Loans, Asset Purchases, and Exit Strategies—Why the WSJ Doesn’t Understand the Fed’s Operations

by Scott Fullwiler

Some may have noticed a few weeks ago when the European Central Bank – the counterpart to the Federal Reserve in the Eurozone – conducted a one-day operation that resulted in $622 billion in 1-year loans to the European banking system. At the time, I and others wondered where the fanfare was, as a similar operation by the Fed would surely have resulted in an outcry about the inflationary impact of such a large “liquidity” injection. But a piece by Simon Nixon in the Wall Street Journal explains why there was so little fanfare. As Nixon put it:

“True, the ECB isn’t increasing the quantity of money in circulation in the same way as the Federal Reserve and Bank of England, which are buying assets outright. But providing unlimited liquidity to banks is designed to achieve a similar goal. Since the crisis began, the ECB has expanded its balance sheet as aggressively as the Fed or the BOE.”

So, as Nixon correctly points out, the ECB has been expanding its balance sheet much like the Fed, but apparently this is less worrisome than the Fed’s actions, since the ECB is “only” lending, not purchasing assets as the Fed is doing. But let’s take a quick look at the Fed’s balance sheet. As of July 2, 2009, the Fed reported the following assets on its balance sheet:

  • Term Auction Credit: $283 billion
  • Commercial Paper Lending Facility: $115 billion
  • Central Bank Liquidity Swaps: $115 billion
  • “Other” Loans (includes the Primary Dealer Credit Facility): $119 billion

Together, these assets amount to $632 billion. And what do they all have in common? They are all LOANS. Note also that they comprise the bulk of the $726 billion in reserve balances held by banks the Fed reports on its liability side. So, it seems rather strange to give a blanket characterization of the Fed’s operations as asset purchases.

In fairness, it’s true that in the past few months the Fed has changed its strategy to purchasing Treasuries and ABS-related assets (the TALF), but overall the Fed’s balance sheet has shrunk, by over $600 billion (to around $2 trillion), while reserve balances have fallen from a peak of above $800 billion. Yet we still see the financial press and many economists continue to tell us that inflation is “just around the corner” since the Fed is “printing more and more money.”

More puzzling is that these lending programs were the overwhelming majority of the increase in the Fed’s balance sheet beginning last fall from around $800 billion in assets to a peak of over $2.6 trillion.

So, it again seems more than a little strange that when the Fed made a lot of loans back then, it generated concerns about inflation (witness Bernanke’s speeches earlier in the year trying to calm fears about the size of the Fed’s balance sheet), but when the ECB does it, there is no such outcry. John Taylor, for instance, was practically screaming bloody murder to anyone who would listen (see, here and here).

At their peak, the above Fed lending programs looked like this:

  • Term Auction Credit: $493 billion on March 12, 2009
  • Commercial Paper Lending Facility: $350 billion on January 22, 2009
  • Central Bank Liquidity Swaps: $600 billion-plus in December 2008 (precise figures not listed until 2009)
  • “Other” Loans: $438 billion on October 16, 2008

So, yet again, why do commentators consider it safe for the ECB to expand its balance sheet via lending, but they become alarmist when the Fed does it? Nixon explains that one of the main reasons why he and others (like Taylor) are so worried about the Fed’s purchases of assets is the Fed, unlike the ECB, lacks a clear “exit strategy.”

“The advantage of the ECB’s approach: It doesn’t require an elaborate exit strategy. As the three-month, six-month and one-year liquidities mature, the ECB’s balance sheet will automatically shrink. In contrast, the Fed and BOE are struggling to explain how they will mop up the excess liquidity they have provided, causing uncertainty and fueling fears of inflation.”

Apparently he and others don’t realize that the same is true for the Fed. John Taylor certainly didn’t cede that point when the Fed was extending all that credit. But all the above listed loan programs represent short-term lending, so the Fed’s balance sheet and (ceteris paribus) the quantity of reserve balances circulating will automatically shrink when the loans are paid back, just as with the ECB’s loans.

In fact, this has already happened. As one can see by simply comparing the peak numbers to the current ones for these lending programs, the Fed has already been “exiting” from these programs, while the financial press apparently hasn’t noticed.

But to address Nixon’s main point . . . are the Fed’s current asset purchases any more difficult to “exit from” than loans would be? To answer this, one must consider in a bit more depth the Fed’s operations. To start, obviously, the Fed’s holdings of Treasuries could be liquidated whenever desired. The TALF-related assets are admittedly a bit trickier, so I won’t address these directly. Instead, let’s look at the main issue being raised here, namely, the rise in reserve balances.

That is, it is the rise in reserve balances (from about $20 billion in August 2008 to over $800 billion within short order, and $622 billion now) that bothers the Fed’s critics the most, as I know of no economic theory that argues a rise in the central bank’s assets alone creates inflation.

But are reserve balances more difficult to drain if the central bank’s assets are mostly assets it has purchased instead of credit it has extended? Certainly not. The Fed already does this by selling Treasuries, engaging in reverse repos, and so forth. It could similarly issue its own bonds, as San Francisco Fed President Janet Yellen said, the Fed is considering (provided Congress approves), effectively turning overnight reserve balances into time deposits. Several other central banks already do this.

Of course, the underlying issue here is the incorrect belief that the rise in reserve balances will be inflationary as banks now have more “high-powered money” with which to create loans. As Nixon says, the ECB “hopes this funding will both encourage banks to redeploy the cash in other assets, driving down yields, and to resume lending.”

But this is incorrect, as I explained at length in a previous post.

To recap that post: Banks DO NOT use reserve balances to create loans. They create loans and deposits simultaneously out of thin air. They use reserve balances to settle payments or meet reserve requirements ONLY. If a bank is short reserve balances for either of these purposes, the Fed provides an overdraft AUTOMATICALLY at a stated penalty rate, which the bank then clears via the money markets or the cheapest alternative. Whether banks in the aggregate hold $1 or $1 trillion in reserve balances, there operational ability to create loans is the same . . . infinite! (Though the creation of even 1 loan requires a willing, credit-worthy borrow in the first place, of course.) Thus, the ECB is not providing reserve balances that banks can “lend or use” to purchase assets, but instead setting a cap on the cost of bank liabilities at different maturities when they do make loans or purchase assets (which, Nixon correctly points out, the Fed ought to do a lot more of).

That is how loan creation works in a modern monetary system. The belief that banks need reserve balances in order to lend is only applicable in a gold standard-type of monetary system.

What does this all mean? First, until recently, the operations of the ECB and the Fed weren’t all that different in the sense that both were using various standing facilities to provide credit, but for some reason only the Fed’s actions appear to have raised inflation fears. And second, talk of a need for an exit strategy is simply baseless. Neither the current rise in the Fed’s total assets nor the increase in reserve balances is inflationary. Banks don’t use reserve balances to create loans, so whether there are more or fewer of them is irrelevant. Indeed, an appropriate “exit strategy” might actually be for the Fed to continue to pay interest on reserve balances at its target rate—thereby enabling it to raise its target if it so desires at some point—and simply leave whatever reserve balances there are circulating, as I proposed a few years ago.

18 responses to “Loans, Asset Purchases, and Exit Strategies—Why the WSJ Doesn’t Understand the Fed’s Operations

  1. Can't banks with excess reserves purchase assets?Clearly the Fed wants some inflation, whether the 2% as they maintain to the public or a higher figure to more quickly relieve the debt burden on borrowers.So why is that $622B sitting in excess bank reserves if it isn't going to be lent out? That is the question your article does not answer.

  2. Isnt ECB's approach more inflationary considering that the eurozone currency is not technically "sovereign"?Thanks

  3. Dear Oregon GuyBanks don't need reserves to purchase assets since that's just using reserves to settle a payment (to pay for the asset) as I described in the post. If they ultimately need reserves to complete the transaction (often they don't due to net settlement) then if they are short they automatically receive an overdraft . . . which they then clear by the lowest cost method available. Again, it's just like settling any other payment.Also, as I said, when banks and others pay back the Fed's loans, that will drain reserve balances. But the Fed has I'm sure has its own reasons for not draining the reserve balances that I'm not going to try to decipher on my own . . . clearly some on the FOMCS erroneously think this helps bank create loans. The reasoning back in the fall was they didn't have enough assets to actually sell to drain the reserve balances being created, but that's not true now. The main point is banks can't "do anything" with these reserve balances that they couldn't already do without them.Best,ScottDear AnonymousI don't see how the ECB's approach is more inflationary since regardless of sovereignty banks again can't use the reserve balances for anything they couldn't do without them. The ECB is setting cost of liabilities at different maturities, that's it. Perhaps that's inflationary (some might think so) since it could help bank profitability (by raising spreads between assets and liabilities), but the reserve balances aren't the point in any case.Best,Scott

  4. Hi Scott – You have considered a simple bank model which does loan banking. However lets say JP Morgan enters into a swap contract with an I-Bank called IB which does not have an account at the Fed. Let us say JP Morgan loses billions in the swap – Wouldn't the amount of money increase ? IB can do things – buy stuff, pay huge salaries etc. Money is in circulation even without loans. Of course this example is hypothetical, but trading can create lot of deposits – dont you think ?

  5. Dear Oregon GuyI actually spoke too soon in the previous response, as it's quite obvious why the Fed hasn't drained the excess reserves. While they do have enough purchased assets to sell, the point of current operations is to be purchasing these assets to try and affect their prices . . . to turn around and sell them would be self-defeating, in their view (most likely). Besides these recently purchased assets, short of banks and others repaying the outstanding loans to the Fed, the Fed doesn't have enough other assets to sell to drain these excess reserves.As I noted in the post, they could do other things if they really wanted to drain them, like reverse repos of their purchased assets, ask Congress for the authority to issue debt at longer maturities than overnight (which is what reserve balances are), and so forth. But they aren't doing this.The broader point I'm making, though, is WHO CARES? (or, who should care? since many who don't understand Fed ops certainly do care about the qty of excess balances.) The balances aren't inflationary and the Fed can still adjust its target rate as desired since the target rate and the rate paid on reserve balances are the same. So, it doesn't matter if there are excess reserves or not.Best,Scott

  6. Dear Ramanathan (the same from Bill's blog? Welcome!)Two things. First, EVERY bank has an account at the Fed, whether they are members or not, since the Monetary Control Act in 1980. Might not work the same in other countries, but it does here. Second, we aren't talking about monetary aggregates here, we're talking about reserve balances. It doesn't appear that the example you are giving had anything to do with reserve balances, which is my point.As an aside, I think it best to eliminate the word "money" from usage when we want to talk about the monetary system in detail. In any form, "money" is a liability for someone, so it's better to just be specific about which liabilities one is talking about. Here, I am talking about a particular liability on the Fed's balance sheet (reserve balances) and the fact that the amount of this particular liability in circulation has nothing to do with the quantity of liabilities banks (or non-banks, for that matter) can or do create. Also, it's actually not a simple model of banking. It's a "general" model, if you will. You can add whatever on to it that you would like, but that core remains unchanged, unlike a simple model where simplifying assumptions have been made. I didn't need any simplifying assumptions . . . nothing has been assumed away, there's just some things I didn't talk about but they don't change the model at all when we do talk about them.Best,Scott

  7. Scott what about my question ? Cant bank reserves create trading losses for the banks and increase the deposits of the counterparty and increase deposits in general ?

  8. Dear Ramanan (apologies for spelling your name wrong in previous reply)As I noted, I don't see how reserve balances played any role in your example. First, if a bank doesn't have a reserve account, then there were no reserve balances changing hands in the first place. Second, ignoring that point, how do reserve balances create a trading loss? Don't confuse the trading loss with the means of settling (between those with reserve accounts) the trading loss . . . and remember that in Canada there are 0 reserve balances circulating overnight and trading losses are still settled. Hope I dealt with your question suffiently this time. If not, let me know what I'm not getting.Best,Scott

  9. Hi Scott,Thanks for the welcome. Yeah I am Ramanan from Billy Blog. Let us say a hedge fund H1 trades with hedge fund H2. When there is a payment which has to be made, the deposit in H1's account increases/decreases and that of H2's account decreases/increases – for the banking system nothing much has changed except that the two deposits may be in different banks. However if the hedge fund H1 trades with a bank B, say an option, it may happen that H1 needs to be paid $1m. Q: In that case, does a bank simply raise the deposit of H1 kept in the same bank B by $1m ? How precisely does it work ? Doesn't the increase in reserves make entering the trade (and other trades) more attractive for the bank ?

  10. Talking of comparing ECB and the Fed's actions – I think European banks have done silly stuff. In the US, the banks got reserves for treasuries and atleast get paid on the reserves. In Europe they get reserve balances for a year and in the end banks have to pay an interest to the ECB! Plus they have to give a collateral and there is a slight/minute risk in that – during one year they cannot sell that collateral.

  11. Dear Ramanan,Yes, if B pays for the option and H1 banks at B, B simply credits H1's deposit account. No reserve balances necessary.Again, banks have absolutely NO use for reserve balances except to settle payments and meet reserve requirements, and for those to purposes they are always provided by the central bank at stated rates, no demand.Best,Scott

  12. Yes Scott – learned it well from Billy Blog. My question was if there are indirect effect in the sense that having very high reserves changes the risk-appetite of the banks. But I guess it shouldn't matter. Thanks. – Ramanan

  13. Wow, did you ever get the wrong time slice!! The balance sheet was about 750 billion dollars two years ago. On 12/28 it is now 2.2 trillion. If they just close out those .6T in loans they've still doubled the balance sheet with new money.So which liablity entry constitutes that "new money"?

  14. Bank reserves are not irrelevant regarding the creation of new loans. You're arguments don't address the profit motive of banking. Banks are not charity organizations making loans only for the benefit of borrowers.It’s true they don’t need reserves to make a loan, but that doesn’t make it irrelevant. The pressure comes from the need for return on investment. Banks (used to) make no money on reserves. They only make money on loans (or to a lesser extent, fees). Second, reserves limit what they can lend. True, they can borrow fed funds to satisfy reserve requirements, but that increases the cost, and it is not infinite. One can only borrow reserves from a bank with excess reserves. Absent that, banks have to go to the discount window, which is frowned upon (except hopefully temporarily in the last crisis). I suspect somewhere, too, the Fed limits any bank from borrowing infinite fed funds or discount money. But I digress, and it isn't necessary to the claims.Profit comes from the spread on the cost of reserves vs. the profit on loans. Excess reserves provide profit opportunity at zero cost – the money is in hand already. Before The Fed offered to pay interest on reserves, they represented a ROI of zero. Unacceptable!! Banks pay to acquire and hold deposits; they can't let them sit there making no return. In fact, Ben’s policy to pay interest is the only proof one needs that this motivating factor is true (assuming Ben isn’t just playing Santa Clause and paying interest on reserves to be nice).

  15. Jade 1Please be more specific . . . I'm not sure what you are referring to.Best,Scott Fullwiler

  16. Jade 2 . . . my responses in CAPS below. Best, Scott FullwilerBank reserves are not irrelevant regarding the creation of new loans. You're arguments don't address the profit motive of banking. Banks are not charity organizations making loans only for the benefit of borrowers.AS I SAID IN THE POST, I ALREADY EXPLAINED ALL THIS IN MY POST IN JUNE AND LINKED TO IT. THERE, I DID SPECIFICALLY MENTION THE PROFIT MOTIVE.It’s true they don’t need reserves to make a loan, but that doesn’t make it irrelevant. The pressure comes from the need for return on investment. Banks (used to) make no money on reserves. They only make money on loans (or to a lesser extent, fees). OK. AGREE.Second, reserves limit what they can lend. True, they can borrow fed funds to satisfy reserve requirements, but that increases the cost, YES, AGAIN, AS I SAID IN THE POST I LINKED TO, THE NEED TO HOLD RESERVE BALANCES CAN AFFECT THE PROFITABILITY OF A LOAN.and it is not infinite. One can only borrow reserves from a bank with excess reserves. Absent that, banks have to go to the discount window, which is frowned upon (except hopefully temporarily in the last crisis).THE FED WILL SUPPLY THE ER IN THE CASE YOU ARE DESCRIBING, BECAUSE IT DOESN'T WANT THE FED FUNDS RATE TO BE BID ABOVE THE TARGET. THAT'S WHAT IT MEANS TO SET AN INTEREST RATE TARGET. ALSO, THE FED PROVIDES AN OVERDRAFT AUTOMATICALLY TO THE BANK, SO THE RBS ARE PROVIDED WITHOUT LIMIT (SEE REGULATION J). THE PROBLEM, AS YOU NOTE, IS THE FED FROWNS ON USING THE DISCOUNT WINDOW TO CLEAR THIS OVERDRAFT, OR AT LEAST USED TO. AGAIN, THOUGH, IF THE BANK CAN'T FIND OTHERS WILLING OT LEND IN THE FED FUNDS MKT AT THE TARGET, THE RATE IS BID ABOVE THE TARGET WHICH BRINGS AN OPEN MKT OPERATION TO PROVIDE THE ER AND BRING THE RATE BACK TO THE TARGET. THE FED'S "FROWNING" JUST DEMONSTRATES ITS LACK OF UNDERSTANDING OF ITS OWN OPERATIONS. I suspect somewhere, too, the Fed limits any bank from borrowing infinite fed funds or discount money. But I digress, and it isn't necessary to the claims.AGAIN, IF IT DOES, IT SHOWS IT DOESN'T UNDERSTAND ITS OWN OPERATIONS. AND IT PROVIDES THE ER ANYWAY WHEN THE FED FUNDS RATE IS BID UP ABOVE THE TARGET RATE.Profit comes from the spread on the cost of reserves vs. the profit on loans. Excess reserves provide profit opportunity at zero cost – the money is in hand already. Before The Fed offered to pay interest on reserves, they represented a ROI of zero. Unacceptable!! Banks pay to acquire and hold deposits; they can't let them sit there making no return. AGREE. AGAIN, I ALLUDED TO THIS IN THE POST I LINKED TO IN THE TEXT.In fact, Ben’s policy to pay interest is the only proof one needs that this motivating factor is true (assuming Ben isn’t just playing Santa Clause and paying interest on reserves to be nice).OF COURSE, BANKS IN THE AGGREGATE CAN'T RID THEMSELVES OF RBS. ONLY CHANGES TO THE FED'S BALANCE SHEET CAN DO THAT.

  17. Thanks for the stimulating discussion. About profit motive, I went to the link on June 11. Is that what you were referring to regarding your addressing of the issue of profit motive?I'll study that and several other links in detail, but to be clear (in case we misinterprete each other) I was referrring to the context of quarterly statements and returns of various types reported to stockholders. I'm seeing profitability of individual loans and credit-worthiness, not institutional strategic objectives.I'm hinting at the use of off-balance-sheet entities as we've see with Enron and with the latest R/E bubble collapse. Those cash flow entities have to do with strategic objectives more than profitablility of individual loans. Plus, they create "profitablitly" by accounting measures, not traditional old-fashiond loan underwriting practices. Be reducing risk with complex accounting, "profitablity" is a whole 'nuther ball game.My objective is to better understand how free reserves provide accounting tools to facilitate broader corporate strategic objectives as we've seen used thoroughly in the last two easy-money policy periods.

  18. Dear Jade"I'm seeing profitability of individual loans and credit-worthiness, not institutional strategic objectives."I'm intending the institutional strategic objectives there. I would agree that it's certainly not a loan-by-loan decision.Regarding "how free reserves . . . " the point I am making is that banks are not OPERATIONALLY constrained in creating loans/credit or purchasing assets by the qty of reserve balances they hold. The overwhelming majority of economists, policymakers, financial press, etc., believes banks are operationally constrained by reserve balances, and this is contributing to bad policy, so that's my main concern in these posts. There are certainly other considerations regarding reserve balances and one can quibble over how important they are or can become, but this is separate from the operational issues.Best,Scott Fullwiler