A very subtle distinction

It will be easier to understand this post if you first read the previous post.  The Fed delivered two monetary shocks today.  The first occurred at 2:15pm, and caused yields on 2 and 5 -year bonds to increase.  The second occurred about 30 minutes later, and caused yields on 2 and 5-year bonds to fall back, and end the day slightly lower.  And yet both shocks seemed “contractionary” in some sense.  Obviously there are some very subtle distinctions here, which require an understanding of Keynesian and Fisherian monetary shocks. More specifically, the first shock was Keynesian contractionary and the second was Fisherian contractionary

At 2:15 the Fed raised its target rate as expected, and also indicated that another two rate increases are likely next year.  This announcement was a bit more contractionary than expected, especially the path of rates going forward.  As a result, 2 and 5-year yields rose, while 10-year yields fell on worries that the action would slow the economy, eventually leading to lower rates.

Later, the markets became increasingly worried that the Fed was not sufficiently “data dependent”, that it would plunge ahead with “quantitative tightening” and that the Fed stance on rates (IOR) would be too contractionary.  Watching the press conference, I had the feeling that Powell might have wanted to be a bit stronger in emphasizing that no more rate increases are a clear possibility, but felt hemmed in by his colleagues at the Fed.  But perhaps I was reading into it more than was there.  In any case, Powell said “data dependent”, but didn’t really sell the markets that he was sincere, or as sincere as the market would wish.  This monetary shock probably reduced NGDP growth expectations, and drove 2 and 5 year yields lower.

Do you see how utterly inadequate it is to talk about monetary policy in terms of interest rates?  Even the “sophisticated’ new Keynesian view that it’s the future path of rates that matters is nowhere near adequate.  The 2 and 5-year yields can go up with tightening, or they can go down with tightening.  Today they did a bit of both, within one hour.

If you insist on using Keynesian language, you might say the 2:15 action tightened primarily by raising expected future rates relative to the natural rate, by raising the expected path of the policy rate.  In contrast, the press conference impacted the gap by depressing the natural interest rate by depressing NGDP growth expectations.

PS.  The discussion on CNBC involved lots of people dancing around the “circularity problem”, i.e. the Fed looking at markets while the markets look at the Fed.

PPS. I laughed when a CNBC person suggested Trump might use the Fed as a “scapegoat”.  Hello, who appointed all the top Fed officials?  Has scapegoating the Fed ever worked for any President, even when they weren’t his appointees?

Update:  I forgot to mention Powell’s most interesting comment.  He says the Fed reappraisal of policy techniques planned for this summer in Chicago will look at the zero bound problem in monetary policy.  I view that as really good news.



36 Responses to “A very subtle distinction”

  1. Gravatar of Brian Donohue Brian Donohue
    19. December 2018 at 14:17

    Good post.

    I observe, however, that long-term yields (10-year and especially 30-year) fell a lot on the first piece of news, and some more on the second. In the past six weeks, the 30-year Treasury yield has fallen 0.46%. Extraordinary.

  2. Gravatar of Marcus Nunes Marcus Nunes
    19. December 2018 at 14:41

    It gets worse!
    From Vitor Constâcio, until earlier this year ECB VP:
    “The FED correctly prioritized economic data over stock market greed wanting no more hikes next year. A significant economic downturn is predicted for 2020 and the FED wants to have by then higher rates to respond to the downturn.”
    Maybe he wants the “significant downturn” to happen in 2019!

  3. Gravatar of TallDave TallDave
    19. December 2018 at 17:05

    That is good news.

    Looks like we might get a chance to see if the Fed learned the right lessons in 2007-9.

    People have almost forgotten that the markets didn’t always jump around on tweaks to Fed policy. Scott, didn’t you once have a graph that showed when that correlation started? Wonder how that’s held up.

  4. Gravatar of Benjamin Cole Benjamin Cole
    19. December 2018 at 18:55

    Meh. The Fed has been too tight since 2006 or so. It still is.

    Tyler Cowen, who professed the 2008 downturn was due to huge and long-term structural imponderables, now has reversed and says it was cyclical. The Fed did it, in other words.

    Dudes, print more money.

    The world has changed. The macroeconomic risks are on the downside. See Japan. See Europe. See the US. I will take bouts of 3%-4% inflation anytime to dodge recessions. The present insanity is that a recession is worth it to avoid 3% inflation. This is index-worship.

    Yes, the Fed will hit zero bound in the first inning of the next recession. They will go to QE slowly, but no one is really sure if QE works, at least in the volumes that are politically palatable.

    (Side question: If the globe has global capital markets, and they are fluid, what does $2-$4 trillion of QE really accomplish? The globe has hundreds of trillions of assets. Now, Michael Woodford contends that QE plus federal deficits are really money-financed fiscal programs. Man, this is blurry, no?)

    The Fed needs to ponder helicopter drops, aka money-financed fiscal programs. I suggest tax holidays on payroll taxes in every recession.

    Next question: The Fed is conducting reverse QE, in conjunction with decreasing endogenous money creation (due to slackening property-lending).

    How will this end up?

  5. Gravatar of Michael Sandifer Michael Sandifer
    19. December 2018 at 23:41

    Benjamin Cole,

    I think Scott is right when he says the Fed has a pattern of letting inflation drift above target and then overreacting with tightening, which has caused many US recessions. This is due, in part, to usinginflation targeting regime.

    I also think policy is too tight, but if we raise inflation, it should be within an NGDPLT regime.

  6. Gravatar of Benjamin Cole Benjamin Cole
    20. December 2018 at 01:55

    Michael Sandifer–

    Thanks for your comment. I can’t say you are wrong (in part as no one is ever wrong in macroeconomics).

    In that case, perhaps the Fed needs an IT range, ala the RBA and some other central banks. Such as 2% to 3%.

    BTW, inflation in Australia did reach into 3% and even 4% in the last several decades but somehow it drifted back down into the target range without a recession. Whether this was luck or skill on the part of RBA….

    But in general, I think the RBA has been successful because the Inflation Bogeyman today is largely a phantom. See Europe, Japan, US, Thailand….

    Meanwhile….from Reuters today.

    “Bank of Japan Governor Haruhiko Kuroda warned of heightening risks to the economic outlook and signalled the central bank’s readiness to ramp up stimulus if needed, brushing aside concerns it lacks policy ammunition to stave off another recession.”


    Now, that is a central banker.

  7. Gravatar of Michael Rulle Michael Rulle
    20. December 2018 at 05:57

    It’s nice to have more theoretically explanatory models, which does also add the risk of being wrong on more variables, both from a model making perspective and from a variable input perspective. But it seems prudent to do so. How about the model of letting lnterbank lending decide the interest rate. Your entire discussion, and certainly various markets’ reaction to the Fed, was a reaction to the Fed. Why should we have a central body which appears to engage in hand to hand combat with market participants? Because models are better than free markets. I don’t understand your overall perspective how markets should work. You believe in efficient markets yet you also believe in the giant spider in the center of the web fanning around unknowables. I like Greg Ip’s closing comment this morning. “The stock market has forecast 9 out of the last 5 recessions. But the Fed has forecast none”. He may not really believe it, but today he does. Your NGDP theory is nice because it seeks market feedback and the Fed is supposed to respond. But it is still not free markets. I know before 1914, or whenever, we had more sharp but short recessions. But in the last 18 years we have seen the stock market drop 50-60 % twice.

  8. Gravatar of Michael Rulle Michael Rulle
    20. December 2018 at 06:31

    To clarify a bit. I am not suggesting we have no Fed. I am asking the question “ what is wrong with a “model” of banks determining the interbank rate”. Why are rates (regardless of whether up or down necessarily means tightening or loosening, after all free markets presumably factors all of that in). Why do we have one market where the government determines the price? I don’t understand.

  9. Gravatar of Michael Rulle Michael Rulle
    20. December 2018 at 06:44

    One more point. You believe, as does apparently Bernanke, that he was tight in 2008 and before, thus (Causing”?) making the recession worse. I always liked Greenspan. Even Greenspan does not like Greenspan today. When Russia defaulted on its own currency bonds, and Long Term Capital revealed that everyone had the same trade on in extreme leverage (long risky, short less risky levered 98-99 to 1——-short equity volatility——-long all merger arb trades etc) there was panic. Banks wanted some guarantee of bailouts. Greenspan, in my opinion, replayed JP Morgan in 1907. He basically said “we are not bailing you out” and essentially forced everyone to finance everyone else. The panic soon ended. What Bernanke did (I believe Paulson led the charge) was just the opposite. They exaggerated future losses. Markets do nothing if they still think they will be bailed out. They really really try when they know they won’t.


  10. Gravatar of Matthew D McOsker Matthew D McOsker
    20. December 2018 at 07:27

    I see your point about the focus on rates, but that seems where the Fed is focused – no?

    And, how can they believe that the neutral short rate be equal to or higher than longer rates?

  11. Gravatar of myb6 myb6
    20. December 2018 at 08:45

    Scott, it seems like the market’s worried about a yield-curve inversion but the Fed needs to hit the inflation target. This might be a stupid question, but why doesn’t the Fed just limit the tightening to the long-term market? (And publicize such intentions for expectations-management, of course). They’ve got the long-term assets to sell, which is actually historically abnormal, right?

  12. Gravatar of ssumner ssumner
    20. December 2018 at 12:34

    Brian, I’m not sure that that sort of change in long term bond yields is all that unusual.


    Michael, I agree that banks should determine the interbank rate.

    Matthew, The short term neutral rate is higher because growth right now is much higher than it’s likely to be in two years.

    myb6, This is where the Lucas critique comes in. If you try to manipulate the yield curve then it’s no longer a forecasting tool. Yield curve inversions don’t cause recessions, they forecast recessions.

  13. Gravatar of bill bill
    20. December 2018 at 13:16

    I think the Fed actually did forecast 4 of the last 9 recessions. Those are the recessions we didn’t have.

  14. Gravatar of James Alexander James Alexander
    20. December 2018 at 13:18

    And what did you make of Powell’s mumbling and fumbling about a symmetric 2% inflation target?

    Also, is there any way to get Hypermind to roll over there 12m NGDP Forecasts … it is more important than ever we know what the betting is for Q1 2020

  15. Gravatar of myb6 myb6
    20. December 2018 at 13:22

    Thanks Scott.
    1 Are we sure about that (forecast, not causal)? [[I’m being genuine, not rhetorical.]]
    2 Wasn’t the Twist a manipulation and prices are convenient for untwisting?

  16. Gravatar of Matthew McOsker Matthew McOsker
    20. December 2018 at 13:52

    Scott, it seems that stance absent inflation going past their target (which does not seem likely right now) the Fed is picking recession over inflation.

  17. Gravatar of Christian List Christian List
    20. December 2018 at 18:15

    My gut feeling of a recession in 2019/2020 seems more likely now. The Democrats would be well advised to do anything for it. The FED is incompetent enough to enable it. And Trump and the GOP are incompetent anyway. And isn’t the 10-year cycle over anyhow? Many businessmen believe in this cycle, so it’s also kind of a self-fulfilling prophecy.

  18. Gravatar of Michael Sandifer Michael Sandifer
    21. December 2018 at 05:28

    Christian List,

    Yes, after the Fed’s performance this week, I’ve thrown in the towel. I bought puts, hedged with a smaller position in calls and am prepared for the worst. Before, I was hedging with long duration Treasuries.

    Even if they want to do more next year, Trump’s effort to pressure them could make them delay loosening.

    By the way, this should not be construed as investment advice.

  19. Gravatar of Michael Sandifer Michael Sandifer
    21. December 2018 at 05:32

    I should be clearer and say that I bought calls, but the call positions are smaller than the put positions. Seems pretty safe to assume the high volatility will continue as long as Tariff Man is on the job.

  20. Gravatar of Garrett M Garrett M
    21. December 2018 at 05:41

    Off-topic, but The Atlantic just published an article about why Australia hasn’t had a recession in a long time:


    That’s a topic you discuss frequently professor so I figured you’d be interested.

  21. Gravatar of Christian List Christian List
    21. December 2018 at 09:00


    thank you for the link, it once again confirms my suspicion that monetary policy is virtually irrelevant to the general public and to journalists in particular.

    According to them Australia has gone 27 years without recession because they shower helicopter money on lower-income households, spent heavily on infrastructure, implement fiscal stimulus, never balance the budget, and of course because they hate this evil thing called “austerity”.

    So there you have it, the recipe for 27 years without recession. =)

    No wonder the party landscape worldwide looks as it is.

    These so-called respectable journalists are full of fake news.

  22. Gravatar of dtoh dtoh
    22. December 2018 at 05:41

    I’ve said this a few times, but the easiest way to think about this is a curve with yield (or price) on one axis and the amount of financial assets the non-banking sector exchanges for goods and services on the other axis.

    If rates change, there is movement along the curve, if NGDP expectations change, the curve shifts.

    That’s all there is to it. It’s trivial. Everything in this post and the previous one are just verbose ways of saying the same thing.

  23. Gravatar of ssumner ssumner
    22. December 2018 at 09:17

    James, I suspect the mumbling partly reflects the problem of having to simultaneously speak for oneself and also the committee. One thing I know for sure is that this is a job that I would not be able to do.

    Yes, I think an NGDP futures market would be very valuable right now. All it would take is for someone to donate money to Hypermind, and they’d extend it for another year. Why doesn’t anyone do that?

    I did a recent post at Econlog mentioning the insanity of the Fed not having already set up a NGDP prediction market. The future will look back on this era in disbelief.

    mby6, AFAIK, all the theories of the yield spread as a predictor suggest that it is not causal. It’s predicting a change in future short term interest rates, which may be an indicator of recession. BTW, the current yield spread is still positive, and does not predict a recession.

    Thanks Garrett. I recently did a post criticizing a similar article on Australia. They don’t seem to realize that commodity production makes an economy less stable. And we also showered money on lower income households in 2008. How’d that work?

    dtoh, You said:

    “the amount of financial assets the non-banking sector exchanges for goods and services on the other axis.”

    How is this measured? Is it NGDP? If not, what is it?

    And don’t you need to know why rates change, before deciding on the impact?

  24. Gravatar of Michael Sandifer Michael Sandifer
    22. December 2018 at 10:51


    How comfortable would you be with an NGDPLT regime using an NGDP futures market, but replacing the FOMC with a computer that is simply programmed to keep NGDP on target?

  25. Gravatar of dtoh dtoh
    22. December 2018 at 16:27

    Scott –

    > How is this measured? Is it NGDP? If not, what is it?

    Yes. Basically NGDP. Or the amount of C+I+G/time resulting from the the non-banking sector exchanging financial assets for goods and services.

    > And don’t you need to know why rates change, before deciding on the impact?

    Not really. If expectations don’t change, the relationship between rates and AD is very straightforward . What makes it somewhat more more complicated is when you add in expectations. This causes the relationship between AD and rates to shift. Another way to think of it is AD on the x axis, rates on the y axis, and NGDP expectations on the z axis. If you’re targeting the x axis (AD/NGDP,) it doesn’t really matter how you get there. The Fed has one main tool, OMO. So they just buy or sell enough assets to get to their x (AD) target. If they are credible in their targeting, z becomes a constant, and OMO becomes very routine and requires very little tuning.

  26. Gravatar of dtoh dtoh
    22. December 2018 at 20:34

    And…. everybody….

    You don’t need a frigging NGDP futures market. Christ, the current policy regime generates 500 bp swings in NGDP growth. With NGDP targeting alone, you would reduce volatility by a factor of 10, and since it’s an LT, it doesn’t matter if you overshoot or undershoot by 50bp. You just adjust in the next quarter.

    Pleeeezzzz….STOP talking about a futures market for NGDP. It’s as if you’ve just read “The Dummy’s Guide on how NOT to sell NGDP targeting” ….Lesson 1 – Distract your audience with the details so they forget you were talking about NGDP targeting. Lesson 2 – if you’re selling a golden flying unicorn, get the audience to inspect its hooves.

  27. Gravatar of James Alexander James Alexander
    23. December 2018 at 09:36

    Well, we may not have Hypermind but we do have a market-influenced (part based on reeemt recent trends, too) NGDP Forecast and it isn’t a recession but it isn’t pretty either – text update soon, watch that space.

  28. Gravatar of ssumner ssumner
    23. December 2018 at 09:57

    Michael, I’m fine with that, if it’s a good computer.

    dtoh, You said:

    “Everything in this post and the previous one are just verbose ways of saying the same thing.”

    Then you said:

    “the amount of financial assets the non-banking sector exchanges for goods and services on the other axis.”

    Instead of simply NGDP. Isn’t that being “verbose?” :)

    I do agree that with NGDPLT we could do far better, even without a futures market.

  29. Gravatar of Michael Sandifer Michael Sandifer
    23. December 2018 at 11:03

    James Alexander,

    Your forecast for the change in NGDP this fall mirrors my own. My estimate is that growth has slowed by a bit more than 1%. If we’re right, monetary policy is definitely too tight. I’ve insured my portfolio against the worst with a nice put-weighted strangle.

  30. Gravatar of dtoh dtoh
    23. December 2018 at 16:06


    > Instead of simply NGDP. Isn’t that being “verbose?”

    I just wanted to make sure you knew I was deliberately including G.

  31. Gravatar of James Alexander James Alexander
    24. December 2018 at 06:46

    Our updated comment on the dreadful forecast. Good to see some leadership from the Treasury Secretary, at least.


    Michael Sandifer, it’s my colleague Justin who runs our market-influenced model.

  32. Gravatar of Michael Sandifer Michael Sandifer
    24. December 2018 at 10:22

    James Alexander,

    My estimate after the continued meltdown today is that NGDP growth expectations have now declined about 1.2% over the past 3 months.

  33. Gravatar of dtoh dtoh
    24. December 2018 at 17:14

    When are you going to call out the Fed for complete incompetence.

    If professional economists don’t speak up things will never change.

  34. Gravatar of ssumner ssumner
    25. December 2018 at 09:42

    Dtoh, Are there any “professional economists” who’ve done to criticize Fed policy than me?

  35. Gravatar of dtoh dtoh
    25. December 2018 at 13:10

    No, but that doesn’t mean that you have done enough or that you have expressed your criticism in strong enough language.

  36. Gravatar of myb6 myb6
    4. January 2019 at 11:54

    Scott, I did some google-fu into the yield-curve causation issue and came back with “nobody knows”. Even the individuals in favor of non-causation were very cautious with their language. I’d personally add that none of that research addressed the novel situation of the CB being a major player in long-term markets.


    If we want to unwind eventually anyways, prices are convenient, and there’s (an unknowable) chance it could help macroeconomically, why not?

Leave a Reply