The “Liquidity Trap” Theory Was Never Right….

Apparently there’s some confusion over the title of my last post.  Perhaps I should have been clearer.  I am aware that Dr. Krugman and others have been claiming there are no bond vigilantes in the USA.  But not for the right reasons.  Maybe I am jumping to conclusions, but there appears to have been some move in the right direction away from this idea that the “liquidity preference” was of central importance to understanding the rate debate and the impact of bond vigilantes….After all, even Dr. Krugman has mentioned that he “rethought” his model.  And late last year he clearly defined how Japan’s ability to issue its own currency made it different from countries who were users of currency.

Anyhow, most of the mainstream has been working under some form of the “liquidity preference” theory or “liquidity trap” theory to explain why rates have remained low in the USA.  This was the thinking that economic agents were choosing to just sit on the excess liquidity provided via the government and the Fed.  It sounds right in theory, but it’s wrong in reality.

First, the government doesn’t increase the money supply through government spending (already confused?  See here).  It redistributes inside money (bank money) and adds a net financial asset in the form of a government bond.  This improves private balance sheets and is particularly useful during a balance sheet recession, but it’s not the equivalent of firing dollar bills out into the economy (though it does increase the velocity of spending as the government becomes the “spender of last resort” when an economy is stagnant for whatever reason).  Second, when the Fed implements QE they swap reserves for bonds.  No change in net financial assets.  And since banks don’t lend reserves there is no firm transmission mechanism through which this policy can impact the money supply.   In short, fiscal policy hasn’t been the equivalent of increasing the money supply in the traditional “money printing” sense that most believe and QE has most certainly not resulted in an increase in the money supply because the primary transmission mechanism is busted (the lending channel).

So, how does the money supply primarily increase?  It increases primarily when banks make loans which create deposits.  And why has this mechanism been broken?  Because private actors were saddled with debt from the credit bubble and no longer had the incomes to service these debts when the bubble burst.  Private actors weren’t choosing to sit on their liquid balances.  They had no choice either because they didn’t have the incomes to meet their debt obligations or they were banks who ALWAYS sit on their reserves and don’t lend them out…The liquidity trap theory was never right.  And the liquidity preference argument for bond vigilantes was also never right….

* In case you’re still wondering, a liquidity trap is essentially an environment in which economic agents choose to sit on their liquid money balances as opposed to spend or lend them.  The theory posits that monetary policy becomes ineffective in such an environment.  I agree that monetary policy has become ineffective, but not because of the liquidity trap….


Got a comment or question about this post? Feel free to use the Ask Cullen section, leave a comment in the forum or send me a message on Twitter.

Cullen Roche

Mr. Roche is the Founder of Orcam Financial Group, LLC. Orcam is a financial services firm offering research, private advisory, institutional consulting and educational services.

More Posts - Website

Follow Me:

  • LVG

    Yes, the LQ theory says monetary policy stops working because we sit on excess money balances. This is based on a misunderstanding of banking. I don’t agree that Krugman is making progress though. The recent debate with Keen exposed a totally flawed view of banking.

  • LVG

    LT theory. Sorry.

  • SS

    Cullen, I think you’re reading into PK’s position changes a bit too much.

  • Cullen Roche

    Yes. Monetary policy was never ineffective due to a liquidity preference. It was ineffective due to the fact that banks don’t lend their reserves and QE didn’t change the net financial asset position of the private sector.

  • Cullen Roche


  • Anon

    So does this:

    Not show north of 2 trillion in savings, still abnormally high? I.e. liquidity sitting idle – a “savings glut”?

  • Anon

    Or at minimum this:

    Showing half a trillion dollars, unspent?

    Corporations also have close to 2 trillion dollars in excess cash-equivalents. Those are certainly “real” – and could be seen as “trapped liquidity at the zero rate boundary”?

  • Huh

    “because the primary transmission mechanism is busted (the lending channel).”

    Liquidity trap is more of an interest rate story, specifically the economic model itself is ENTIRELY about interest rates and does not have any mechanism for reserves > lending, it really is quite a trivial claim, it just says once interest rates are at or near zero the central bank cannot make rates any lower and doing more open market operations wont push them much lower so monetary policy is ineffective. Which part of that do you disagree?

    Also isnt the statement “banks do not lend reserves” a tautological fallacy? The definition of a reserve is money not lent, so a lent reserve cannot exist. But that does not mean banks do not lend money (instead of creating money out of thin air, not procuring funds from anywhere first).

  • BT London

    Many people group the liquidity trap with the balance sheet recession and private sector deleveraging as all the same idea. I think Krugman is one of these people.

    But Krugman likes to use the word liquidity trap because he claims to have ‘started the whole liquidity trap literature’ back with his work on Japan.

    Now some of Krugman’s ideas have recently shifted – particularly on currency issuing countries versus say eurozone countries – in the direction of post-keynesians and MR as Cullen points out. That’s a good thing.

    It’s a bit harsh to say the liquidity trap idea was never right, as it is conceptually similar to private sector deleveraging in many ways, except that it is couched in the language of ISLM instead of the language of monetary accounting.

    So in the end this argument is about ISLM being flawed, even though it can actually come up with the correct policy responses to the current crisis.

    Minor quibble – your line about gov spending not increasing the money supply might confuse people. Bond issue creates both a debt (the bond) and a credit (deposits in the account of the recipient of government spending) – correct?

  • BT London

    Re: reserves.

    Cash is only a tiny tiny proportion of total credit on bank balance sheets. In the modern world most transactions are credit transfers between accounts, not cash transactions.

    This means the focus on cash as a means of liquidity is no longer a useful framework. Reserves just serve to settle balances between banks and the central bank/treasury. The main means of central bank policy is via manipulation of the rate of interest on bonds, not via the quantity of reserves as some limitation on cash liquidity.

    Better to ditch the money multiplier and just talk about loans creating credit and repayment destroying credit. Simon Wren-Lewis seems to agree:

  • charlesd

    Hi Cullen
    You correctly explain that the money supply is not increased because of deficit spending. However, you do explain that Treasury Bonds are created “ex nihilo” for the private sector as a result of deficit spending. Consequently, savings in the form of “long duration” “near-cash” has been increased as a result of deficit spending, even though this is not part of most definitions of the “money supply”. Is my understanding basically correct? Thanks much.

  • Romeo Fayette

    I think I disagree or misunderstand…

    The Treasury auctions bonds to Primary Dealers, who pay with inside money. (Technically, this is a bank making a loan to the Treasury, which creates a deposit that’ll all soon be assumed by the Fed.) The Fed buys the T-bonds from PDs at a slight markup, which profit is a modest stimulus to the banking system.

    The net effect of this tri-party exchange is the Treasury has financed its spending (increased its leverage and its spending capacity). The PD has essentially collected a fee for acting as an intermediary. The Fed has made a roundabout loan to the Treasury (swapped a T-bond for financing).

    The key here is how the Fed pays the PD for the T-bond. Sure, it “swaps reserves for bonds,” as you say, but is the Fed creating new reserves to do so? I don’t think the Fed borrows reserves from Bank A’s account to pay Bank B for a T-bond transaction. That’d be rehypothecation, which is hardly different than creating the reserve ex nilho. That’s the shark you jump, which is where I think I disagree.

  • Romeo Fayette

    Reserves are merely the form of settlement payment among banks & the Treasury/Fed. I get that, but what I’m saying is that the Treasury has more spending power at the end of QE than it would’ve had otherwise. That spending works into the real economy, and I think the Fed creates reserves ex nilho (collateralized by its T-bond assets), rather than shuffling reserves among Primary Dealers’ excess reserve accounts. The latter “shuffling” seems to be your suggestion, so I’d be interested to hear what you guys have to say about that, Cullen included…?

  • hangemhi

    I think I get it now… the 3rd paragraph “First, the government doesn’t increase the money supply through government spending” isn’t something I quite got before. That’s a big diff with your former school (won’t name it since it always gets my comments spam controlled) and MR. I don’t recall seeing it stated this clearly – that deficit spending isn’t new money – or printing.

    I wonder if this is correct: That banks create (most of) the money, and the gov borrows it and spends it back in, and the gov only gets the money to borrow if the private sector has borrowed enough money from banks for the gov to then re-borrow it. That, plus the private sector keeps getting its money handed back to them via deficit spending, which the biggest savers let the gov re-borrow it again.

    It still feels a bit semantic-y to me – meaning, deficit spending sure looks like new money and printing. But at least I get it now

  • Cullen Roche

    It’s a very technical point, but operationally accurate. Deficit spending creates NFA, but redistributes money.

    What Krugman would say in response to something like my comment is that bonds and money are near equivalents in a liquidity trap. I would say wrong. The problem is not that bonds and money are near equivalents, but in our current environment monetary policy is broken and no money is being created (because the credit mechanism is broken). Therefore, the liquidity trap idea is irrelevant. It’s not that bonds and money are near equivalents. It’s that there is no money creation via fiscal or monetary policy to begin with. Fiscal redistributes existing money. And monetary policy isn’t reslulting in more lending. Hence, no money creation. A NFA add via deficit spending which eases the burden, but this is a totally separate understanding from the idea of the liquidity trap.

  • Cullen Roche

    I don’t think you’re quite connecting the dots. QE is monetary policy. Deficit spending is fiscal. Think of them as two separate things. The monetization argument confuses the two. When QE occurs there has already been a predetermined amount of deficit spending. These bonds were going to get sold whether the Fed bought them or not because the PD’s were required to buy them. So QE doesn’t enable deficit spending. What QE does is swap reserves for bonds in an attempt to alter monetary policy. These are two very different things.

  • Cullen Roche

    Charles, yes, that is correct. Bonds are securities. These are less liquid money-like instruments whose price can vary substantially.

  • Cullen Roche

    Govt spending redistributes deposits and adds a NFA in the bond. So, before deficit spending the pvt sector has a deposit and no bonds. After deficit spending the pvt sector has a deposit AND a new bond. Make sense?

  • Cullen Roche

    No, bank reserves are reserves held on deposit at the Fed. They cannot be spent. They do not serve any purpose in the real economy except to meet reserve requirements and settle interbank payments. It’s not tautology. Most people just don’t know what reserve are to begin with. Especially economists who don’t understand banking.

  • GLG34

    This is a very interesting comment, Cullen. I think something with MR is starting to click with me. What you’ve done is really excellent. In essence, you’ve proven how the neoclassical obsession with the government creating money is wrong. It is the private sector that creates money. That’s the whole problem today.

  • charlesd

    Thanks Cullen and all for your time and comments.
    While I can’t say I understand all the operational aspects,
    you answered the “big picture” if I understand correctly.
    That is, when the government deficits spends, in effect,
    the money taken from the private sector (to buy the bonds)
    is handed back to the private sector as the deficit “payment” (lower taxes or higher spending). So this is a wash. However, the private sector also has now has the bonds which it did not have before. Thus fiscal policy is
    creating new private sector financial assets in the form of Treasury securities. This is not pure “money”, as Cullen
    has explained, but it is new private sector financial wealth
    created by fiscal actions, not by the private sector. I apologize if I seem to be repeating in a different way what Cullen has already explained (many times I’m sure!)
    but this is what I think the “big picture” looks like
    regarding the impact of deficit spending on financial assets.

  • Cullen Roche

    Charles, precisely right!

  • Andrew P

    QE may affect interest rates, but it is hard to say exactly how it does so because the short term movements after QE is announced tend to be upward ticks in rates. Perhaps, QE holds rates down for the long term because the dealers know that the Fed is making sure the Government is financed at an acceptable interest cost.

  • Andrew P

    If the “debt ceiling” hits the wall and Obama does the trillion dollar platinum coin thingy, doesn’t that create new money? What if the Congress then realizes they can spend bigger deficits without serious inflation and goes on to double the deficit to $2T, and finances it by coin seigniorage. Isn’t that creating new money?

  • Johnny Evers

    You get three of the four right.
    1. Private sector gives money to buy bonds.
    2. Private sector gets government spending.
    3. Private sector gets a financial asset.
    But you leave out 4. Eventually a) future private sector must buy back the bond, or b) Fed buys the bond (monetization.)
    MR assumes that growth and inflation will buy back the bond or that an endless supply of future savers will be available or that the Fed can carry an unlimited amount of bonds on its balance sheet.
    So MR describes what is happening, but doesn’t address what 4) might be, other than assurances that everything will be fine because we don’t currently have inflation or aren’t degrading the standard of living(in their view.)

  • SS

    As long as there is a desire to save in US dollars someone will always be willing to exchange non-interest earning cash for interest earning bonds.

  • BT London

    I’m confused. You’re saying bank loans to the private sector create both debt (the loan agreement) and credit (deposits). But bank loans to the government create debt (the bond) but not credit (deposits).

    Surely, since credit and debt have to equal due to double entry accounting, the sale of bonds via primary dealers must involve credit creation.

  • hangemhi

    how does #4 happen when the deficit spending flows right back into the hands of the largest savers? Like the atmosphere evaporating water from the ocean, raining on the mountain tops, it all/most ends up right back in the ocean for the atmosphere to re-evaporate. Somehow the money has to literally disappear for there not to be money to re-buy the bonds. If I’m misunderstanding your point, I’d like more clarification, but it seems to me that deficit spending allows for more bond buying, and around it goes. The largest savers are so bloody rich that they don’t need their money back. We could always tax the hell out of them and make them poor, but then taxes pay down the debt.

  • Dan M.

    I know I have hit on this before, but the assertion that banks are not reserve constrained seems partially flawed to me. When a bank approaches its reserve equirement, it can get it in a few ways (right)?

    1) attract more non-lent deposits… Basically, reserves. If you can attract someone’s deposit without making a loan you got more reserves. This involves paying higher interest to attract them.

    2) borrow reserves from other banks (at interest)

    3) borrow reserves from the fed (at interest)

    Let me know if these assertions are flawed. So isn’t this a sign that there IS in fact some reserve constrsaint? If I have the money to take a private jet from LA to New York, this doesn’t mean that I am not constrained in doing so, but simply that I have the ability to do so at some cost.

    So please let me know where I’m wrong here. I think there is a constraint there, of sorts, and that we might be oversimplifying the power of the banks to essentially make money with no multiplier-esque constraints.

    Thanks everyone.

  • jt26

    Cullen, similar to liquidity preference could you comment specifically on whether the monetary base has any meaning? As you know, monetarists are obsessed with it.

  • jt26

    Cullen, dumb question, just to clarify …w.r.t. *excess* reserves …
    Banks can loan these if they wanted, right? (One could argue with IOER, they are doing it now?)

  • charlesd

    Regarding “numbers 1-4″ on the impact of deficits. The private sector, as agreed, ends up with a net credit in the form of Treasury bonds. Since there must be a debit for every credit, the public sector has the offsetting debit, i.e, the “debt”. However, as I understand it, this “debt” is no burden on anyone, now or in the future. For example, if so desired, the government (the Fed I guess) could, in principle, create money “ex nihilo” to retire the debt.
    (There may be legal restrictions from doing so with which
    I am not familiar but we are talking economics, not law).
    While this would be “money printing”, no financial assets would be added to the private sector because the private sector loses the bonds. The composition of private sector assets would change, but not the amount. And the debt is gone!
    The “platinum coin” idea being discussed in various places is a “smaller” version of this concept. If the government created a, say, $2 trillion platinum coin, this
    money created “ex nihilo” could be drawn upon to reduce the
    debt by $2 trillion. And no financial assets would be added to the private sector because the private sector loses the bonds. And the “debt ceiling” issue goes away for now.

    I believe Cullen has explained that eliminating the “debt” would not be desirable. The Treasury market serves several purposes: (a) it is used in the implementation of important aspects of monetary policy (b) there are many constituencies who wish to own Treasuries and (c) Treasury yields are a useful benchmark for other fixed income securities.
    But the point is that, in principle, the debt could be eliminated and therefore is no burden on the future. Indeed, if my understanding is correct, the “National Debt” is nothing more than a private sector savings account which represents part of our wealth. It represents the cumulative amount of financial assets (Treasuries) which have been created in the past as a result of the normal consequences of deficit spending as described in Johnny Ever’s steps 1-3 (above).

  • Linda G

    When I think of a “deposit account” being created, however it comes about – whether via a loan agreement or someone bringing in cash for deposit, etc. – I think of a “claim on money” being created.

    It’s that claim that shows up as a bank liability, while the cash or loan agreement, etc., shows up as a bank asset.

    When a customer spends from their bank account, they exercise that claim on money.

    Only then does the bank have to get the money to fulfill the claim then being exercised.

    At that point, the bank can obtain the money from a few different sources: from money that other customers have deposited (the cheapest route to take), form borrowing money from other banks, or from borrowing money from the Fed.

    Because that last option is available (borrowing money from the Fed), the potential to “run out of money” to fulfill a “claim on money” being exercised does not exist for a bank within the Federal Reserve system, at least in the U.S. and in countries with similar monetary systems.

    And when the customer spends money upon the exercise of his/her/its “claim on money,” he/she/it can do so in more than one form: via cash or electronically. And one way or another, the bank then needs to come up with the money for the customer to spend at the time that he/she/it demands that money, or exercises their “claim on money,” from the bank… whether the bank then does so from another customer, another bank, or the Fed.

    Anyway, I find that to be a helpful step in the explanation of “creation of deposits” in banking that occurs with all forms of deposits – that a “claim on money” has been created, which has the potential, from that point on, to be exercised… whenever a customer spends (or makes a transfer to another bank) from their deposit account.

  • Cullen Roche

    Well, reserves are an asset of the bank so if you want to claim they’re reserve constrained because they’re asset constrained (really capital constrained) then be my guest. But the argument is against the standard money multiplier so it’s a very different point.

  • Linda G

    A bit of a revision to my statement above:

    When I wrote of one route that banks have for obtaining money to fulfill the exercise of a “claim on money,” I shouldn’t say “from money that *other* customers have deposited,” but simply “from money that customers have deposited.”

    I also should add that, of course, cash has the constraint of physical/logistical limitations, and so spending money in electronic form is, of course, by far the most common way to do so.

  • Cullen Roche

    Excess reserves are just reserves in excess of required. There’s nothing special about them really.

  • Johnny Evers

    The rain analogy forgets about 4)
    Consider: Water evaporates (savers buy bonds); savers get a financial asset (clouds form); rain falls to earth (government spends), clouds dissipate (um, bonds or net financial assets remain.
    What happens with all those ‘clouds’. Do they eventually rain? Or do they just continue to form, blotting out the sun.

    As to SS’s point, who says there will always be enough savers to exchange their savings for bonds? We are seeing in Europe and Japan that either savers are afraid of the risks or just not there anymore, so central banks are stepping in.
    We don’t have the problem, here, because we’re sovereign, which means we’re going to monetize the debt.
    Which is what it comes down to — let’s talk about what happens when the debt is monetized and if that’s a good idea to finance government spending.

  • Cullen Roche

    That’s a good explanation. But I also think that explanation might misinterpret the purpose of the Fed system so it’s important to be clear here. It might even lead some to believe that the govt creates all of the money in the system. The truth is that the Fed system exists to help settle interbank payments. If we had one big banking system under one bank there would be no need for this settlement process. All money would be the same in essence. The Fed helps to facilitate the existence of private competitive banks who create money. Thinking of bank loans as a claim on government money creates the impression that banks don’t actually create money. That’s not accurate. The govt plays an important facilitating role in helping the banks operate smoothly, but it’s important to get the order of importance right. Banks rule the monetary roost. The govt supports through its various powers.

  • Johnny Evers

    We’re not Greece because we can — and will — print money to redeem debt and facilitate more borrowing.
    Krugman knows this, but the minute he writes that in the New York Times, then political hell breaks out.
    The plan is to monetize debt quietly, our of view of the public, to deny it is taking place.

  • Cullen Roche

    No, he actually just realized that in the last year or so. He’s even admitted that his model changed.

  • Cowpoke

    “First, the government doesn’t increase the money supply through government spending (already confused? See here). It redistributes inside money (bank money) and adds a net financial asset in the form of a government bond. ”

    The WSJ has a neat little tool that you can play with that shows how you can redistribute and what happens to the deficit by taxing more or less. It’s called “Make Your Own Deficit-Reduction Plan”

    I was able to take our govt’s 1 trillion dollar deficit to a 1.4 Trillion dollar surplus.
    Problem I saw though was that I pretty much had to redistribute every bit of taxes and gut funding for almost all programs to put the US govt in the Green.
    I think I like it better when the govt is in the red.

  • Linda G

    Thanks, Cullen. I think I’m beginning to see what you’re talking about (beginning to):


    If I enter into a loan agreement with Bank A, then that would create a loan agreement (bank asset) and a “claim of X amount of dollars” in my account (bank liability).

    If I wanted to then make a payment/transfer to my sister, also a customer at Bank A, then the only thing Bank A would need to do would be to lessen the “claim of dollars” in my account and to increase the “claim of dollars” in my sister’s account. No other asset for the bank would be needed to complete that transfer.

    And if there was only one bank for transaction purposes, then that’s all that would be needed, that bank’s created assets via loan agreements.


    Only if I were to exercise my “claim of X amount of dollars” in the form of demanding cash or in the form of payment to a customer at another bank would Bank A need to borrow (from customers, from other banks, or from the Fed).

    In those cases, Bank A’s created asset of a loan agreement would not be able applicable. Other assets would be needed, whether bank notes or electronic reserves (the definition of “reserves” can be confusing to me), in order to meet those demands for payments.


    Still, with the Federal Reserve in play via lending at the Fed level (when needed), there is the definite sense for me of the tail (private banks) wagging the dog (the Fed), as Steve Keen puts it.

    That’s what I get from the “order of things” inherent in private bank lending… from the understanding that private banks lend first, without regard to reserves, and then borrow later (if needed – when payments between banks or cash payments are called for, but not when payments between customers within the same bank are called for)… and with the understanding that banks can pretty much borrow as much as they need to with the Fed always available as lender of last resort.


    That brings up a question for me regarding interbank lending. Do banks need to lend from excess reserves in that case? Or is that lending done in the same way as lending to customers is done?

  • Jonathon McKitrick

    “First, the government doesn’t increase the money supply through government spending”

    Yes, I’m confused. I thought that was exactly what the government does when the accounting entry is made and money is produced from ‘thin air’ for use by the Treasury….

  • Cullen Roche

    Govt spending is a redistribution of inside money. Taxes take from Peter to pay Paul. Bond sales raise money from Peter to pay Paul. BUT, when the govt runs a budget deficit (spending in excess of tax receipts) it finances this spending by issuing a bond. That means the private sector obtains a bond AND a deposit. The deposit is a redistribution of money. The bond is a brand new net financial asset. So, the “money printing” (if one were inclined to call it that) comes from the new bond, which is technically a security. Calling this “money printing” is silly though. It’s like claiming that a corporation prints money when it issues stock.

  • Finn0123

    Hi Cullen,

    I believe I have a variation of the above, but I’m not sure if you’ve addressed it in other posts (if you have and I’ve missed it, my apologies).

    First, I agree with your statements on QE, bank reserves, and government spending. My variation, though, is what if a central bank bought Treasuries in the primary market (i.e., straight moneitzation of the debt)? My assumption is this would increase velocity and money supply (at least until the money entered the banking system) and thus prove inflationary? If this is true, could a central bank engage in infltion targeting (assuming it wanted to) by continuing the process indefinitely? Finally, is this Japan’s future?

    Sorry for the string of questions; just something I’ve been meaning to bring up for a while now.