, I use a lot of cash because in the UK builders avoid VAT, but I suspect that any personal stories will involve the fallacy of composition.
My final point is this: if committed experts like those on this thread can express uncertainty about whether and over which lags such operations result in a rising NGDP, does this not undermine the expectations mechanism in the wider economy? If Professors and brainy types like Kling can be uncertain, why would man on Main Street adjust his behaviour?
Many thanks
11. December 2009 at 06:06
SSumner- I am not sure that’s fair to their research. In essence, it seems to me to say that credit is delinked from broad money, and has historically been more important than broad money measures in causing financial crisis.
In addition, credit itself then started to decouple from broad money and grew rapidly, via a combination of increased leverage and augmented funding via nonmonetary liabilities of banks. In recent decades, we have been living in a different world, where financial innovation and regulatory ease have permitted the credit system to increasingly delink from monetary aggregates, resulting in an unprecedented expansion in the role of credit in the macroeconomy. By 2007, the typical level of broad money had risen to about 70% of GDP, but bank loans exceeded 100% and bank assets were over 200%.
In a sense consumer credit only pulls demand forward, this can be good in terms of consumption smoothing. But, at a certain point it creates contractual demands for nominal payments that may not be stable.
At this point I’d argue that authorities need to raise NGDP up to a level where those contractual demands become sustainable. If credit is largely independent of monetary policy without further regulation, is your contention that money supply should always rise to meet credit creation? Or that there research is incorrect?
11. December 2009 at 06:37
Or, rather, ‘their research’ in incorrect?
11. December 2009 at 07:29
Giles, I partly agree about the savers, but remember that lots of high savers (like me) have stocks and corporate bonds, which were both adversely affected by the deflationary policies.
You said;
“And if the Banks don’t see an adequate, risk-adjusted return on lending the money out, won’t they either stick it in vaults as cash or buy inert items like gold etc – or back into Tbonds?”
No, banks cannot hold it as vault cash either, as VC is a part of reserves. They’d have to spend it, and that boosts AD. If they spend it on bonds the cash will go into circulation, where it will help boost spending.
Your Ratex question is interesting, and there is a lot I could say. But for me the bottom line is that the markets are smarter than the experts. Economists are all over the map on inflation predictions. But the TIPS market expects low inflation. And I belive that is the correct answer. So I am not discouraged by the disagreement among economists. In earlier months I had posts on the “wisdom of crowds” idea.
Sorry OGT, I missed something. Which research paper were you referring to? (When I answer many posts I often lose the original thread of our discussion.)
11. December 2009 at 09:31
Scott, you said;
“123, Not necessarily as much as you’d think, it depends on how much ERs are hoarded. If the central bank prevents that with a penalty rate, the base need not increase that much.”
If banks are capital constrained, quantity of ERs is fixed, and penalty rate has much smaller effect than QE.
11. December 2009 at 14:15
Scott, dd’s may not replace all functions of cash, but cash can replace DD’s. Imagine I start a Vault business that will store your cash and allow withdrawals at any of my locations. We do not make loans and are not required to hold a reserve with the fed. We are highly efficient and charge only 0.1% per annum.
Banks subject to an excess reserve penalty will first try to reduce their deposits (eg no fees on balances below $100!). Customers withdraw their cash but presumably don’t have any better investment opportunities than the bank did, so they stick it in my Vaults. Once deposit accounts are empty, if the penalty is lower than the expected loss of the unattractive loans, the bank will just eat the penalty (but of course they sold their bond to the fed for enough money to make it nonetheless profitable).
I think the penalty will have to be pretty big and disruptive to get the banks’ helicopters off the ground. Of course the expectation that the fed will take it this far will reduce the effort required…etc, that’s another topic.
11. December 2009 at 17:06
123, You lost me there. If ERs are fixed, then all new injections of base money goes to currency held by the public. That is likely to be quite inflationary.
My point about the panalty rate was not to use it so much as a tool in and of itself, but to make QE much more effective.
JJ, The problem you are talking about is safety deposit boxes, which are a safe way to hoard. But the penalty rate doesn’t have to be high at all, because banks have the alternative of 0% T-bills. So if the penalty rate was even negative 2% on reserves, ERs would be replaced by T-bills. I have a very long post back in the spring that discusses this in detail. It was called Reply to Mankiw.
12. December 2009 at 04:39
SSumner- I am not sure that’s fair to Taylor and Schularick’s research. In essence, it seems to me to say that credit is delinked from broad money, and has historically been more important than broad money measures in causing financial crisis.
http://www.voxeu.org/index.php?q=node/4349
In addition, credit itself then started to decouple from broad money and grew rapidly, via a combination of increased leverage and augmented funding via nonmonetary liabilities of banks. In recent decades, we have been living in a different world, where financial innovation and regulatory ease have permitted the credit system to increasingly delink from monetary aggregates, resulting in an unprecedented expansion in the role of credit in the macroeconomy. By 2007, the typical level of broad money had risen to about 70% of GDP, but bank loans exceeded 100% and bank assets were over 200%.
In a sense consumer credit only pulls demand forward, this can be good in terms of consumption smoothing. But, at a certain point it creates contractual demands for nominal payments that may not be stable.
At this point I’d argue that authorities need to raise NGDP up to a level where those contractual demands become sustainable. If credit is largely independent of monetary policy without further regulation, is your contention that money supply should always rise to meet credit creation? Or that there research is incorrect?
http://www.voxeu.org/index.php?q=node/4349
(Reposted with link and clarified references).
12. December 2009 at 06:56
OGT, Now I understand what you are asking. When I say tight money caused the crisis, I do not mean that the monetary aggregates fell. I don’t regard the aggregates as a good indicator of monetary policy. I mean that NGDP fell. If the Fed prevents NGDP from falling, and even better keeps it rising at about 5% a year, which is what I have proposed, then any credit crises that occur will only have a modest impact on the economy. And the best way to prevent credit from getting out of hand is not monetary policy, but rather better regulation (not more regulation, better regulation.)
So to summarize, most of this recession is caused by tight money, and the part that is caused by reckless banking is better addressed through regulation than monetary policy.
12. December 2009 at 14:49
Sorry for the lack of clarity.
Let’s say the size of the central bank balance sheet is fixed, and central bank changes only the rate on ERs. If all commercial banks are capital constrained, then they don’t change the size of their balance sheets in response to different rates on ERs or as you say continue to “hoard” them. So my point is if that banks are capital constrained, the size of QE is much more important than the interest rate on reserves. I’d really prefer Gagnon with double QE (even if it was with 1 percent Fed funds rate) to Bernanke.
15. December 2009 at 14:15
Another point – bad banks reduce money multiplier, so central bank has to perform an unprecendented expansion of the quantity of reserves. As a result for many economists it takes almost a year to realize that their models with narrow range of money multipliers are obsolete, and in the meantime their recomendations cause various market anomalies, such as huge rallies in gold.
16. December 2009 at 06:01
123, I agree with your criticism of economists in your second comment. Rgarding the first, suppose the Fed went to a negative 3% rate on excess reserves. Don’t banks have the option of swapping them for zero interest T-bills? And aren’t T-bills able to be used just as effectively as reserves in terms of capital requirments? I think the answer is yes, but am not certain. If I am right, this still causes the ERs to go out into circulation as currency held by the public.
16. December 2009 at 12:39
Negative 3% rate would change almost nothing. Banks would shrink their balance sheets, on asset side ERs will be removed, on liability side deposits will be removed, and former depositors will hoard currency.
16. December 2009 at 14:30
123, I would not hoard currency, nor do I know many people who would. If you dump another $800 billion in currency into circulation, some will be hoarded, but probably not all.
In any case, I’m not recommending this as an optimal monetary policy, so the discussion is purely academic.
17. December 2009 at 12:10
“I would not hoard currency”
If you have a deposit that yields zero percent and if your bank calls you and says “we have a new product – currency in our vault – if you switch to it we will not raise you credit card fees”, I think you will start hoarding currency.
The point that QE is more important than interest rates on ERs when banks are bad has significant practical importance.
18. December 2009 at 07:38
123, I don’t follow, safety deposit boxes are costly—cash yields a negative rate of return.
19. December 2009 at 00:45
But only slightly negative. Net result of minus 3 percent rates on ERs is equivalent to the result of Fed funds rate of minus 0.1.
19. December 2009 at 12:02
123, Yes, I’ve always admitted that negative rates on reserves don’t drive market rates significantly below zero. That’s not the purpose of QE. (And negative rates on reserves are a backdoor way of dioing QE.) The purpose of QE is to get the excess cash balance mechanism into operation. It is not to work through the nominal interest rate transmission mechanism. If risk free rates are near zero, I completely agree with Keynesians who say thet can’t go much lower. But if inflation expectations rise because of QE, then real rates can fall.
19. December 2009 at 12:44
When banks are bad properly executed QE can work even when market rates are sky high. Best example here:
http://www.dallasfed.org/research/swe/2001/swe0103d.pdf
21. December 2009 at 13:53
123, Yes, that’s a good example of what monetary policy can do under difficult circumstances. I was buying lots of HK stocks in 1998 as well. The HK central bank and I think alike.
BTW, I don’t see the purchase of stocks as a big deviation from laissez faire as long as they are done in a neutral way (like an index fund) and as long as they are sold when the crisis is over.
21. December 2009 at 15:56
Then why are you always mentioning interest on ERs instead of size of QE in your blog?
21. December 2009 at 15:58
Purchase of index is almost no deviation from laissez faire compared to TARP, AIG etc…
22. December 2009 at 18:08
123, I don’t quite follow. Did the HKMA pay interest on reserves?
I didn’t talk about the option of the Fed buying stock, but I think Nick Rowe did. It is a way of reducing the time inconsistency problem, because now the central bank has an incentive to make sure the economy recovers.
I suppose I assumed that idea was not politically feasible here, which is why I didn’t mention it.
23. December 2009 at 14:15
HKMA did nothing to prevent interbank rates skyrocketing.
I don’t think that Fed stock purchases are feasible. But QE works well with many other asset classes, HKMA example just shows the power of QE during high interest rate situation.