Does the Fed believe in magic?
Some posts almost write themselves. A number of commenters sent me a WSJ story by Jon Hilsenrath, which is worth a half dozen posts.
Federal Reserve officials are considering a new type of bond-buying program designed to subdue worries about future inflation if they decide to take new steps to boost the economy in the months ahead.
Under the new approach, the Fed would print new money to buy long-term mortgage or Treasury bonds but effectively tie up that money by borrowing it back for short periods at low rates. The aim of such an approach would be to relieve anxieties that money printing could fuel inflation later, a fear widely expressed by critics of the Fed’s previous efforts to aid the recovery.
This reminds me of the “magical AD” theory that you see kicked around so much in popular discussion. It works this way:
1. Monetary stimulus creates inflation, and hence is bad.
2. Fiscal austerity reduces real growth, and hence is bad.
You see these ideas discussed quite frequently in the British press. In theory, both monetary and fiscal stimulus directly impact AD, and the effects on prices and output depend on the slope of the short run aggregate supply curve. But the “magical theory” of AD says that demand-side policymakers can get whatever P/Y split they prefer, by using a special “magic dust” which they sprinkle over the economy. Then when the Fed buys long term bonds to reduce interest rates and boost growth, the magic dust prevents any inflationary impact.
You might wonder how this magic dust is supposed to work:
The Fed’s approach to a bond buying program matters a lot to many investors. More money printing could push commodities and stock prices higher, or send the dollar lower, if it sparks a perception among investors that inflation is moving higher, said Michael Feroli, an economist with J.P Morgan Chase. However, if the Fed chooses a course aimed at restraining inflation expectations, the impact on those markets might be more muted.
Fed officials have used different types of bond-buying programs since 2008. In each case the aim has been to drive down long-term interest rates to spur investment and spending by businesses and households. In case they decide to act again, they’re exploring three different approaches, according to people familiar with the matter. Those approaches are:
– First, they could use the method they used aggressively from 2008 into 2011, in which the Fed effectively printed money and used it to purchase Treasury securities and mortgage debt. The Fed has already acquired more than $2.3 trillion of securities in several rounds of purchases using this approach, widely known as “quantitative easing,” or QE.
– Second, the Fed could reprise a program launched last year in which it is selling short-term Treasury securities and using the proceeds to buy long-term bonds. This $400 billion program, known as “Operation Twist,” allows the Fed to buy bonds without creating new money.
– Third, in the new novel approach, the Fed could print money to buy long-term bonds, but restrict how investors and banks use that money by employing new market tools they have designed to better manage cash sloshing around in the financial system. This is known as “sterilized” QE.
The Fed’s objective under any of these programs would be to reduce the holdings of long-term securities in the hands of investors and banks. The Fed believes that reducing the amount of long-term bonds in the hands of investors drives down long-term interest rates, encourages more risk-taking, and thus spurs spending and investment by households and businesses.
The differences between the three approaches involve where the money comes from and where it ends up. The Fed hasn’t literally print more money, but it has electronically credited the accounts of banks and investors with new money when it purchased their bonds under quantitative easing. The Fed has pumped more than $1.6 trillion in new money into the financial system this way, and has also rejiggered it existing holdings, as part of its bond-buying efforts.
Many Fed officials believe strongly the bank reserves it has created as part of this money creation aren’t an inflation threat. But they are acutely aware of a popular perception, also held by a few inside the Fed itself, that the money the Fed has created could cause an inflation problem down the road. An approach that limits the amount of new money flowing into the system””through another Operation Twist or a sterilized operation””could help them manage that perception.
Where does one even begin? The Fed’s already been sterilizing 100% of all monetary injections since October 2008; sterilization is certainly not a new program. So what’s really going on here? Despite the mocking tone of this post the Fed isn’t stupid, so there must be an explanation. I’m not sure what the explanation is, but I’ll take a wild guess:
1. There are two types of inflation expectations; those of the unwashed masses, and those held by sophisticated investors with lots of money on the line.
2. The naive masses think there’s a direct correlation between the size of the monetary base and inflation. They also think inflation is bad, as they believe in the magical theory of AD. They believe some types of increased AD (produced by something like quotas on Chinese imports) create jobs, and other types (say printing money) create inflation. The Fed doesn’t want to frighten this group by doing something that would produce scary stories about “printing money.”
3. The sophisticated observers understand what’s really going on. They understand that higher inflation expectations are good for the economy right now. And if they expect more inflation, it will drive up asset prices and trigger more AD, more growth.
In my view the real problem here is that the sophisticated investors are even smarter than the Fed. They don’t believe in magic dust, nor do they believe that lower long term nominal interest rates are an indication that higher inflation is on the way. They don’t buy into the Fed’s basically Keynesian worldview, and hence the policy is not likely to succeed unless . . . and this is where things get really tricky, unless it succeeds in convincing the sophisticated observers that the lower long term rates are a coded signal that the Fed intends to allow faster than expected growth in currency in circulation once we exit the zero rate bound.
In other words, Fed policy might “work” but for the wrong reason. How would we know it’s working? One sign is if they are not successful in reducing long term rates as much as they’d hoped.
I said there was enough here for a half dozen posts. Consider this:
Fed officials in New York and at the board in Washington are considering the costs and benefits of the different approaches. They have been pleased with the results of Operation Twist. But they would face some practical limits if they wanted to repeat it.
Louis Crandall, a money-market analyst with Wrightson ICAP LLC, estimates the Fed would have only about $200 billion of short-term securities left to work with in the second half of the year if it decides to repeat the Operation Twist program as currently designed. The current program ends June 30.
The third approach has some benefits the other options don’t have. Unlike Operation Twist, the size of the program wouldn’t be constrained by the Fed’s own holdings of short-term Treasurys. This approach would also give officials an opportunity to try out some of their new tools to see how they work on a large scale.
To a casual reader, this might sound like it’s a program to overcome the fact that the Fed is “out of ammunition.” In fact, just the opposite is true. This quotation refers to a low stock of government debt held by the Fed. However, being “out of ammunition” actually means there is a low stock of public debt not held by the Fed. This says the Fed has lots of unused ammunition, they simply don’t want to use it; they don’t want a larger monetary base.
Moreover, the program could be conducted with financial institutions other than banks, like money-market funds, increasing the Fed’s flexibility in managing reserves. The reverse-repo program was designed to include money-market funds and these institutions don’t participate directly with the Fed in other operations.
Yes!! Can we please completely separate monetary policy from the banking system? I get tired of commenters trying to bait me into talking about the banking system, when in fact the banking system doesn’t actually play an important role in monetary policy. It’s all about currency in circulation. It would make more sense to treat those interest-bearing ERs as government debt, not “high powered money.” Money is only “high-powered” when it earns no interest. And sometimes (as in Japan prior to 2008 when they began IOR) not even then.
Here’s the conclusion:
Still there are potential drawbacks. Reverse repos could push short-term interest rates””now near zero””higher than the Fed wants them to go. Moreover, the Fed has described the reverse-repo program as a tool to use when it wants to tighten credit. Using it in combination with a bond-buying program meant to ease credit could “send a lot of conflicting signals” to the markets, Mr. Feroli said.
The Fed’s an expert on mixed messages. The most famous example occurred in October 2008, when they adopted an avowedly contractionary policy in the midst of the mother of all financial crises, and the biggest fall in NGDP since 1938. I am referring, of course, to the infamous interest on reserves program, their first attempt at sterilizing monetary injections. This recent proposal is merely some accounting trickery to make the base look smaller, there’s no meaningful difference between interest bearing ERs and these short term repos.
HT: Michael Darda, Clare Zempel, Tommy Dorsett, John Hall
Tags:
8. March 2012 at 08:20
“I get tired of commenters trying to bait me into talking about the banking system, when in fact the banking system doesn’t actually play an important role in monetary policy. ”
Sounds similar to what Arnold Kling thought you’d say here.
Note that all the panel of expert economists thought that bailout out specific banks was the right thing to do. None spoke up for what Kling said would be your view, “We used to believe that employment depended entirely on the path of NGDP relative to trend. We used to believe that this path could be controlled by the monetary authorities, regardless of what is going on with individual banks. I have not changed my beliefs, even if everyone else has.”
I think Kling accurately characterized your views. Did he?
8. March 2012 at 08:39
All i can say is that its called “sterilization” for a reason. If it were productive it would be called something else. I have to refrain from thinking about this because my head will explode. its basically just an extension of operation twist. do we really really think lower interest rates are the problem???!!!!
8. March 2012 at 09:02
Stop with the currency in circulation. It is ridiculous.
If the Fed does reverse repurchase agreements, especially over night ones, the Fed is creating money.
Gee, it isn’t currency in circulation. So what?
The Fed borrows money overnight, and then funds long term bonds.
Those lending overnight to the Fed have what amounts to money in their checking account.
If such a person writes a check, then can cover it by having the Fed wire the fund to their bank to pay them back for part of the money the Fed owes them today.
If they have a sweep account with their bank, no wiring is needed. It _is_ money in their checking account which their bank is lending to the Fed overnight. The rules can be gamed so the bank doesn’t report it as a checking account and the Fed doesn’t report it as base money.
Of course, these overnight repurchase agreements will pay some tiny bit of interest.
Scott, get over it. Paying interest on money doesn’t matter!
The amount of money issued that pays no interest (currency in circulation) is no more important than other types of money.
More importantly, if the goal is some level of nominal GDP in the future, the Sumner/Krugman thought experiment of promising to leave some particular amount of base money (interest bearing or not) outstanding (say, doubling) after the zero bound goes away is _pointless._
Base money is endogenous and base money should be endogenous. What would happen if the quantity was changed a fixed amount matters for nothing. Not with the real world, and not with any sensible policy.
What does matter is that the Fed has to give up on treating interest rates (including expected future ones) as central . And it needs a level target–preferably nominal spending on output. But it cannot work by promising to keep base money at some fixed level at some future time at a level consistent with nominal GDP being on target after the zero bound goes away.
A noimnal GDP target means that base money will not be at any particular fixed level at any particular future time. It will adjust as necessary to keep nominal GDP on target.
Currency in circulation? How much worse.
I am not saying that if the Fed created a target for currency in circulation that this couldn’t cause inflation. I am saying that this would be bad thing to do. That the Fed should target a growth path for nominal GDP.
I don’t see any purpose for using reverse repurchase agreements to fund bonds rather than reserve balances. But it is still creating money.
8. March 2012 at 09:07
“You see these ideas discussed quite frequently in the British press. In theory, both monetary and fiscal stimulus directly impact AD, and the effects on prices and output depend on the slope of the short run aggregate supply curve. But the ‘magical theory’ of AD says that demand-side policymakers can get whatever P/Y split they prefer, by using a special ‘magic dust’ which they sprinkle over the economy. Then when the Fed buys long term bonds to reduce interest rates and boost growth, the magic dust prevents any inflationary impact.”
Shorter ssumner: One cannot control the P/Y split with only one policy variable. Therefore, using two policy variables to control the P/Y split is “magic”
Do I have that about right?
8. March 2012 at 09:12
“Paying interest on money doesn’t matter!”
Really, Bill Woolsey? I’d say it makes a big difference, if you’re talking about AD and the economy. Simply because money *can* be withdrawn doesn’t mean it will be withdrawn.
Do you dispute that if the Fed suddenly paid 10% interest on reserves, then banks (and anyone else) would be extremely unlikely to loan money out for less than that rate, and people would be less likely to consume, less likely to invest in any other projects that had a lower expected return?
Money that people have a strong incentive to keep with the Fed and not lend out might as well not exist, in the same way that money buried and never spent might as well not exist.
There’s no difference between money that you’ll never, ever spend, and simply burning the currency.
8. March 2012 at 09:14
Given that the Fed has been testing reverse repos a few times in the last few weeks, the MMF angle makes sense. The whispers are that those are who they hae been testing with.
I wonder if this will start the move towards targeting the O/N Repo rate instead of FF like is outlined here: http://www.ecb.europa.eu/events/conferences/shared/pdf/pocrides_opfram/Stebunovs.pdf?a49eabd5fdac178131a9cb046aaef78b
Getting more control over the non-bank monetary channels seems to me like a very good idea.
I think you are spot on that IOER is sterilization.
8. March 2012 at 09:15
D R:
The objection would be resolved if monetary stimulus and fiscal stimulus actually operated in different ways. But almost all the economists commenting supposedly teach, research, and subscribe to theories arguing that both affect AD exactly the same way. Yet they have a magical belief that in the real world, they’re somehow different.
If they think so, then they should produce research or at least theories supporting their views, instead of having a split between their academic work and personal views that would make a young earth creationist astrophysicist proud.
8. March 2012 at 09:15
“unless it succeeds in convincing the sophisticated observers that the lower long term rates are a coded signal that the Fed intends to allow faster than expected growth in currency in circulation once we exit the zero rate bound.”
Scott, I think your coded signal argument is spot on. The unwashed masses believe inflation starts and ends with gasoline and grocery bills. The Fed is afraid to feed those insecurities especially with the media fear mongers and wall st gold promoters ready to pounce on any increase in gasoline prices.
Meanwhile, do you know what has happened to the long-dated price of oil over the past 12 months?
Dec 2019 WTI has declined from $106 to $87.
Dec 2019 Brent has declined from $108 to $89.
That means we can look forward to 4% annual commodity DEFLATION for almost a decade. The Fed will really have to juice core PCE in order to hit a 2% headline target!
8. March 2012 at 09:26
@Bill,
i dont see how this is printing any money at all (“sterilization”). they buy long term bonds then take the money back out with the reverse repos (effectively selling the short term bonds). dollar-for-dollar no change in the base. It’s basically more operation twist, an attempt to further flatten the yield curve.
8. March 2012 at 09:36
@John Thacker,
I think what Bill Woolsey is saying is that the defining characteristic of money is that it is the medium of exchange. Whether or not interest is paid on it doesn’t matter for that.
Where interest on reserves does matter is that it affects the demand for money. When, as now, interest paid on excess reserves is as high or higher than alternative highly liquid assets, like short term T-bills, money demand goes way up. Even though the base has tripled since August 2008, money is tight because the demand for excess reserves has skyrocketed.
I agree with Bill on all of that. I also agree with him that Scott’s currency kick is misguided. Currency and reserves are both Fed liabilities, and the Fed has always been perfectly willing to exchange one for the other to whatever extent banks want to do so. That is, banks can deposit currency in their reserve accounts, or withdraw reserves as currency. To directly control the amount of currrency in circulation, that would have to change.
Can you imagine the hue and cry that would result if the Fed no longer accepted its own Notes for deposit, or refused to allow currency withdrawals from reserve accounts? I can’t believe and Fed chairman would ever want to find out.
8. March 2012 at 09:37
s/and/any/ in that last sentence.
8. March 2012 at 10:03
dwb:
The repurchase agreements _are_ money to those holding them.
Those who make these transactions with the Fed are lending funds to the Feds. If they are term repurchase agreements, then it is short term lending. If it is overnight, then it is money.
Now, these loans to the Fed are secured by the bonds. The Fed is selling bonds promising to pay them back. (In practice, these are particular bonds, just a set of goverment securities that belong to the seller/borrower.)
The Fed is effectively borrowing overnight. Those lending to the Fed are lending overnight. The transactions are rolled over, and so any day, the person lending to the Fed can get their money. It is like having money in a checking account.
The Fed, on the other hand, borrowing on these terms, has to be ready to pay off at any time.
Now, this money the Fed is creating won’t meet reserve requirements. Which hardly matters these days.
This money the Fed is creating can be directly held by firms other than banks offering transactions accounts.
It is true that this will not create more official “base money.” It will not create excess reserves for banks to lend.
But it creates money.
What if the Fed directly offered checking accounts to you and me and used the proceeds to purchase long term government bonds.
Well, the banking system wouldn’t be issuing more checking accounts. And the banking system wouldn’t have more resreves.
But there would be more money.
That is what this amounts to.
of course, if these are 5 year reverse repurchase agreements, I take it back.
The closer they are to overnight or continuing, the more this amounts to the Fed directly creating money to fund bond purchases.
8. March 2012 at 10:15
I’m torn about this. I think your explanation of a ‘coded signal’ is plausible but I’m not sure it’s correct. I think the distinction between interest/non-interest bearing money is largely irrelevant to the issue though. With the exception of this “there’s no meaningful difference between interest bearing ERs and these short term repos.” Just like the IOR policy this will not push up inflation, i.e. it will not boost AD.
Rather than a deft ‘coded signal’ I think it’s far more likely that the Decision Makers at the Fed are stupid. The fact that Fed staff are intelligent does not matter one iota if that is the case.
8. March 2012 at 10:46
ssumner:
I get tired of commenters trying to bait me into talking about the banking system, when in fact the banking system doesn’t actually play an important role in monetary policy. It’s all about currency in circulation.
Currency in circulation is intimately tied up with the banking system. Indeed, a substantial portion of aggregate money stock ORIGINATES in the banking system!
The Fed’s monetary base is now almost $3 trillion. M2 (which isn’t even the most aggregated money stock measure) stands at almost $10 trillion. That difference is what the BANKING SYSTEM created by credit expansion and the concomitant demand deposit balances. That is a huge quantity of money from the banking system relative to the base money created by the Fed.
If we go by private measures of M3, which is about $15 trillion, then banking system effect inherent in “monetary policy”, i.e. inflation, is even more pronounced.
8. March 2012 at 11:03
That is a huge quantity of money from the banking system relative to the base money created by the Fed.
if by huge you mean in fact the lowest on record, sure that kind of huge – the ratio of m2 to base money is lower than at any time since fred has data. If by “huge” you mean the “I am completely divorced from reality and have been endlessly rebutted” kind of huge, yes, humongously gimormously huge.
the velocity of money is at the lowest point since the depression.
http://research.stlouisfed.org/fredgraph.png?g=5yh
8. March 2012 at 11:04
“The objection would be resolved if monetary stimulus and fiscal stimulus actually operated in different ways. But almost all the economists commenting supposedly teach, research, and subscribe to theories arguing that both affect AD exactly the same way. Yet they have a magical belief that in the real world, they’re somehow different.”
Funny. I’d say it’s entirely reasonable to think they are different in the real world, but there may be bizarre *theories* in which they are magically the same.
8. March 2012 at 11:36
“Funny. I’d say it’s entirely reasonable to think they are different in the real world, but there may be bizarre *theories* in which they are magically the same.”
D R: What is bizarre is to claim to believe theories that they are magically the same, do research according to those theories, and teach that they are magically the same– and then behave differently in the real world.
I’m perfectly okay with economists acting in the real world as if they are different, but they have a responsibility to teach it that way, to research and publish explanations for why they are different.
They are not. Economists are behaving like creationist astrophysicists.
8. March 2012 at 11:37
Steve is right. What the Fed needs to do is to increase inflation expectations without increasing expectations of inflation. Or maybe it’s the other way around.
8. March 2012 at 11:40
This should be under the ‘Media in Wonderland’ thread (though someone seems to ‘believe in magic’). From my hometown newspaper’s econ writer;
http://seattletimes.nwsource.com/html/soundeconomywithjontalton/2017698307_women_and_economic_progress_in.html
‘…women produce some 80 percent of the food in Africa but own only 1 percent of the farmland. Women in the developing world are 5 times less likely than men to own land, and their farms tend to be smaller and less fertile.
‘Women are less like to have access to financing to make their farms more productive.’
8. March 2012 at 12:13
Major Freedom:
The price of a dollar deposit in a savings account at JP Morgan is fixed at one dollar by JP Morgan’s contract to pay it off. The dollar isn’t defined in terms of deposits at JP Morgan. For the price of a one dollar deposit at JP Morgan to stay at one dollar, JP Morgan must adjust the quantity supplied to the demand. In equilbrium, the quantity of each and every type of private money must adjust to the demand to hold it. Othwise, the price of it would change–it would trade at a discount or premium. But that would be in violation of the contract, and so the quantity adjusts so that no premium of discount is necessary.
Base money, on the other hand, is not fixed by contract in that way. The Fed isn’t obligated to give people JP Morgan deosits of Bank of America deposits in exchange for currency or reserve balances.
The price level must adjust so that the real quantity of base money is equal to the real demand to hold it. Or, conversely, the price level must adjust so that the nominal demand to hold it matches the nominal quantity.
That is the argument for why it is currency (and reserve balances) that count for the determiniation of the price level. The market process that causes the price level to make the needed adjustments is changes in nominal spending on output. And so, the idea is that if the nominal quantity of base money is adjusted properly, the nominal spending on output can remain on target. In equilibirum, what that requires is that the nominal quantity of base money adjust with changes in the nominal demand for base money, though that nominal demand depends on expecations about nominal expenditures on output.
I think this all pretty much correct, except the impact of disequlibrium works through the banking system. The immediate effect of excess supplies or demands for currency is deposits or withdrawals of currency from banks. And this generates excess supplies or demands for bank deposits.
It is only in equilibrium that the banks don’t matter.
Where Scott and I agree is that this isn’t about getting the banks to lend more so that people can borrow more and spend more.
Your error is that you think it is all about trying to get banks to lend more, and you think that is bad because it isn’t matched by “real saving.” For the most part, it is just confusion.
8. March 2012 at 12:29
Bill Woolsey – If the Fed isn’t committing new resources (base) to fund the bond purchases, then how does it creat money? It simply shifts monetary resources *already in the system* in a zero sum way. This won’t boost expected future NGDP unless the process in some way boosts velocity, which is a stretch, to say the least. Usually I agree with you but in this one instance we diverge.
8. March 2012 at 12:29
I saw a quote from a Fed inflation hawk that bond purchases are bad because it locks the Fed into a low interest rate policy, which is inflationary since it prevents the Fed from responding to inflation by raising rates.
This is, in my view, backwards. Buying bonds locks the Fed into a low inflation policy. Only with low inflation can they keep the short term rate low, and thus avoid losing money on the purchases.
In any case, the Fed doesn’t want investors to buy treasuries. They want investors to buy risky bonds, stocks, houses, etc. So it’s a confusing “do as we say, not as we do” message.
8. March 2012 at 13:18
Repurchase agreements are money, and reverse repurchase agreements are creating money.
If the Fed offered checking accounts to households and firms directly, it would be directly creating money.
The firms that end up with the repurchasing agreements have more money, and no one has any less.
Of course, repurchase agreements are not included in any of the official statistics. To the degree that this increase in the quantity of money raises spending on output, then the velocity for MZM or M2 would increase. But the quantity of a measure of the equantity of money that includes repurchase agreements might not rise. The quantity rises.
8. March 2012 at 13:25
Bill Woolsey:
The price of a dollar deposit in a savings account at JP Morgan is fixed at one dollar by JP Morgan’s contract to pay it off. The dollar isn’t defined in terms of deposits at JP Morgan. For the price of a one dollar deposit at JP Morgan to stay at one dollar, JP Morgan must adjust the quantity supplied to the demand. In equilbrium, the quantity of each and every type of private money must adjust to the demand to hold it. Othwise, the price of it would change-it would trade at a discount or premium. But that would be in violation of the contract, and so the quantity adjusts so that no premium of discount is necessary.
I am referring to property rights to dollars. In terms of property rights, banks do increase the number of property rights titles, and those titles circulate as money.
Base money, on the other hand, is not fixed by contract in that way. The Fed isn’t obligated to give people JP Morgan deosits of Bank of America deposits in exchange for currency or reserve balances.
Irrelevant. The point is that the Fed does give the banks base money created out of nothing.
The price level must adjust so that the real quantity of base money is equal to the real demand to hold it. Or, conversely, the price level must adjust so that the nominal demand to hold it matches the nominal quantity.
The demand is primarily a function of how much exists. The more exists, the greater the nominal demand, ceteris paribus. Or, more accurately, the only explanation for more demand in terms of money, is more money in existence.
That is the argument for why it is currency (and reserve balances) that count for the determiniation of the price level. The market process that causes the price level to make the needed adjustments is changes in nominal spending on output. And so, the idea is that if the nominal quantity of base money is adjusted properly, the nominal spending on output can remain on target. In equilibirum, what that requires is that the nominal quantity of base money adjust with changes in the nominal demand for base money, though that nominal demand depends on expecations about nominal expenditures on output.
The target is arbitrary.
I think this all pretty much correct, except the impact of disequlibrium works through the banking system. The immediate effect of excess supplies or demands for currency is deposits or withdrawals of currency from banks. And this generates excess supplies or demands for bank deposits.
There is no such thing as “excess” demand or supply of money. It presupposes a reference point that is itself as arbitrary as the term “excess”.
It is only in equilibrium that the banks don’t matter.
The market is never in equilibrium, so the banks always matter.
Where Scott and I agree is that this isn’t about getting the banks to lend more so that people can borrow more and spend more.
That’s inevitable if the Fed prints money and gives it to the banks. The Fed is a banker’s institution.
Your error is that you think it is all about trying to get banks to lend more, and you think that is bad because it isn’t matched by “real saving.” For the most part, it is just confusion.
What do you mean by I think “it” is all about trying to get banks to lend more? Do you mean NGDP targeting? No, that’s not what I am saying. I am saying that much of the money the Fed creates in the real world does enter the banking system first, and so the policy of Fed easing does require the assumption of more bank lending to boost NGDP.
Your confusion is that you believe I think that banks are the ONLY institutions that can receive Fed money.
8. March 2012 at 13:29
dwb:
“That is a huge quantity of money from the banking system relative to the base money created by the Fed.”
if by huge you mean in fact the lowest on record, sure that kind of huge – the ratio of m2 to base money is lower than at any time since fred has data. If by “huge” you mean the “I am completely divorced from reality and have been endlessly rebutted” kind of huge, yes, humongously gimormously huge.
I mean neither. I mean “huge” relative to Sumner’s comment that the banking system is primarily irrelevant to monetary policy as it pertains to NGDP. I am not talking about history and I am not talking about whatever that second part is that you’re babbling on about.
If the banking system is responsible for the difference between $15 trillion total money and $3 base money from the Fed, then it is wrong to say that the banking system is primarily irrelevant to monetary policy as it pertains to NGDP. Clearly the banking system is responsible for a very large portion of total money.
the velocity of money is at the lowest point since the depression.
8. March 2012 at 13:48
@Bill Woolsey
My understanding is that with the reverse repurchases, the Fed would sell short term bills with the agreement to buy them back at a later date (typically a couple days, but the expecation would be to roll them over i guess). Repurchase agreements add reserves (then later drain them) while reverse repurchase agreements drain them (and later add them), so i think the idea here with this plan is to drain the reserves from the system created by buying the long term bonds on a dollar-for-dollar basis (and keep rolling over the reverse repos, i guess??). I dont think this will have much different effect than operation twist. Maybe the promise to continue to spin the hamster wheel is worth something, but i dont see how this is going to get us anywhere.
8. March 2012 at 13:54
@Scott
@Bill Woolsey
“Where Scott and I agree is that this isn’t about getting the banks to lend more so that people can borrow more and spend more.”
Labor markets function better under stable monetary environment. Monetary instability has caused unemployment.
Credit markets function better under stable monetary environment. Monetary instability has caused a suboptimal reduction of credit/GDP ratio.
Monetary policy is about changing the money supply to keep NGDP along a stable path.
Stable NGDP path is about getting employers to hire people, and it is about getting people to become employed. Stable NGDP path is about getting banks to lend more and it is about getting borrowers to borrow more.
8. March 2012 at 14:13
Corollary. The Fed could correct the labor market imbalances by hiring people. This would be wrong, as the Fed does not know the right price of labor. The Fed should print money instead.
The Fed could correct the credit market imbalances by making loans. This would be wrong, as the Fed does not know the right price of credit. The Fed should print money instead.
The ECB is smart and humble. 3 year LTRO’s do not set the price of credit. The interest rate on LTRO loans will be determined afterwards, taking into the account the macro developments. The only thing the ECB knows is the required degree of overcollateralization. The purpose of 3 year LTRO’s is to print money.
The Fed is dumb and arrogant. The Fed thinks it knows the right price of long term government credit. The purpose of the proposed third novel approach is to manipulate the government credit, and to “twist” the liquidity of newly created liabilities. On one hand, these liabilities would be a little bit less liquid if they will have one week duration. On the other hand, these liabilities would be more liquid as non-bank institutions would be able to hold reverse repos.
8. March 2012 at 14:13
John, Close, I left a comment over there.
dwb, I get frustrated too.
Bill. You said;
“More importantly, if the goal is some level of nominal GDP in the future, the Sumner/Krugman thought experiment of promising to leave some particular amount of base money (interest bearing or not) outstanding (say, doubling) after the zero bound goes away is _pointless._
Base money is endogenous and base money should be endogenous.”
I agree that with a proper monetary regime (NGDP futures targeting) the base would be endogenous, and that this would be desirable. But the base obviously isn’t endogenous right now–just consider QE2.
My point about currency is that if we fully replaced the base with currency, and eliminated reserve requirements, and got positive interest rates again via much higher NGDP, the banking system would hold very little base money, just a tiny bit of vault cash. You can complain all you want that nothing fundamental would change, and I agree. But it makes it easier to explain to others that monetary policy is all about controlling the supply and demand for base money, not the banking system and loans, securities, deposits, etc.
I completely agree that the Fed should not target currency in circulation.
DR, If they are both demand side policies then the answer is yes. If one is a supply-side policy, then yes, you can control the split. But the confusion I’m referring to is over demand-side policies. The British press is saying more AD would be bad (inflation) and less AD would be bad (recession).
Randy, Thanks for that link. Those comments sound reasonable.
Steve, That’s very interesting. Might that reflect the growing view that technological improvements are going to allow much more oil production than previously expected?
Cthorm, I mostly agree with you, I didn’t mean to suggest the coded signal approach was all that likely to work, just that it might work.
Major Freeman, All those ERs sitting in banks are more like T-bills than currency. They don’t really have much impact on the economy.
DR, You said;
“Funny. I’d say it’s entirely reasonable to think they are different in the real world, but there may be bizarre *theories* in which they are magically the same.”
Interestingly, most of the theories I’ve seen predict the fiscal stimulus will be the more inflationary (as the private sector is more efficient), whereas monetary stimulus will tend to boost real growth. But commentators tend to assume the opposite. Why?
John, You said;
“Economists are behaving like creationist astrophysicists.”
Good analogy.
Joe, Exactly!
Max, Good point.
123, That might be a positive result, I don’t know enough to know whether we currently have suboptimal lending.
8. March 2012 at 14:19
@Bill Woolsey
ok lets take the checking account example. I have a 10 yr UST in my checking account. I sell it to the Fed and get $1000 (leg 1, the “QE” part). Then the Fed enters into a reverse repo with me, where the fed sells me a 3mo Bill with the promise to buy it back a few days later (leg 2, the “sterilization” with a reverse repo). End of day, i have a 3mo Bill and a promise. All i’ve done is exchange the 10yr for the 3mo in my account, plus the promise from the Fed to buy the 3mo a few days from now. Sorry, i just dont see the Fed’s promise to buy back the 3mo Bill in a few days as worth much because 3 mo UST Bills are the most liquid instrument in the world, and 10 yr UST bonds are pretty close to being as liquid. I can make this exact exchange anytime i want, 24-7, nearly 365 days a year, with practically any bank in the world. I dont really see this as creating any money.
I think we are in agreement that the interest rate channel effect is negligible. maybe there is some signaling.
8. March 2012 at 14:35
Scott:”That might be a positive result, I don’t know enough to know whether we currently have suboptimal lending.”
After listening to some hawks, I don’t know enough enough to know whether we currently have suboptimal levels of labor market activity. But I do not deny that higher employment might be a positive result of NGDP targeting 🙂
But seriously, if you knew that the NGDP growth would be stable in the future, you would require lower risk premium on risky lending. And you would like to borrow more, as borrowing is especially uncomfortable during negative NGDP shocks.
8. March 2012 at 14:55
“Steve, That’s very interesting. Might that reflect the growing view that technological improvements are going to allow much more oil production than previously expected?”
Yes, more production and also tech improvements in the internal combustion engine. The implications get me all excited. The Fed is going to have to find something else to inflate: service wages, housing, manufacturing, etc.
8. March 2012 at 14:58
Scott,
“It’s all about currency in circulation. It would make more sense to treat those interest-bearing ERs as government debt, not “high powered money.” Money is only “high-powered” when it earns no interest. ”
Two questions –
1) Does currency in circulation include currency that is being horded as store of value and locked in a safety deposit box.
2) If Tbill yields become negative, what happens to the amount of currency locked in safety deposit boxes.
8. March 2012 at 15:14
dwb:
The Fed buys someone elses 10 year bonds for $10b billion, credits their bank’s reserve deposits, which credits their checkable deposit.
The Fed “sterilizes” that by borrowing $10 billion overnight. This is secured by an unspecified set of government bonds. Maybe including the 10 year bonds.
The Fed takes 10 billion out of the overnight lenders’ checking accounts. The Fed takes the funds out of the reserve balances of the lenders’ banks.
Checkable deposits and reserve balances are unchanged.
But the overnight lenders have money. A promise by the Fed pay money back tomorrow is the same thing as money in a checking account. The differences are that there are no reserve requirements on them. In this case, the lending is to the Fed and so the banks have no extra funds to lend.
By the way, you don’t give the Fed your T-bills. The Fed sells you the T-bills (and really it is just some government securities it holds) and promises to buy them back. You have the T-bills and you have given the Fed your money which it will pay back tomorrow.
What you have less of is money in your checking account. You have a perfect substitute. A promise by the Fed to pay you back tomorrow. But starting tomorrow, the Fed will pay you back today. Whenever you need the money.
If the Fed fails, you get to keep the government securities. Not that this matters one bit, but the receivers would sell these government securites and use it to pay you back.
So, the sellers of the 10 year bonds have new money. You have the same money as before, but rather than in the form of an official checking account, it is in the form of a overnight loan to the Fed.
Suppose the Fed sells 10 year Fed bonds (which don’t exist) and use the proceeds to buy 10 year Treasury bonds.
Sterilized, right?
Now, suppose the Fed sells 1 year Fed bills (which don’t exist) and use the proceeds to buy 10 year Treasury bonds. Under current conditions this is creating money, though nominally no.
Now, the Fed selles 1 month Fed bills and buys 10 year bonds>
One week Fed bills, and buys 10 year bonds.
One day Fed bills.. which is what they are doing.
Do I need to take it to one second Fed bills?
The Fed is borrowing overnight secured by government bonds. They care creating money.
8. March 2012 at 16:00
“Interestingly, most of the theories I’ve seen predict the fiscal stimulus will be the more inflationary (as the private sector is more efficient), whereas monetary stimulus will tend to boost real growth. But commentators tend to assume the opposite. Why?”
Because commentators are acting on [political] belief not scientific theory (like those creationist astrophysicists work on or teach one theory in their work, and believe something contradictory when off the clock), and most of them who favor AD stimulus lean left, and so favor fiscal stimulus over monetary. The exception that proves the rule: those who see the need for stimulus who do lean right, like me or you or Ramesh Ponnuru, strongly favor monetary stimulus over fiscal.
8. March 2012 at 16:18
Major Freedom:
The Fed doesn’t give money to banks.
I suspect that this error is the source of many others.
8. March 2012 at 16:38
@Bill Woolsey
i see your point, you are saying the promise for the Fed to pay you back is as good as money – it’s future money. the net effect tho depends on what the market expects the Fed to continue to do – i.e. roll over the rev repos or not. This is the kind of op that would be more effective if the repo market were dysfucntional because of bank credit issues, or if overnight lenders were having issues borrowing. none of which are the case.
right now, the down side could be worse than doing nothing because by running the hamster wheel creating the illusion of motion, if it turns out we have not gone much of anywhere they foster the illusion of impotence.
8. March 2012 at 20:56
@Bill Woolsey
But in your chain of events, at the end of the day what’s happened is $10bn of private sector assets have been shifted between different govt liability categories, all of which exhibit meaningful “moneyness” characteristics. In other words I’d agree that reverse repos aren’t that different than reserves, but I’d take it a step further and say neither are that different from the original $10bn of Treasuries, at least in turns of impact on private sector propensity to spend/inflation. What is different is their interest rate risk characteristics, which is what the Fed is really after anyway.
I also think Scott is giving “sophisticated investors” waaaaaay too much credit. HFs got all types of bulled up over monetary base expansion = hyperinflation trades. This was the reaction function the Fed didn’t anticipate and is trying to avoid in the future.
8. March 2012 at 21:36
The “monetary policy makes inflation but fiscal policy makes jobs” mentality goes right up through Congress. I was watching Bernanke’s semiannual report, wherein one of the congressmen asked Bernanke whether the Fed could narrow its focus to just inflation so that Congress could focus on jobs.
It’s everywhere, and frankly it’s a dangerous idea.
9. March 2012 at 01:01
Scott, I agree completely with your post, except with the characterization of “sterilizing” that you attribute to the fed paying interest on reserves. I would actually favor the fed charging a slightly negative interest on reserves, but it seems kind of arbitrary to characterize positive interest rates as “sterilizing”.
True S&M style sterilizing was done by the ECB under Trichet, which would not only pay an interest on reserves but also auction off term deposits corresponding to the amount of bonds they bought.
Also I don’t get why you call the Fed’s thinking “Keynesian”, I imagine that (new) Keynesians would completely agree with you that there is no point in sterilizing QE when the output gap is so huge.
9. March 2012 at 08:10
Bill Woolsey:
The Fed doesn’t give money to banks.
Yes, they do. Every time they purchase securities from the banking system, they are giving the banks new money that did not exist before.
9. March 2012 at 08:51
Major_Freedom,
I’m going to give you a Cheeseburger. That’ll be $10.
Yup, sounds like giving to me.
9. March 2012 at 13:15
@mf
I’ve got this great bridge on the East River, I’d like to give to you.
9. March 2012 at 21:19
Great post, Scott!
10. March 2012 at 04:47
Major Freedom:
When someone buys something for money, the money they pay is not a gift.
While the Fed can buy bonds from banks, not everyone they buy from directly is a bank, and all that they buy from directly are dealers–middlemen–and those who sell the bonds to the dealers are receiving new money that didn’t exist before in exchange for the bonds they used to hold.
Anyway, as I said before, that you characterize this as giving money to the banks is the probably the source of much of your confusion. I suspect that Rothbard and a few friends decided a long time ago that this would be a useful propaganda ploy in helping to promote a 100% reserve gold standard. We can say that the Fed is just giving money to the evil, rich bankers. Sure, it isn’t really true, but stupid people will get angry. And when their anarchist political officers order–storm the barricades–they will charge into the machine gun fire, overrun the counterrevolutionary police and army lines, and bring about victory for the anarchist revolution.
There are any number of elements of Rothbardian ideology that have the same rationale–motivating the revolutionary cannon fodder.
10. March 2012 at 05:11
dwb:
I don’t think the Fed’s promises to pay you back is future money and as good as present money. I think that it is money, especially if your excess checking acount balance at your bank is being swept into them every night.
jrp:
I think money is very different from 10 year bonds. What I will grant is that if the yields on 10 year bonds are driven down to zero or less, then money will be an all-around better than 10 year treasury bonds, and so purchasing any more beyond that point will not ease monetary disequilibrium.
Step one is an explicity committment to a growth path for nominal GDP.
Step two is to reduce short rates the tiny bit still possible–slightly negative interest rates on reserve balances at the Fed.
And finally, it is time to privatize the issue of hand to hand currency and get rid of this zero nominal bound permanently.
10. March 2012 at 06:50
“And finally, it is time to privatize the issue of hand to hand currency and get rid of this zero nominal bound permanently.”
You don’t have to get rid of government currency. Just auction currency the same way as treasury bonds, so it can trade at a premium to face value. It’s really a very simple and non-disruptive change.
10. March 2012 at 08:07
Bill Woolsey
When someone buys something for money, the money they pay is not a gift.
It is if they created it out of nothing.
If I printed money in my basement, and I bought goods and services from you, then you will receive higher prices than you otherwise could have gotten in a world with fair trade only. Therefore, the higher nominal income you get on account of my money printing, is in fact a free gift to you.
While the Fed can buy bonds from banks, not everyone they buy from directly is a bank, and all that they buy from directly are dealers-middlemen-and those who sell the bonds to the dealers are receiving new money that didn’t exist before in exchange for the bonds they used to hold.
Not everyone sells bonds to the Fed. Those who don’t sell bonds to the Fed, pay higher prices and lose that way.
Anyway, as I said before, that you characterize this as giving money to the banks is the probably the source of much of your confusion.
You haven’t shown how it is not a gift, and you haven’t shown how I am confused. You are confused, not me. You don’t realize that the Fed doesn’t even deal with everyone in the economy.
I suspect that Rothbard and a few friends decided a long time ago that this would be a useful propaganda ploy in helping to promote a 100% reserve gold standard.
Oh please. Spare me with this accusatory rabble rousing.
We can say that the Fed is just giving money to the evil, rich bankers.
Or we can say what Rothbardians say, that the Fed is giving free money to only a select few market participants before everyone else.
Sure, it isn’t really true, but stupid people will get angry. And when their anarchist political officers order-storm the barricades-they will charge into the machine gun fire, overrun the counterrevolutionary police and army lines, and bring about victory for the anarchist revolution.
Not really interested in your paranoia and emotive rationalizing.
There are any number of elements of Rothbardian ideology that have the same rationale-motivating the revolutionary cannon fodder.
All ideologies that are anti-status quo are revolutionary.
10. March 2012 at 08:07
Bill Woolsey
When someone buys something for money, the money they pay is not a gift.
It is if they created it out of nothing.
If I printed money in my basement, and I bought goods and services from you, then you will receive higher prices than you otherwise could have gotten in a world with fair trade only. Therefore, the higher nominal income you get on account of my money printing, is in fact a free gift to you.
While the Fed can buy bonds from banks, not everyone they buy from directly is a bank, and all that they buy from directly are dealers-middlemen-and those who sell the bonds to the dealers are receiving new money that didn’t exist before in exchange for the bonds they used to hold.
Not everyone sells bonds to the Fed. Those who don’t sell bonds to the Fed, pay higher prices and lose that way.
Anyway, as I said before, that you characterize this as giving money to the banks is the probably the source of much of your confusion.
You haven’t shown how it is not a gift, and you haven’t shown how I am confused. You are confused, not me. You don’t realize that the Fed doesn’t even deal with everyone in the economy.
I suspect that Rothbard and a few friends decided a long time ago that this would be a useful propaganda ploy in helping to promote a 100% reserve gold standard.
Oh please. Spare me with this accusatory rabble rousing.
We can say that the Fed is just giving money to the evil, rich bankers.
Or we can say what Rothbardians say, that the Fed is giving free money to only a select few market participants before everyone else.
Sure, it isn’t really true, but stupid people will get angry. And when their anarchist political officers order-storm the barricades-they will charge into the machine gun fire, overrun the counterrevolutionary police and army lines, and bring about victory for the anarchist revolution.
Not really interested in your paranoia and emotive rationalizing.
There are any number of elements of Rothbardian ideology that have the same rationale-motivating the revolutionary cannon fodder.
All ideologies that are anti-status quo are revoolutionary.
10. March 2012 at 08:09
Cthorm:
I’m going to give you a Cheeseburger. That’ll be $10.
Yup, sounds like giving to me.
It is if I printed that money in my basement. If I didn’t, then you’d have to accept a lower price in the market that would have less money and spending. Booya.
10. March 2012 at 09:10
steve, I hope your are right.
dtoh,
1. Yes. 2 There’d be more currency in lock boxes.
John Thacker, Good point.
Integral, That’s sad.
orionorbit, The thinking at the Fed is “Keynesian” in the sense that it focuses on interest rates as the instrument and/or indicator of policy.
IOR is “sterilization” in the sense of trying to prevent new money from having an expansionary effect.
Thanks Doc Merlin.
15. March 2012 at 22:47
[…] Scott Sumner: Does the Fed believe in magic? […]