Is there anything facially stupid about the following argument that fed open market operations don't affect the money supply? And if not, are there any professional economists who have made this argument?

When the fed buys or sells treasury bonds they are not changing the broader stock of highly liquid assets that include treasury bonds (is that m3?), since they are just exchanging one of the components (cash) for another (govt bonds). Since there are no reserve requirements for commercial banks -- only capital requirements -- this doesn't actually affect the rate of credit creation since a bank that exchanges a dollar of central bank reserves for a dollar of us treasury debt is just as well capitalized as it was before.

  • $\begingroup$ google.com/url?sa=t&source=web&rct=j&url=https://… explains the connection between banks reserves and broad money. $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 17:24
  • 1
    $\begingroup$ Do you have a source for your question? In fact, "When the fed buys or sells treasury bonds they are not changing the broader stock of highly liquid assets." is certainly correct (and most likely refers to LCR I suppose). $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 17:56
  • $\begingroup$ @Alex no it was just something that occurred to me, cobbled together from my limited understanding of central and commercial banking $\endgroup$ Commented Aug 8, 2022 at 18:29

2 Answers 2


M3 isn't even published anymore in the US (it did not include bonds though).

As Alex noted, the LCR is entirely unaffected if the FED purchases government bonds because both are level 1. See for example this BIS paper on P.58 for a simple example demonstrating this.

Moreover, the Fed cannot control the federal funds

rate through routine changes in the quantity of reserves, also known as open market operations (OMO)

as copy pasted from this FED note about the ample reserves regime. This is a direct effect of the ample reserves regime, where the reserve requirement ratio is zero.

Generally, OMO can mean a lot of things (purchase of all sorts of securities) but (ignoring the title), I think the main argument is about the purchase of government bonds? For short, this does neither affect the interest rate (in an ample reserves regime), nor the liquidity ratio (LCR) / capital requirement of a bank. The latter statement requires that the central bank is purchasing a highly liquid (level 1) asset though (which US government bonds are).

Insofar, as written in your post, a purchase of government bonds by the FED has

no direct impact on credit creation,

at least not via the LCR or interest rate channel. I am inclined to believe that this was the argument made by whoever you heard this from. I am less convinced they mentioned money supply because traditionally the monetary base is defined as currency in circulation plus reserve balances and while there are more liquid types, none include bonds in the US in any case (there used to be M4 and L which included T-bills but these were stopped long ago).

Lastly, the implementation of monetary policy has evolved considerably and repeatedly since the financial crisis. The FED also does not (anymore) target monetary aggregates when conducting monetary policy. I believe reading this answer might be interesting, although many others have different opinions about this subject.


Commercial bank credit is NOT part of M1. When a bank provides a loan, the borrower receives a deposit. From the perspective of commercial banks balance sheets, this increases credits on the assets side and customer deposits on the liabilities side. Most of the times (with mortgages it is frequently directly sent to seller of the house / or mortgage notary's account) this deposit is withdrawn quickly and usually sent to another bank (unless the seller's bank account is at the same bank as the buyer's). Therefore, an individual commercial bank cannot generate lasting increases in its deposits by granting loans.

However, the banking system as a whole does see an increase in deposits (thus money supply) despite constant amounts of central bank money. Leaving potential reserve requirements aside, the bank will still have risk/return considerations to take into account when providing loans. From an asset liability management perspective, loans are long-term claims on the assets side, while sight deposits are typically liquid and short-term liabilities. In essence, that is one of the main reasons justifying commercial banking - it's an intermediary that brings together investors and savers despite their diverging requirements. Banks can offer this service because they can diversify the credit and liquidity risks better than individuals.

For risk management, banks need to consider factors such as current and future interest rate on loans and deposits, the likelihood of deposit withdrawals and credit defaults and the like. On top of this, banks are tightly regulated. Banks need to follow the standards of Basel III, which consist of 3 Pillars

Long story short, banks cannot simply give away loans without end. Also, credit creation by banks relies on a lot more details than the traditional reserve requirement (money multiplier) argument suggests. Specifically, it is not just the consideration of outflows (of deposits) and inflows being spread over time and normally only being a fraction of total deposits (which is where the name fractional reserve banking originated, as only a fraction of customer deposits have to be covered all the time). Since deposits created by the banking system belong to the banks’ customers, the main driving force behind credit creation are customers, not banks themselves.

In fact, the FED does not even aim to affect the broad money supply. St.Louis FED research claims it's best to forget what we learnt about money aggregates in undergrad Econ. The so called dual-mandate aims to (somewhat interestingly) promote three (not two) goals: maximum employment, stable prices, and moderate long-term interest rates

Neither of these goals mentions money supply. Also, New York FED - Money Supply states that

In 2000, when the Humphrey-Hawkins legislation requiring the Fed to set target ranges for money supply growth expired, the Fed announced that it was no longer setting such targets, because money supply growth does not provide a useful benchmark for the conduct of monetary policy.

and FED - what is money supply claims that

Over recent decades, however, the relationships between various measures of the money supply and variables such as GDP growth and inflation in the United States have been quite unstable. As a result, the importance of the money supply as a guide for the conduct of monetary policy in the United States has diminished over time

This answer has a few charts illustrating that total money supply is not really directly impacted by monetary policy. For example, normalizing each series to Dec 2005 = 100 (monetary base - BOGMBASE, M1 - be careful, the definition changed in May 2020 if you look at the data on FRED, M2), the data looks like this (the code to replicate this is in the link: QE stands for quantitative easing)

enter image description here

The dual-mandate got its name from employment and the price level, because under full employment (not no unemployment), and stable prices, interest rates settle at moderate levels. For details, you can read Frederic S. Mishkin (2007) speech at Bridgewater College.

To summarize, the main goals of the FED are to keep prices stable and to promote full employment (closely related to a healthy economy / GDP). However, as just seen, there is only a weak relationship between measures of the money supply and GDP growth and inflation. Therefore, the central bank focuses on interest rates, not only the FED Funds market, but also longer term interest rates. Changing interest rates has a direct impact on demand for credit (by customers) and the risk metrics of banks (current and future interest rate on loans and deposits, the likelihood of deposit withdrawals and credit defaults, Basel III calculations, ...).

As written above, monetary policy has evolved considerably. For example, in the EUR area, so called Targeted longer-term refinancing operations (TLTROs) aim to offer commercial banks attractive long-term funding conditions. To stimulate lending to the real economy, the interest rate is negative and in a nutshell, the more loans participating banks issue to non-financial corporations and households (except loans to households for house purchases), the more attractive (the more negative) the interest rate. While ultimately there needs to be demand from customers, these favourable funding conditions make it easier (cheaper) to offer loans.

  • $\begingroup$ Thanks lots of great resources here -- the stuff on the ample reserve regime is particularly enlightening and puts a name to something i wondered about for a while (namely, how can changes in total bank reserves really matter when banks are already sitting on massive excess reserves?). I'm a little confused where you come down though. It seems like you agree that there's "no direct impact on credit creation" but not that there's no effect on the money supply. When I said "money supply" in my post i wasn't referring to the monetary base (continued...) $\endgroup$ Commented Aug 8, 2022 at 19:53
  • $\begingroup$ The monetary base obviously increases with the creation of new reserves when the fed buys assets like govt bonds from commercial banks. By "money supply" i meant a measure like M1, i.e. something like cash and coins plus commercial bank credit. So if the fed's bond purchases don't expand commercial bank credit I don't see how they can affect the money supply. $\endgroup$ Commented Aug 8, 2022 at 20:01
  • 1
    $\begingroup$ Thanks so much again for the added info. The only remark I have is when you say "Commercial bank credit is NOT part of M1" I think you and I must be using the word "credit" in different ways. All I mean by "credit" is electronic deposits at commercial banks -- ie what's in my checking account right now. Surely that is part of M1, right? Probably I am just misusing the word "credit". Although i'm not sure what "commercial bank credit" would refer to if not these deposits. $\endgroup$ Commented Aug 8, 2022 at 23:33
  • 1
    $\begingroup$ Commercial bank credit is an asset of banks, whereas deposits are liabilities. Bank credit is the first entry on the asset side, and deposits are the first entry on the liability side of the Assets and Liabilities of Commercial Banks in the United States - H.8. Frequently credit and loans are used interchangeable to refer to financing, but strictly speaking these are distinct types of financing. $\endgroup$
    – AKdemy
    Commented Aug 9, 2022 at 0:04
  • $\begingroup$ A checking account derives its name from the standard method of making a withdrawal, which was by writing a cheque. While this method is a bit archaic, the name checking account still exists and is essentially the same as a current account or deposit account (a type of account that accepts deposits, where you can withdraw in a number of ways - cheque, debit card, ATM machine etc.). Either way, they are liabilities for banks (as long as there is a positive sum on the account). $\endgroup$
    – AKdemy
    Commented Aug 9, 2022 at 0:14

OMOs do increase money supply. When central bank conducts OMOs it does not use pre-existing money in the economy, it creates new money to conduct OMOs. OMOs would also increase money supply even without existence of private bank.

When the fed buys or sells treasury bonds they are not changing the broader stock of highly liquid assets that include treasury bonds (is that m3?)

US bonds are generally not M3 because average maturity of US bonds is 65 months (see Peterson Foundation) and M3 includes only debt securities with maturity less than 2 years (see OECD). In any case even if you would count all bonds in your preferred measure of money supply the bond would still exist and new cash that previously not existed was created so the measure of money supply would still increase. When central bank buys bonds it does not automatically destroy them the way it automatically destroys cash when it sells bonds.

Since there are no reserve requirements for commercial banks -- only capital requirements -- this doesn't actually affect the rate of credit creation since a bank that exchanges a dollar of central bank reserves for a dollar of us treasury debt is just as well capitalized as it was before.

No this is actually not true. The new capital requirements depend on type of debt banks are holding. Banks generally need less capital for government bonds then private loans under Basel III.

Moreover, the term 'capital requirements' is somewhat misleading. Capital requirements under Basel III also specify that banks have to have some level of liquid assets (see Basel III overview). To be more specific banks have to maintain certain levels of Liquidity Coverage Ratio (LCR) and reserves are highly liquid.

Hence even though de jure capital requirements are abolished, de facto banks still need those reserves to function under Basel III. However, now there is no simple rate of required reserves for every single loan banks make, rather each bank needs different fraction of reserves depending on the riskiness and liquidity of the loans they issue.

  • $\begingroup$ Deposit accounts are frequently backed by US bonds, aren't they? So how could US bonds not be money? $\endgroup$ Commented Aug 8, 2022 at 16:30
  • $\begingroup$ @user253751, debt is never money. Credit cards are not money either. $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 17:03
  • 1
    $\begingroup$ To compute LCR you need to figure out HQLA,which consists also of Level 1 assets, which in turn include Federal Reserve bank balances, foreign resources that can be withdrawn quickly, securities issued or guaranteed by specific sovereign entities, and U.S. government-issued or guaranteed securities. Insofar, it does not change the LCR as much as I know. $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 17:04
  • 1
    $\begingroup$ Bonds are simply not part of US monetary aggregates. You usually cannot pay anything with a bond. You can sell the bond to receive money which you can in turn use to pay for things. The repo market is gigantic and wouldn't exist if you could outright use bonds like cash (money). $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 17:49
  • 1
    $\begingroup$ @1muflon1, but government bonds do matter for LCR because these are part of the same level 1 hqla as reserves. $\endgroup$
    – Alex
    Commented Aug 8, 2022 at 20:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.