The actual “ammunition” issue

Pundits often make the mistake of assuming that expansionary monetary policy somehow uses up “ammunition” by reducing interest rates, which gives central banks less room to cut rates in a future recession. Actually, the reverse is true. Expansionary monetary policy boosts the equilibrium (or natural) nominal interest rate, and hence gives central banks more “ammunition” in the conventional sense of the term. This is precisely why some economists advocate a higher trend rate of inflation—to give the Fed more ammunition.

There is another sense in which conserving ammunition does make sense. For any given size of the Fed’s balance sheet, monetary policy is more expansionary when bank reserve demand is lower. Unfortunately, since 2008, the Fed has made a number of changes that boost reserve demand and hence reduce “ammunition”. (I use scare quotes because in a technical sense central banks have near infinite ammunition; the problems are legal and psychological.)

One recent change is obvious; the payment of IOR has increased demand for bank reserves. Some would argue that this isn’t really a problem, as the rate of interest paid on reserves can be reduced to zero in a crisis. Maybe, but it wasn’t reduced to zero during the worst banking crisis of the past 80 years.

A bigger problem is changes in regulation that artificially boost the demand for bank reserves, discussed in an excellent post by Greg Baer and Bill Nelson:

In September 2019, the repo market broke down because banks and broker-dealers did not step in to equilibrate a supply-demand imbalance caused by corporate tax payments and Treasury securities settlement. Banks were unwilling to use their reserve balances to lend into the repo market, in part because of a supervisor preference that banks hold reserves rather than Treasuries.  Banks also were unwilling to raise funds to lend into the repo market for two reasons: leverage requirements and GSIB surcharges generally made the resulting balance sheet expansion too expensive, and banks were reluctant to upend their capital and liquidity allocations to respond to a temporary event.

The Fed responded to the breakdown by further expanding its balance sheet and actively considering creating a standing repo facility that would mirror its standing reverse repo facility.  Reportedly, the standing facility might not just lend funds to commercial banks (which already borrow through the discount window) and primary dealers (which already participate in Fed open market operations) but also hedge funds and other participants in the FICC sponsored repo program.  In short, the Fed stepped in to solve a problem caused by its increased role in financial markets by further expanding its role in those markets and by planning to increase its role in them still further.

Their analysis leads to a policy recommendation that is music to my ears:

We encourage the Fed to . . . conduct policy in a manner similar to how it did before the crisis, with a level of excess reserves about one thousandth of its current level.

I believe that policy is more likely to avoid unexpected problems such as we saw in September 2019 if it is kept simple. Prior mid-2008, the monetary base was roughly 98% currency. That’s a very simple monetary regime. It’s true that marginal changes in the base were first implemented via adjustments in the other 2% (bank deposits at the Fed), but base injections tended to eventually impact the currency stock, as required to keep inflation close to 2%.

I acknowledge that there are potential efficiency gains associated with superabundant reserves, but I worry that the demand for bank reserves in our current system is so high and so unstable that it might make monetary policy (especially QE) less effective in a future crisis. It takes a heap of Harberger triangles to fill an Okun’s gap. K.I.S.S.

Baer and Nelson also discuss other downsides of current Fed balance sheet policy, such as the risks associated with the Fed’s much greater involvement in our financial system.

PS. This issue relates to the current debate over the floor vs. corridor system. I’m obviously with supporters of the corridor system. George Selgin has a number of illuminating posts on this issue.

PPS. Stock prices and bond yields fell today in response to one additional coronavirus case in the US. I believe the greatest risks to NGDP are to the downside, which is why the Fed should adopt a slightly more expansionary policy. We need a policy where the risks are balanced. We are close, but not quite there yet.



18 Responses to “The actual “ammunition” issue”

  1. Gravatar of Benjamin Cole Benjamin Cole
    24. January 2020 at 20:22

    I confess to doing the intellectual pretzel when standing repo facilities vs. standing reverse repo facilities are bandied about. Huh? Does this stuff matter? I can’t tell.

    I am further twisted when people call for smaller or larger central-bank balance sheets.

    The Swiss National Bank has a balance sheet of about $100,000 per resident. They have modest growth and scant inflation. The US Fed has a much smaller (relatively) central-bank balance sheet, and similar results (but a larger annual federal deficit), and then there is Japan…I am looking for some consistent results, but where?

    This is snipe-hunting for the right monetary policy!

    Adding to the fog, Scott Sumner has often advocated “whatever it takes” in central-bank QE to beat back a recession. I think I agree (though I might prefer money-financed fiscal programs).

    But if Fed does go to the mattresses on QE to beat recession, then…it will have a larger balance sheet.

    Sumner will say the fact that the Fed is willing to do “whatever it takes” means it won’t have to…

    But to convince markets to keep the Swiss franc stable, the Swiss National Bank had to buy $100,000 in sovereign bonds (of other nations) per resident. There was a lot of financial convincing that had to be done before anyone took the SNB seriously…or maybe the market turned for another reason. Or maybe the SNB satiated the market…but it took a lot of satiating. No one knows.

    Add on: Every central banker on the planet is publicly tossing the policy ball into the fiscal court…pleading no ammo. So…how do central bankers convince anybody of anything, with so many of the tribe whimpering the ammo boxes are filled with emptied shells?

    Put on your tin-foil hats, grab your favored intellectual tools, and let us continue on the great snipe-hunt for the best monetary policy!

    Or is that monetary-fiscal policy?

  2. Gravatar of bill bill
    25. January 2020 at 05:01

    Basically, the Fed is doing more than reasoning from a price change. They are trying to steer the whole economy with a price change, forward guidance about a price change, etc.

    Related topic. What are some of the better sites or papers that support the new Fed ways? These are smart people who used the 2008 crisis as an opportunity to switch to a different set of tactics. The changes don’t seem to bring much value (at times, even cause harm), so why the change?

  3. Gravatar of Michael Rulle Michael Rulle
    25. January 2020 at 07:24

    I am glad you raised these issues, in particular the cause of the repo illiquidity problem and the possibility that policy makers can and will often not do what MM calls for. As for the former, I asked you to comment on that in Sept/Oct and you specifically responded as if it were not a relevant question—-you did. I am not saying this to say “see I told you so”—-I am not that stupid—-I raise it because it concerns me as to why at that time you did not think it was important (I didn’t know—-I was just asking). I am trying to be helpful here. I also worry about the randomness inherent in all bureaucracies, central banks included. I say that with regard to your “no recession” future. Just a comment as always—-thinking out loud as always—-I know very little.

  4. Gravatar of Michael Rulle Michael Rulle
    25. January 2020 at 07:29

    Re: repos——I said you said it was not relevant.It is more accurate to say you expressed the opinion in a way that you did not see it as important

  5. Gravatar of ssumner ssumner
    25. January 2020 at 10:37

    Bill, It was something they had wanted to do for a long time, and they used 2008 as an opportunity to implement the change.

    Michael, The reason I don’t believe the September spike in rates was important was because it was very brief, and thus had no macroeconomic consequences. I still believe that. Others do view it as important, and a simpler policy regime is more likely to avoid that problem.

  6. Gravatar of rayward rayward
    25. January 2020 at 12:42

    Banks increased reserves because, why? Because IOR appealed, yes, but also because the alternative (lending) didn’t. And, unfortunately, still doesn’t, at least not at the level one would expect at this point in the business cycle. Trump’s tax cut was supposed to free up billions for investment in productive capital, increase productivity, and increase wages. Not. Sumner’s friend Cowen blames lack of dynamism: folks are, well, not very ambitious. What? Business shifts production to China, ordinary Americans become less dynamic, and who’s to blame? The Fed? The Banks? Un-American Americans? The Chinese? [An aside, I am on the front lines in dealing with bank lending. And guess what: it’s about short-term lending (five years). Would you be “dynamic” with a five-year horizon? Economists focus on their friends investing in tech, with the quick and often enormous returns, while the rest of us deal in the real world.]

  7. Gravatar of Thaomas Thaomas
    25. January 2020 at 13:17

    I think the “ammunition” fallacy is simpler: it is based on the assumption that the only instrument of monetary policy is control of ST interest rates and that rates cannot be reduced to or below zero.

    “I believe the greatest risks to NGDP are to the downside, which is why the Fed should adopt a slightly more expansionary policy. We need a policy where the risks are balanced. We are close, but not quite there yet.”

    Even if the risks to NGDP are on the upside I think policy should be more expansive and explained as such “because price level is still below the target needed to achieve symmetric inflation since 2008.” It cannot be clear to markets that the Fed has really repented of it’s failures of 2008-present. For someone who favors NGDP targeting, the announcement would be explained as “because NGDP is still below the target needed to achieve 5% p.a growth since 2008.”

  8. Gravatar of Thaomas Thaomas
    25. January 2020 at 13:27

    @ rayward:
    “Trump’s tax cut was supposed to free up billions for investment in productive capital” No one could possibly believe that borrowing money to reduce the taxes on a bunch of rich people when the economy was near full employment could “free up resources” for investment. Unless the rich have a MPC of zero,or less, resources for investment must have gone down.

  9. Gravatar of dtoh dtoh
    25. January 2020 at 20:38

    Basically agree with everything you say, but in order for the Fed to reduce banks reserves, they will have to buy a ton of assets (otherwise the money gets pushed out into the non-banking section and you get massive inflation.)

    So here’s another policy approach. Fed completely stops OMO (except for overnight or seasonal adjustments) and just pushes money out into the economy by causing changes (i.e. reductions) in bank reserves by adjusting the IOR rate. Given the level of reserves and assuming a 4.5% NGDP growth rate and constant velocity, the Fed could go 25 years without ever having to buy any assets.

  10. Gravatar of dtoh dtoh
    25. January 2020 at 20:42

    Correction > Fed would have to SELL a ton of assets

  11. Gravatar of Willy2 Willy2
    26. January 2020 at 00:31

    – There is no “ammunition problem”. The FED or any other bank can increase the “excess reserves” at will. But lending by (commercial) banks is unlimited (in theory) because these banks can create as much “money out of thin air” as they want. But lending requires 1) someone who wants to borrow 2) rising asset prices (stocks, bonds, real estate & commodities).

  12. Gravatar of Thaomas Thaomas
    26. January 2020 at 10:46


    Well then just sell enough not to get “tons” of inflation but enough to get the PL back to where it would be if the Fed had maintained average inflation of 2% since 2008 (or NGDP if they had maintained average nominal growth at 5%).

  13. Gravatar of Benjamin Cole Benjamin Cole
    26. January 2020 at 16:17


    If the Federal Reserve buys or sells assets, it is doing so by entering or leaving globalizedc capital markets. Also, money is a fungible commodity.

    What impact do you expect Federal Reserve buying or selling of assets to have on the domestic economy of the United States? How will those impacts be manifested?

    What will the actions of other central banks, also buying and selling assets in globalized capital markets, have on the domestic economy of the United States?

    If we have globalized capital markets and money is a fungible commodity, is there any difference between what a central bank does in Europe versus what a central bank does in Asia versus what a central bank does in North America?

  14. Gravatar of Brian Donohue Brian Donohue
    27. January 2020 at 07:10

    Things move quickly, unless they are the Fed. Right now, we are paying banks 1.55% to stash reserves risk free. Meanwhile, the yield on the 7-year Treasury is 1.53%. Why? Markets are forecasting 25-50bp cuts this year. Hop to it Jerome.

  15. Gravatar of msgkings msgkings
    27. January 2020 at 08:46


    As mentioned before, Trump’s constant ripping of the Fed is actually making it harder for Jerome to actually cut rates and look like he’s Trump’s lapdog.

  16. Gravatar of Brian Donohue Brian Donohue
    27. January 2020 at 10:04

    @msg, maybe but Trump hasn’t said anything about the Fed for the past couple months and Powell should be professional enough to do his job.

  17. Gravatar of msgkings msgkings
    27. January 2020 at 11:31


    Actually, September 18:

    September 11:

    And so on. He never stops.

  18. Gravatar of Thaomas Thaomas
    28. January 2020 at 12:03

    @ B Cole
    I expect that if the Fed said (and then appeared to be doing what it said) that it would buy assets until the Price level in the US were back where it would have been if it had achieved 2% inflation since 2008, the price level would soon be back to where it would have been if they’d achieved 2% inflation since 2008. And since this would probably make businesses less likely to fear a recession in the future, investment demand (including investing in finding and training less suitable workers) would pick up.

Leave a Reply