Don’t target interest rates

John Cochrane has a very long and interesting post, which advocates that the Fed target interest rates. I respectfully disagree.

Here’s Cochrane:

But money has disappeared from more recent economic thinking. My preferred model of the world (fiscal theory of monetary policy)  has an interest rate target, which sets expected inflation, fiscal theory which sets unexpected inflation, and money is not needed.  Conventional academic wisdom uses new-Keynesian models with `active’ interest rate rules to produce determinacy. Older-school ISLM style models, which are still used by the Fed and capture  completely the verbal explanations Fed officials offer for monetary policy, also are based entirely on interest rate targets.

I recall that Cochrane favors a NeoFisherian model of interest rates and monetary policy, which suggests that a lower interest rate peg will lead to lower inflation. But central banks have the opposite view, that lowering interest rates will raise inflation.

That doesn’t reassure me. It’s as if the designer of a bus insists that turning the steering wheel to the right makes the bus go to the right, while the actual drivers of the bus believe that turning the wheel to the right makes the bus go left. What could go wrong?

Even worse, I believe that turning the wheel to the right sometimes makes the bus go right and sometimes makes the bus go left. I’d prefer a different steering mechanism—one that wouldn’t be misused, causing NGDP to plunge 8% below trend in 2008-09.

I don’t think Friedman would be happy, as he might say inflation happens with long and variable lags, and the money demand shift happens before you can see the inflation. But Friedman is no longer with us, and doesn’t get the chance to modify his views with the evidence that inflation has ended under interest rate targets, and the amazing period of the zero bound, which turns on its head the experience of 1940s and 1950s interest rate pegs that so influenced him.

Friedman died in 2006. I don’t see anything that has happened in the past 13 years that would have surprised Friedman. He lived through long periods of the US being at the zero bound in the 1930s and 1940s, and also the Japanese experience of the late 1990s and early 2000s. We know his views on Japan. Our recent experience offers nothing new.

Monetarists always talked about how velocity is interest elastic in the short run it was “stable” in the long run. But it’s not. Money demand — reserve demand especially — has exploded by a factor of 300 at i−imi−im, and it’s not ever coming back as long as that is the case.

Monetarists did not favor targeting reserves, or even the monetary base. They favored targeting a broad aggregate such as M1 or M2. Just to be clear, I don’t favor targeting any sort of M, but recent moves in M2 velocity are nothing like recent moves in reserve velocity. Since 1959, M2 velocity has stayed within a range of 1.4 to 2.2. Again, reserve demand rose sharply during 1930s and more recently in Japan. It’s a predictable response to the zero bound, made worse by the payment of interest on bank reserves. (A contractionary policy that Friedman would have opposed in late 2008, even if he supported the general concept in normal times.)

In sum, the screaming lesson of the last 10 years in the US, and 25 in Japan, is that a “liquidity trap” with arbitrary reserves does not cause any inflation. 

And what was the lesson of 1932-51 (when short-term rates were near zero?) I’d say the lesson was that near-zero rates and large reserve holdings don’t cause inflation when the natural rate of interest is low, and do cause inflation when the natural rate of interest raises above the policy rate. Have we learned anything new from the recent zero bound episodes? I’d say no.

Instead of targeting interest rates, we should target NGDP futures prices.


Tags:

 
 
 

27 Responses to “Don’t target interest rates”

  1. Gravatar of rayward rayward
    2. April 2019 at 11:50

    Why do investors choose to invest? Wouldn’t it depend on the investment? If the investment were productive capital, then the expectation of economic growth should trigger the investment. On the other hand, if the investment were in financial assets or real estate, then the expectation of rising asset prices should trigger the investment. Powell learned a hard lesson this Fall: for many investors, the expectation of rising asset prices is the most important factor. The neo-Fisherian model is inconsistent with investment determined by an expectation of rising asset prices; instead, it works, if at all, when investment is determined by an expectation of economic growth – an increase in interest rates sends a message that the economy and thus inflation will be growing and this expectation triggers investment in productive capital. [Note to Sumner: just ignore my comments; insults aren’t becoming.]

  2. Gravatar of Christian List Christian List
    2. April 2019 at 13:49

    John Cochrane has a very long and interesting post, which advocates that the Fed target interest rates.

    Scott,

    Cochrane’s post is indeed very long but does he really advocate targeting interest rates? Is this really the main point of his post? I don’t think so.

    For example, he seems to end with:

    Next up: The target is up for review too. Price level target? Nominal GDP target?

    There’s still hope!

  3. Gravatar of ssumner ssumner
    2. April 2019 at 15:05

    Rayward, You said:

    “Powell learned a hard lesson this Fall: for many investors, the expectation of rising asset prices is the most important factor.”

    Slow down, I want to write this stuff down before I forget it.

    Christian, You said:

    “Cochrane’s post is indeed very long but does he really advocate targeting interest rates?”

    If he didn’t, he sure fooled me.

  4. Gravatar of Benjamin Cole Benjamin Cole
    2. April 2019 at 15:44

    Money demand — reserve demand especially — has exploded by a factor of 300 at i−imi−im, and it’s not ever coming back as long as that is the case.

    What is “i-imi-im”?

    I agree with this post; central banks should target a result, not a process. Get the result you want, whether through interest rates, quantitative easing, money financed fiscal programs.

    If it is true that central banks can buy back national debt without incurring inflation, then there are a few jokers in the deck—more than a few.

  5. Gravatar of John Cochrane John Cochrane
    2. April 2019 at 18:32

    I actually agree with most of this. Scott points out that we’re really not sure which way interest rates are connected to inflation. I’m not sure either. I am sure that scientific knowledge on this point is weak. Scott talks a lot about M and M2, but does not advocate a return to money targeting. Neither do I. Scott ends by arguing “Instead of targeting interest rates, we should target NGDP futures prices.” I sort of agree. My preferred wild idea is to target CPI futures, or, equivalently, the spread between indexed and non-indexed debt. Like Scott (I think) these should be not only target but instrument — the Fed should trade these, not use them as a guide to short term interest rates, which is what most people mean by “target.” I do not advocate an interest rate target, either in the post or generally. In the post, I think about how the Fed should implement an interest rate target given that the Fed has decided it wishes to follow an interest rate target. That’s the question the Fed is asking right now. Sometimes it’s useful to say “you shouldn’t ask that question, you should do something else entirely.” Sometimes it is useful to say “well, ok, given that you want to target interest rates, here is how to do it.”

  6. Gravatar of Ralph Musgrave Ralph Musgrave
    3. April 2019 at 02:47

    Scott ends by saying “Instead of targeting interest rates, we should target NGDP futures prices.” I’m baffled. I’ve nothing much against aiming to increase NGDP by some specified amount per year, but that leaves open the question as to how to do that. Adjusting interest rates is a certainly a possibility. I.e. an annual increase in NGDP is what is aimed at, while interest rate adjustments are way to get to the latter target. Ergo the two are not alternatives.

  7. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    3. April 2019 at 06:53

    ‘Friedman died in 2006. I don’t see anything that has happened in the past 13 years that would have surprised Friedman.’

    I know it wouldn’t have, as he told me in the 1990s (and Walter Heller in the 1960s): Interest rates are not the price(s) of money.

    Why is this so hard for the world’s economists to understand.

  8. Gravatar of Carl Carl
    3. April 2019 at 08:14

    Ralph:

    One challenge for interest rate targeting is periods of good deflation due to increased productivity. Interest rate targeting could end up being pro cyclical.

  9. Gravatar of ssumner ssumner
    3. April 2019 at 08:26

    John, Thanks for replying. Here is one reason I assumed you were advocating an interest rate target:

    “The Fed should not target the supply of reserves at all. The supply curve of reserves should be horizontal. The Fed should just say, “bring us your treasuries, and we’ll give you reserves and pay the IOER rate.” Or, “Bring us your reserves and you can have treasuries.”

    Why? Well, if you want to target a price, you offer to buy and sell freely at that price. If you want to target an interest rate, target an interest rate. ”

    It seems to me that if you want abundant reserves (as you do), and you favor the Fed setting IOER, then the setting of IOER is effectively an interest rate target.

    Your post also says:

    “An interest rate target that varies with the price level or inflation rate has all the advantages of a peg — it provides elastic money when there are velocity shocks — but it avoids the monetarists’ complaint, in that it does not accommodate higher price levels. We can have our cake and eat it too.”

    Rereading, I could not find any outright endorsement of interest rate targeting, so it looks like I was reading between the lines. Sorry about that. But the passages I quoted sort of created that impression.

    Glad we agree on futures targeting. I am happy to use futures as a policy instrument, and indeed proposed that in a 1989 paper. However, for pragmatic reasons my latest proposal (dubbed “guardrails”) still uses OMOs of Treasuries as a policy instrument. I agree that your approach is a more elegant solution.

  10. Gravatar of Ewan Ewan
    3. April 2019 at 09:26

    I’ve come late to this. Could someone refer me to a summary of how NGDP targeting is different from and better than the broad money monetarism of Profs. Laidler and Congdon, which in my day (last century!) was the only effective guidance to central bankers? Tim Congdon even used NGDP as a rule of thumb.

  11. Gravatar of John Cochrane John Cochrane
    3. April 2019 at 10:52

    Scott:

    I admit to being sneaky. The post says “IF you want to target interest rates here’s how to do it.” I carefully didn’t endorse the if, but didn’t do anything to argue against the holy writ that this is what the Fed should do. In this discussion they don’t want to hear from us about fundamental changes, or from others about money targeting, gold standards, etc. So, save it for another day. And while I think there is something better, I don’t see an interest rate target as a terrible or unworkable policy.

    The 10 years of lessons business is a reference to “michelson morley fisher and occam”

    https://faculty.chicagobooth.edu/john.cochrane/research/papers/320201.email.pdf

    That we lived under a peg for 10 years, and Japan 25, and all that happened is the Friedman rule, is, I think, a really decisive experiment. No deflation spirals, no multiple equilibria, no hyperinflation, all widely predicted. But that’s a topic for another day.

  12. Gravatar of Christian List Christian List
    3. April 2019 at 10:54

    Thank you John, this is gold. Someone was understanding you better than Scott. As a non-native speaker and non-economist. Sometimes mental distance is quite helpful.

    Scott, you said about half a dozen times that I have “zero reading comprehension”. That might have been true or not, I found it too harsh, so I might remember this episode here, as a little counter example, in case you ever say it again.

    Ewan, on the right side of his blog, Scott has a lovely section called “Quick intro to my views”. That’s a good start. Not to mention Google.

  13. Gravatar of ssumner ssumner
    3. April 2019 at 13:53

    Ewan, You might google one of my papers on using futures markets to guide policy, particularly the more recent “guardrails” approach paper.

    Thanks John. But I’m still not sure how you feel about my claim that IOER is basically an interest rate target, at least in an economy with abundant reserves. Am I correct in that claim? Does your comment about not endorsing interest rate targeting also mean you do not endorse IOER combined with abundant reserves, as that regime seems to be equivalent to interest rate targeting? Or is it possible to set IOER at a different position from short-term market rates, in an economy with abundant reserves?

    You said:

    “I think, a really decisive experiment. No deflation spirals, no multiple equilibria, no hyperinflation, all widely predicted. But that’s a topic for another day.”

    It’s always nice to be reminded that market monetarists went out on a limb and said Fed policy during 2008-18 would not lead to high inflation or deflation spirals, and we were later shown to be correct. 🙂

    Christian, Let’s see if we can get clarification from John, in a reply to my latest comment. Or maybe you can answer those questions for him. 🙂

  14. Gravatar of Michael Rulle Michael Rulle
    3. April 2019 at 14:16

    Scott, this was interesting. Even very interesting smart people like you and John can have uncertainty about e.g., how interest rates relate to inflation. But I need clarification on a point. John said he would like to use CPI futures as the instrument by which to target inflation. You seemed to respond, if I understood correctly, that you found that “more elegant” than targeting nominal GDP using futures. I believe that it is impossible for my perception to be accurate. I would faint. Are those two targeting ideas equivalent? If not, what did you mean? If they are I will need to think about it.

  15. Gravatar of Benjamin Cole Benjamin Cole
    4. April 2019 at 02:43

    https://www.bostonglobe.com/opinion/2019/03/20/there-better-model-explain-economic-trump-era/gguXuRUUcaxlyZPeSe3kIO/story.html

    Not a bad defense of MMT.

    I am sure about MMT voodoo. But then orthodox macroeconomic shamanism is not compelling either.

  16. Gravatar of Ewan Ewan
    4. April 2019 at 04:21

    My education continues apace – I am working through the articles and papers. I know I will sound archaic (of my time), but would appreciate guidance from someone.

    A “natural” or equilibrium real interest rate (which I understand plays a large part here in determining how tight monetary conditions are) used to be impossible to infer with any practically useful accuracy. The TIPs market used to be too small (likewise the UK ILG market) to escape the needs of pension funds and their actuaries – who pushed prices up.

    Market monetarism, as I understand, differs from traditional monetarism in relying on market forecasts, but concurs with it in rejecting the consensus economic model – which is the basis for the market’s forecasts.

    I also remember the markets forecasting lots of recessions that didn’t happen and missing one or two humdingers that hit it in the face.

    I’m getting tangled up in terminology: NGDP is a target, a “target” interest rate is an instrument, is it not, for hitting the ultimate target?

    As I say, my education continues, but I’m already a fan.

  17. Gravatar of ssumner ssumner
    4. April 2019 at 08:25

    Michael, No, I meant that his proposal to use futures prices as the instrument is more elegant than just using them as a target. Thus in his plan, CPI futures would be bought or sold in unlimited amounts to keep the price pegged, at least if I understand him correctly. That was an idea I proposed (invented?) in a 1987 paper. I used NGDP, however.

    Ewan, Good questions:

    The natural rate of interest is almost impossible to measure, as you say. It’s existence can only be inferred from macroeconomic data. In addition, there is no single natural rate. The natural rate with a 2% inflation target is higher than with a 0% inflation target.

    TIPS spreads can be slightly distorted by a risk premium, as you suggest. They are still somewhat useful at forecasting, in a rough sense.

    Yes, MMs favor a policy setting where the market consensus forecasts on target NGDP growth. The $64,000 question is how best to infer the consensus market forecast. I now favor a guardrails approach, looking for unusual trading activity in either direction on long and short NGDP futures contracts.

    The “target” terminology” is a mess, as you suggest. The actual policy instrument, prior to 2008, was open market operations. Interest rates were targeted, and the rate was occasionally adjusted to “target” inflation. It would be better to have a differnet term for the short-term target (like interest rates), and the long-term goal (like inflation or NGDP).

    With the advent of IOR in 2008, interest rates are also an instrument that is directly controlled by the Fed.

  18. Gravatar of Ewan Ewan
    4. April 2019 at 08:50

    I’m becoming a right pest, I know. So, after this, I will keep shtum till I’ve read more papers.

    Before I ask my final (for now) question: In the first several years of TIPs, and in ILGs at the same time, the demand from pension funds was such as to make the real yield highly distorted (not slightly). Is the market so much deeper now?

    My “final” question (you wish) – which may again just be me misunderstanding:

    What does central bank buying and selling of NGDP futures add to an announcement of an NGDP target?

    The central bank surely does not influence the economy just by playing with people’s expectations. It alters the composition of their asset portfolio. It does so chiefly through the banking system by its open market operations (which affect the banks’ ability to extend credit and thus create money), but also by QE when required (which may also be through the banking system). The central bank is also the banks’ banker. It ensures that solvent banks remain liquid. It is of the essence of the central bank’s purpose to interact with the banking system. Why then would it be deemed sensible to separate them? Who else would stop the banking system imploding as in days of yore? Who else has the potential to change the rate of growth of the aggregate money supply?

  19. Gravatar of Ewan Ewan
    4. April 2019 at 08:55

    Oops, two seconds into my promised silence and I’m already blabbbing again.

    My point is that Market Monetarism advocates reliance on market forecasts which, it also deems, reflect a mistaken model of how the economy works. Shome mishtake shurely (as the famous Fleet St. editor Lunchtime O’Booze used to say – in the days before Rupert Murdoch).

  20. Gravatar of Christian List Christian List
    4. April 2019 at 09:27

    @Ewan

    What does central bank buying and selling of NGDP futures add to an announcement of an NGDP target?

    As I read it the cb is of course not trading any NGDP futures. I assume you need a (private) NGDP futures market to know, for example, the stance of monetary policy in real time. Interest rates won’t tell you that.

    To me the connection between money (as medium of account), and wages, and NGDP seems to be extremely important, so the question arises: why not target NGDP as directly as possible (and keeping NGDP growth stable)?

    Or maybe you can answer those questions for him.

    Scott,

    you got me, I was right this one time, I won’t pressure my luck again, with questions that are way over my head. =)

    @Benjamin Cole

    From the piece that you linked to:

    inflation has also been far lower for longer than virtually any economic model would have predicted, given current robust US growth and very low unemployment.

    You know that this is a lie. Scott predicted it over and over again, with pretty conventional monetarism.

    Galbraith is a phony. I never read a single interesting piece by him. Just one old, wrong idea after another. He’s right about one thing though: MMT is just Galbraith 2.0, a cheap copy – and even more stupid.

    “Modern Monetary Theory” should be about market monetarism. The abbreviation does not even have to be changed, it fits well already: MMT = Modern Monetary Theory = Market Monetary Theory.

  21. Gravatar of ssumner ssumner
    4. April 2019 at 18:03

    Ewan, The buying and selling of contracts is a form of QE, as it influences the base.
    My preferred system is actually just to buy Treasuries, and use futures contracts as just a “guardrail”. Like the beeping noise a truck makes backing up.

    You said:

    “My point is that Market Monetarism advocates reliance on market forecasts which, it also deems, reflect a mistaken model of how the economy works.”

    How about this:

    My point is that Keynesian advocates reliance on computer models which, it also deems, reflect a mistaken model of how government bureaucracies work.

    Look, the worst case is that if my plan does work then I get rich. Is that so bad? If NGDP future prices don’t reflect rational forecasts, I can trade the market and get rich. Thus in 2008 I could have sold them short and got rich, knowing NGDP was plunging.

    Yes, markets are not perfectly efficient. But trust me, it’s really hard to beat the market. And recall that only time varying risk premia matter in the long run, so the pension fund example you mention for TIPS actually would not be a problem, if the distortion was stable.

    My sense is that the TIPS market is now pretty deep, but I still think the TIPS spread gets distorted during major crises. Fortunately this distortion overestimates the expected rate of deflation, which triggers easier policy. We need that anyway in a deep crisis as real output is also falling. (Dual mandate, etc.)

    Christian, You were right, but I still don’t understand how you can pay IOR and have abundant reserves without targeting interest rates. Anyone else have an answer? I’m stumped.

  22. Gravatar of Ewan Ewan
    5. April 2019 at 01:13

    Trading futures is surely a very minor form of QE compared to the alternatives?

    I take the point about time-varying risk premia.

    So Market Monetarism and Keynesianism(s) in their policy advice rely on forecasts they assume to be false. Market Monetarism may be an improvement if the wisdom of crowds is a thing.

    On the evidence of the long past (Prof. Laidler in Canada and Prof. Congdon in the UK) broad money monetarism pragmatically applied can do better – indeed Prof. Congdon has demonstrated as much since the 1970s.

    Believe me, I know it is impossible to beat the market. (It does not follow that the market forecasts are particularly good.) The central bank is not in the business of trying to beat the market (as also in the matter of “bubbles”).

    The central bank is at the centre of the banking system and conducts monetary policy primarily through the banking system, whose efficient functioning is its direct responsibility. The notion of separating the two seems to me unwise.

  23. Gravatar of Ewan Ewan
    5. April 2019 at 01:46

    Here’s a thought. Do your comments on statism and bureaucracy not argue for the free banking notion of the banks organising their own clearing house or central bank?

  24. Gravatar of Ewan Ewan
    5. April 2019 at 03:20

    … or, if you accept a role for a central bank, you acknowledge something the state and its bureaucracy can do better than the market. There is then (possibly) something circular about relying on market forecasts to influence market expectations.

  25. Gravatar of JP Koning JP Koning
    5. April 2019 at 05:46

    Scott, I think you have Cochrane wrong about targeting interest rates. In his recently added post-script, he even agrees with you that central banks should target NGDP:

    “Update

    Scott Sumner posts some interesting comments. I agree with most of them. Scott points out that we’re really not sure which way interest rates are connected to inflation. I’m not sure either. I am sure that scientific knowledge on this point is weak. Scott talks a lot about M2, but does not advocate a return to money targeting. Neither do I. Scott ends by arguing “Instead of targeting interest rates, we should target NGDP futures prices.” I actually sort of agree. My preferred wild idea is to target CPI futures, or, equivalently, the spread between indexed and non-indexed debt. Scott says I advocate an interest rate target, which I do not. In this post, I think about how the Fed should implement an interest rate target given that the Fed has decided it wishes to follow an interest rate target. That’s the question the Fed is asking right now.”

  26. Gravatar of ssumner ssumner
    5. April 2019 at 13:52

    Ewan, It’s misleading to talk about the state vs. the market with monetary policy. Because of network effects, there’s likely to be a single medium of account. The provider is a monopoly, although you may allow free private banks to produce banknotes backed by the medium of account. In any case, there must be some sort of “monetary policy” for the medium of account, whether done by a public or private monetary authority.

    Even merely defining the dollar as 1/35 oz. of gold is a sort of “policy”.

    Under the guardrails approach the central bank is free to ignore bubbles. Of course if bubbles don’t exist (which is my view) then central banks will lose lots of money by ignoring market signals seen as a bubble. If they do exist, as most people assume, central banks will make lots of money beating the market. I doubt it. But I’m happy to let them try.

    I’d prefer to completely separate the monetary authority from the banking system. Have the Treasury do bank regulation, bailouts, etc., if needed. Have the monetary authority focus like a laser on base supply and demand.

    I disagree with Congdon on the monetary aggregates, although I agree with him on many issues.

    JP, Yes, I saw that (in the comments here as well.) But he didn’t answer my question of how he can square his non-support for interest rate targeting with his support for IOER combined with abundant reserves. It seems to me that that policy inevitably implies an interest rate target, equal to the administered IOER.

  27. Gravatar of Ewan Ewan
    6. April 2019 at 02:14

    I know I said I wouldn’t keep going on, but…

    There are many happy to see a monopoly in private hands. There are some happy to see the currency. The argument for a central bank, rather than a clearing house run by the banks themselves, is that the central bank will be backed by the full tax-raising powers of the state and so better able to guarantee liquidity. The central bank thus has two duties: to maintain the value of the currency and to maintain the liquidity of the banking system (by providing liquidity to solvent but illiquid banks, not to bail out insolvent). It maintains the value of the currency principally through its transactions with the banking system, which allow it to influence the private sector’s asset allocation, prompting the private sector to adjust it if the central bank’s actions have caused it to move away from what the private sector prefers (i.e. too much or too little money – currency and bank deposits – versus other assets). The focus on broad money follows from this.

Leave a Reply