Bill Dudley on monetary policy lessons

Bill Dudley has an excellent column in Bloomberg:

More importantly, the Fed has yet to adequately recognize two important ways monetary policy contributed to the crisis.

First, by committing to keep short-term interest rates near zero until the economy reached full employment and inflation exceeded 2%, the Fed ensured that it would be very late in tightening monetary policy. When inflation overshot, it had to raise rates faster and higher than it would if it had been more preemptive. As a result, it delivered a significantly larger shock to banks’ funding costs and to the value of the banks’ longer-term investments.

Second, the Fed’s quantitative easing program, by exchanging cash for securities, flooded the banking system with reserves and deposits. In the low-rate environment of 2020 and 2021, this naturally tempted some banks to boost earnings by buying higher-yielding, long-term, fixed-income assets.

But I disagree with Dudley’s conclusion:

I see three lessons for the Fed. First, the Federal Open Market Committee should consider financial stability risks in its monetary policy decisions. 

Dudley correctly observes that the same Fed policy that led to a bad outcome for banking also led to a bad macroeconomic outcome. The alternative policy that he recommended would have led to a better macro outcome. Focusing like a laser on stable NGDP growth won’t eliminate banking problems, but at least it will avoid having monetary policy exacerbate those problems.

PS. Speaking of Bloomberg, it really annoys me when reporters misstate the Fed’s inflation target, which applies to the PCE, not the CPI:

US consumer prices rose 0.4% in April with headline CPI up 4.9% on a year-on-year basis, its first reading below 5% in two years. That’s still well above the 2% level targeted by the Fed as central bank officials juggle the need to curb rampant inflation against a potential recession and banking sector angst.



36 Responses to “Bill Dudley on monetary policy lessons”

  1. Gravatar of Michael Sandifer Michael Sandifer
    11. May 2023 at 01:46

    I’d like to see a responsible central bank somewhere experiment with a monetary policy that doesn’t rely on asset purchases or sales. An NGDP level targeting policy could be conducted with a computer, with NGDP futures information as inputs, or using my stock market method. The former approach would be more precise, at least on paper.

    Instead of relying on the sale of government bonds, for example, to tighten policy, a computer could simply schedule to gently create less money in the future, spread out over a long horizon.

    I wonder more generally how long it will be until we start to hear serious talk about automating monetary policy with machine learning.

  2. Gravatar of Nick S Nick S
    11. May 2023 at 03:20

    RE: “ Instead of relying on the sale of government bonds, for example, to tighten policy, a computer could simply schedule to gently create less money in the future, spread out over a long horizon.”

    The Fed doesn’t create money… it creates bank reserves. Commercial banks create money via loan originations.

    The readership of this blog cannot get this concept through their heads.

    Please listen and read Jeff Snider, who is probably one of the few who actually understands money.

  3. Gravatar of Spencer Spencer
    11. May 2023 at 04:52

    Snider doesn’t know what he’s talking about. He has a grasp on the E-$ market, but that’s it. Snider gets the domestic market very wrong.

    “During this time, the Federal Reserve under Paul Volcker continued to pursue a policy of restricting bank reserves.”

    Monetarism involves controlling total reserves, not non-borrowed reserves as Paul Volcker found out. Volcker targeted non-borrowed reserves (@$18.174b 4/1/1980) when total reserves were (@$44.88b).

    And: “The problem, one of many, was in tying bank reserves to the effective money supply (which bank reserves aren’t) then to overall bank credit and finally to nominal economic outcomes.”

    After the DIDMCA, total reserves increased at a 17% annual rate.

  4. Gravatar of Spencer Spencer
    11. May 2023 at 05:06

    As I said: The only tool in a free capitalistic system through which the volume of money can be properly controlled is legal reserves. The FED will obviously, sometime in the future, lose control of the money stock.
    May 8, 2020. 10:38 AMLink

    link Daniel L. Thornton, Vice President and Economic Adviser: Research Division, Federal Reserve Bank of St. Louis, Working Paper Series
    “Monetary Policy: Why Money Matters and Interest Rates Don’t”

    Link: Daniel L. Thornton, May 12, 2022:
    “However, on March 26, 2020, the Board of Governors reduced the reserve requirement on checkable deposits to zero. This action ended the Fed’s ability to control M1.”

    Whereas Dr. Richard Anderson said: “reserves are driven by payments”

    Quantitative Easing and Money Growth:
    Potential for Higher Inflation?
    Daniel L. Thornton

    “the close relationship between the growth rates of required reserves & total checkable deposits reflects the fact that reserves requirements apply only to checkable deposits”

  5. Gravatar of Spencer Spencer
    11. May 2023 at 05:27

    Re: “Focusing like a laser on stable NGDP growth”

    N-gDp has been too high for 9 quarters:

    2021-01-01 ,,,,, 11.7
    2021-04-01 ,,,,, 13.8
    2021-07-01 ,,,,, 9.0
    2021-10-01 ,,,,, 14.3
    2022-01-01 ,,,,, 6.6
    2022-04-01 ,,,,, 8.5
    2022-07-01 ,,,,, 7.7
    2022-10-01 ,,,,, 6.6
    2023-01-01 ,,,,, 5.1

    Whereas an expansion of legal reserves takes place over a period of time, the contraction of legal reserves takes place immediately:

    Some people think Feb 27, 2007 started across the ocean:

    “On Feb. 28, Bernanke told the House Budget Committee he could see no single factor that caused the market’s pullback a day earlier”.

    In fact, it was home grown. Feb 27coincided with the sharpest decline in:

    (1) the absolute level of “free” legal reserves, &
    (2) & an historically large peak-to-trough reversal of roc’s for proxies on real GDP & the deflator.
    Flow5 January 27, 2009 2:02 PM

    Another example:

    “Black Monday” Oct. 19, 1987, coincided with the sharpest and fastest peak-to-trough decline in the roc in required reserves since 1915.
    Flow5 January 27, 2009 2:17 PM

    Powell didn’t have to wait until inflation took off. He should have known that when N-gDp is too high, an increase in inflation follows.

    The effect of the FED’s operations on interest rates (now largely via the remuneration rate), is indirect, varies widely over time, and in magnitude. What the net expansion of money will be, as a consequence of a given injection of additional reserves, nobody knows until long after the fact.

    The consequence is a delayed, remote, and approximate control over the lending and money-creating capacity of the payment’s system.

    The consequence is FOMC schizophrenia: Do I stop because inflation is increasing? Or do I go because R-gDp is falling?.

  6. Gravatar of ssumner ssumner
    11. May 2023 at 05:29

    Nick, The Fed creates base money, which is one definition of money. There are others, such as M1 and M2.

    No one will take you seriously if you continue posting this sort of nonsense.

  7. Gravatar of Spencer Spencer
    11. May 2023 at 05:46

    Fed Paying Interest on Reserves: an Old Idea With a New Urgency

    “But since required reserves earn no interest, they are a tax, and like all taxes, create distortions”

    Reserves are “Manna From Heaven”, costless and showered on the system.

    Given bankable opportunities (and the Federal Government is the largest creditworthy borrower), on the basis of these newly acquired free reserves, the commercial banks created a multiple volume of credit and money. And, through this money, they acquired a concomitant volume of additional earnings assets.

    How much was this multiple expansion of money, credit, and bank earning assets? Thanks to fractional reserve banking (an essential characteristic of commercial banking) for every dollar of legal reserves pumped into the member banks by the Fed, the banking system acquired about 93 (c. 2006), dollars in earning assets through credit creation.

    See “Bank Reserves and Loans: The Fed Is Pushing On A String” – Charles Hugh Smith

  8. Gravatar of Classical Liberal Classical Liberal
    11. May 2023 at 11:17

    “it really annoys me when reporters misstate the Fed’s inflation target, which applies to the PCE, not the CPI”

    Hear! hear! How can anyone not know the difference between chain and fixed-weighted price indices!!?? I mean, honestly!

  9. Gravatar of Will Will
    11. May 2023 at 13:13

    “it really annoys me when reporters misstate the Fed’s inflation target, which applies to the PCE, not the CPI”


  10. Gravatar of Solon of the East Solon of the East
    11. May 2023 at 15:31

    Maybe the Fed should say its target is 2% to 3% on the CPI.

  11. Gravatar of Spencer Spencer
    12. May 2023 at 06:35

    Bank failures: it’s all about liquidity (

    #1 “the NY DFS’s chart above shows that the deposits that fled were predominantly interest-bearing”
    #2 “a large concentration of uninsured deposits, without sufficient funds management contingency plans,”

    Large Time Deposits, All Commercial Banks (LTDACBM027NBOG) | FRED | St. Louis Fed (

    Dr. Lester V. Chandler (“After earning his doctorate in economics at Yale University, Chandler taught at Dartmouth and Amherst before accepting a position as Gordon S. Rentschler Memorial Professor of Economics at Princeton University”):

    1961 “Professor Pritchard is quite right, of course, in pointing out that commercial banks tend to compete with themselves when the issue savings deposits”

    Commercial banks pay for the deposits that the system already owns. I.e., all time deposits, interest-bearing deposits, have been shifted from demand deposits.

    Savers never transfer their savings outside the payment’s system unless they hoard currency or convert to other National currencies. So, savings flowing through the nonbanks never leaves the commercial banking system. It’s stock vs. flow. The only way to “activate” monetary savings is for their owners, saver-holders, to spend or invest outside of the banks.

    Disintermediation for the DFIs can only exist in a situation in which there is both a massive loss of faith in the credit of the banks and an inability on the part of the Federal Reserve to prevent bank credit contraction, as a consequence of its depositor’s withdrawals.

    There was no disintermediation for the commercial banks until Bernanke “did it again”.

    The removal of Reg. Q ceilings for the banks (the nonbanks were unregulated prior to 1966), was a monetary policy blunder.

  12. Gravatar of Spencer Spencer
    12. May 2023 at 07:52

    “The 3-month annualized rate of total CPI less shelter has been running steadily at 1.2%.”

  13. Gravatar of Spencer Spencer
    12. May 2023 at 08:20

    Readers should note that calculating inflation on a year-to-year basis minimizes, over time, the rate of inflation since the rate is being calculated from higher and higher price levels. A $ today, using 1967 (a former base year), is equivalent to $9.22 of consumer purchasing power today.

    In absolute terms, each year confronts all of us with a higher and higher level of prices with no end in sight.

  14. Gravatar of Nick S Nick S
    12. May 2023 at 09:32

    Scott – Sure, the creation of reserves is technically be included in the textbook definition of some money measures, but my point is that reserve creation/destruction by the Fed does NOT result in credit expansion/contraction. Bank lending does…. The Fed can use their tools to inject or pull reserves from banks all they want, but if banks don’t lend, then this policy has zero impact. You can’t go to the store and buy something with bank reserves. Those reserves need to get transformed into loans/deposits to see any impact to the real economy.

    Spencer – Snider is saying that it is a mistake to link reserves to money supply, bc of what I’ve stated above. The Fed cannot create deposits and loans (I.e. money actually used in the real economy). Only the banking sector can do that.

    Your point about borrowed vs non borrowed reserves is not relevant now, as borrowed reserves are such a small fraction, and only are housed by troubled banks accessing the window, and those banks are certainly not in any position to extend credit.

  15. Gravatar of Spencer Spencer
    12. May 2023 at 10:55

    Nick: Snider was talking about the 1980s. Today, your view is also not right.

    Open market operations should be divided into 2 separate classes:

    (#1) purchases from, and sales to: member commercial banks;
    (#2) purchases from, and sales to: “other non-bank entities”:

    (#1) OMO transactions of the buying type between the FRB-NY’s “trading desk” (the Central bank) and the member commercial banks directly affect the interbank demand deposit volumes in one of the 12 District Reserve banks without bringing about any change in the money stock.

    The “trading desk” credits the master account of the clearing bank used by the primary dealer from whom the security is purchased. This alteration in the assets of the commercial banks (the banks’ IBDDs), increases – by exactly the amount the PD’s portfolios (or acting as dealer agents, NB’s portfolios), of Treasury and coupon securities was decreased.

    (#2) Purchases and sales between the Reserve banks and non-bank investors directly affect both bank reserves (inside money) and the money stock (outside money).

  16. Gravatar of Spencer Spencer
    12. May 2023 at 10:58

    (#2) Purchases and sales between the Reserve banks and non-bank investors directly affect both bank reserves (outside money) and the money stock (inside money).

  17. Gravatar of Solon of the East Solon of the East
    12. May 2023 at 16:36

    “we note that the Atlanta Fed’s wage tracker staged the biggest drop in April for any month since the series began in 1997. If that’s not an outlier, the balance of risks could tilt towards a rate cut.”—-Oxford Economics

    Well, sure, only one data point.

    But…seeing a lot of signs inflation is headed down. Already dead in Japan and China. Most of SE Asia.

    The Federal Reserve is now paying 5.15% on IOER. Why should commercial bank lend when they can collect risk-free work-free 5.15%?

    Technical question: Can banks put excess reserves into Treasuries, and then collect interest on Treasuries and also the 5.15% interest from the Fed?

  18. Gravatar of ssumner ssumner
    12. May 2023 at 21:32

    Nick, I’ve spent my whole life studying money. You come on here and tell me I don’t know what money is and expect to be taken seriously? You need some self awareness.

  19. Gravatar of Spencer Spencer
    13. May 2023 at 04:17

    Remunerating IBDDs induced nonbank disintermediation, an outflow of funds, or negative cash flow. It inverted the short-end segment of the retail and wholesale money market funding yield curve.

    During the GFC the nonbanks shrank by 6.2 trillion dollars while the banks expanded by 3.6 trillion dollars.

    The Covid-19 inversion is even more pronounced.

  20. Gravatar of Spencer Spencer
    13. May 2023 at 05:04


    When the reverse-repo facility was set up, Bill Dudley, then the president of the New York Fed, worried it could lead to the “disintermediation of the financial system”.

    The FED doesn’t know a debit from a credit.

  21. Gravatar of Spencer Spencer
    13. May 2023 at 05:48

    “The I.M.F. has called for tougher regulatory oversight, and U.S. Treasury Secretary Janet Yellen said last month that she wanted to make it easier to designate nonbanks as systemically important,”

    The Austrian Business Cycle theory is bunk. Recessions are generated by nonbank disintermediation.

  22. Gravatar of seer of things seer of things
    13. May 2023 at 20:35

    Interesting quotes from Axios…

    Mary Daly: “Globalization has been a key driver of past goods deflation in the United States”.

    A direct contradiction from the same article is interesting to see.

    Jason Furman: “Globalization, technology, demography had little to do with why inflation was low before the pandemic and have little to do with why it is high now. Ultimately, central banks and monetary policy can pick the inflation rate they want.”

    John Williams: “It is our job to make sure inflation is low and stable. We can do that even in a world with deglobalization or friendshoring”.

  23. Gravatar of Michael Sandifer Michael Sandifer
    14. May 2023 at 06:52

    Nick S,

    You’re confusing money and credit. The price of money is the things it can buy. The price of credit is interest. It’s a common mistake that’s even made by many economists.

    If you’d read Scott’s blog carefully for long enough, and/or other market monetarist blogs, you’d understand that, and even if you disgree, you wouldn’t come here trying to tell him he doesn’t know what money is.

  24. Gravatar of Spencer Spencer
    14. May 2023 at 09:32

    Tried to post this on your other blog. But it doesn’t work.

    It is the means-of-payment money supply measured by the distributed lag effect of money flows. The rate-of-change in M*Vt had already hit historic levels by Sept. 2020.

    Any expansion after that point represented a monetary policy that was too easy.

    2/1/2020 ,,,,, 1558.7 ,,,,, 0.087
    3/1/2020 ,,,,, 1824.1 ,,,,, 0.205
    4/1/2020 ,,,,, 2047.6 ,,,,, 0.360
    5/1/2020 ,,,,, 2128.9 ,,,,, 0.452
    6/1/2020 ,,,,, 2220.4 ,,,,, 0.519
    7/1/2020 ,,,,, 2265.1 ,,,,, 0.530
    8/1/2020 ,,,,, 2279.1 ,,,,, 0.544
    9/1/2020 ,,,,, 2392.4 ,,,,, 0.643
    10/1/2020 ,,,,, 2432.5 ,,,,, 0.641
    11/1/2020 ,,,,, 2757.2 ,,,,, 0.912
    12/1/2020 ,,,,, 3359.4 ,,,,, 1.184
    1/1/2021 ,,,,, 3371.9 ,,,,, 1.271
    2/1/2021 ,,,,, 3533.8 ,,,,, 1.427
    3/1/2021 ,,,,, 3770.5 ,,,,, 1.572
    4/1/2021 ,,,,, 3846.3 ,,,,, 1.562
    5/1/2021 ,,,,, 4016.6 ,,,,, 1.712
    6/1/2021 ,,,,, 4256.0 ,,,,, 1.796
    7/1/2021 ,,,,, 4375.4 ,,,,, 1.836
    8/1/2021 ,,,,, 4437.2 ,,,,, 1.914

    The roc had never been over 1.000 before.

    But even this measurement understates the transaction’s velocity of funds.

  25. Gravatar of Solon of the East Solon of the East
    15. May 2023 at 00:32

    India is solidly into disinflation also. Just reported an 18-month low.

    China in outright deflation?

  26. Gravatar of Randomize Randomize
    15. May 2023 at 12:23

    “Forever” seems like a strong word but it does feel like we could be facing Japan-levels of workforce scarcity for a while. Would love to hear your thoughts on this, Dr. Sumner.

  27. Gravatar of Spencer Spencer
    16. May 2023 at 05:35

    “The rate of remuneration on bank reserves at the Fed (5.15%) is now (May 2023) substantially higher than the yield on 10-year US government securities (3.4%). ”

    This artificial interest rate inversion is restrictive. And because of “core” deposits, the banks are able to outbid the nonbanks for loan funds, inducing nonbank disintermediation.

    Lowest Comfortable Level of Reserves (“LCLoR”)

  28. Gravatar of Solon of the East Solon of the East
    17. May 2023 at 01:56


    Today’s CPI Data by Truflation

    The USA Inflation Rate by Truflation is 3.36%, -0.17% change over the last day. Read Methodology

    Not everything threatens a 1970s replay. (When, btw the PCE core annual never got into double digits)

  29. Gravatar of Michael Sandifer Michael Sandifer
    17. May 2023 at 10:21


    I think many economists, including the Fed, were far too impatient with the recovery of the supply-side of the economy after the trade wars, the pandemic, and now the continued implementation of industrial policy. Yes, there are permanent efficiency losses associated with the trade restrictions and industrial policy, but not all of the losses are permanant, as the economy will adjust.

    The Fed really gave up on its “transitory” language when the criticism started to fly.

    Again, that’s not to say that monetary policy was not too loose for a while, but that most of the inflation problem has been due to temporary, non-monetary factors.

  30. Gravatar of ssumner ssumner
    17. May 2023 at 17:13

    Randomize, It depends on the path of NGDP.

    Michael, Nope, the NGDP data suggests the inflation has almost all been demand side. I’ve shown this in previous posts.

  31. Gravatar of Michael Sandifer Michael Sandifer
    17. May 2023 at 20:01


    Yes, I’m aware that is your view, and as I recall, you’ve offered very defensible arguments. I suspect you consider them far more solid than “defensible”.

    I think more recent data is better supporting my view. I became concerned when inflation breakevens began to approach 3% in core PCE terms, acknowledging at the time that money had perhaps gotten too loose. That situation didn’t last long.

    Now, with inflation breakevens falling below 2% in core PCE terms, but current inflation remaing high, though falling, it increasingly appears that my greater emphasis on temporary non-monetary factors was correct. The evidence also seems to increasingly suggest that real growth potential is significantly higher than most expected.

    And I also think we fundamentally disagree about the short-run impact of the new large language models on productivity growth. I don’t think this should be compared to electricity or PC adoption, as those technologies required signficant capital investment, as well as time to perfect the technology to allow for growth in applicability.

    Just this Monday, I used prompt engineering via ChatGPT-4 to design a retail chatbot for an online company with a few hundred million dollars in retail sales and it works fabulously. I engineered it for a more narrowly focused use case, using automated prompt engineering. It works far better than the chatbot the company currently has on its website, and offers superior search capabilities as well. It has real conversations with customers, having the ability to make product recommendations given a customer’s stated needs, convey in-stock status, explain shipping policies and options, etc.

    This took me 2 hours, with a few hours spent polishing it thereafter. I’ve been coding for 2 months. This would have taken years of experience coding, prior to ChatGPT.

    I will probably deploy this chatbot in various ways on my own website, which I’m buidling from scratch, with ChatGPT’s help.

    I get all this for $20/month.

    18. May 2023 at 18:29

    Committing to specific setting of policy instruments is always bad. I was not aware that the Fed ever DID “to keep short-term interest rates near zero until the economy reached full employment and inflation exceeded 2%” But even if it did, that would not have prevented it from staring to raise ST rates much sooner than it did, at least as soon as TIPS started exceeding its target.

  33. Gravatar of spencer spencer
    19. May 2023 at 07:09

    Brian Sack:
    “the floor system…is designed to achieve effective control of overnight market interest rates under a variety of outcomes for the balance sheet size. That control has been very, very strong. So, the SOFR rate, which is the benchmark repo rate, probably the most important overnight interest rate in our financial system, that has been largely pinned to the rate set on the Feds’ overnight reverse repo facility or at least within a few basis points of that facility rate. And the federal funds rate has remained remarkably stable in the center of the target range set by the FOMC.”

    The time-frame of the FED’s economic analysis is 24-hours rather than 24-months.

  34. Gravatar of spencer spencer
    19. May 2023 at 07:33

    “there are calls to reduce the RRP rate relative to the interest rate on reserves to try to push assets out of money funds and back into the banking sector, but there’s a couple problems with that – the biggest problem being, if you start reducing the rate on the RRP facility, you are going to lower market interest rates and you are going to compromise your policy setting,”

    Guaranteed nonbank disintermediation, a deceleration in velocity, a reduction in R-gDp.

  35. Gravatar of spencer spencer
    19. May 2023 at 07:54

    “So, basically the proposal is, Treasury could have some predictable regular policy for buying age securities and replacing them with newly issued securities, which tend to be more liquid.”

    Good way to start a flight from the dollar, hyperinflation, a run-in short-term liabilities and destruction of the U.S. dollar.

  36. Gravatar of mira mira
    19. May 2023 at 10:04

    absolutely agreed scott.

    the problem with many of these suggestions around ‘considering financial stability’ is that they have a Newton’s 3rd law sort of effect: if the Fed considers how the banks might be impacted, the banks will necessarily consider how the Fed will save them.

Leave a Reply