How many times?

Here’s Tim Duy at Bloomberg:

It was supposed to be easy. When the Federal Reserve started hiking the federal funds rate, longer-term interest rates would rise. After all, they were at very low levels, restrained by a low-term premium. The “Greenspan conundrum” of the past two cycles, when long rates failed to respond in line with higher short rates, couldn’t happen a third time in such circumstances. But it didn’t work out that way. Short rates continue to gain on firming expectations of tighter Fed policy while long-rates stubbornly track sideways.

How many times does this have to happen before people stop assuming that higher interest rates represent tighter money?  In fact (as Duy suggests) a tighter monetary policy will often put downward pressure on longer term rates (relative to short rates):

We shouldn’t be surprised by the flattening yield curve. That is what typically happens during tightening cycles and there was no reason to think it would not be the case this time.

Longer term bond yields tend to track expected long-term NGDP growth, although of course other factors also play a role.  But expected NGDP growth is by far the most important factor, and largely explains why long-term rates are much lower than during 1972-81, when NGDP was growing at double digit rates.  And tighter money tends to slow expected NGDP growth.

The yield curve is one of the better predictors of the business cycle, but it’s not perfect.  The current yield curve is flatter than usual, but not flat enough to predict a recession.  (Research suggests that it would need to be substantially inverted to signal a recession is more than 50-50.) Given that stocks are doing quite well, I’d say that the consensus view of the financial markets is that a recession is unlikely during the next few years, but not impossible.

Duy also points to the continued undershoot of inflation:

Arguably, though, the Fed only reinforces expectations that 2 percent is a ceiling with its commitment to rate hikes even as inflation remains below that level. In fact, monetary policy makers appear dead set to continue rate hikes next month, and into 2018. The message sent is that they stand ready to snuff out any expansion that threatens to push inflation above 2 percent.

It’s difficult to evaluate Fed policy in isolation; one needs to consider the regime over an entire business cycle.  Thus the current sub-2% inflation rate is entirely consistent with the dual mandate, assuming that inflation is appropriately countercyclical (which is what the mandate implies.)  But in the past inflation has tended to be procyclical (in violation of the dual mandate), in which case current policy is inappropriately tight.

So how do we know if the Fed policy today is appropriate?  We don’t know, and won’t know until the next recession.  If inflation rises during the next recession, then current policy will have been appropriate—even though inflation is now under 2%.  If inflation falls during the next recession then current policy will have been inappropriately tight. Based on past experience, the latter assumption is more plausible, but again, current policy is exactly right if the Fed were taking its dual mandate seriously.  That mandate calls for slightly above 2% inflation during periods of high unemployment, and slightly below 2% inflation during periods of low unemployment.  And right now unemployment is well below average.  The dual mandate calls for sub-2% inflation at this point in time.


Tags:

 
 
 

27 Responses to “How many times?”

  1. Gravatar of Brian Donohue Brian Donohue
    21. November 2017 at 08:57

    Isn’t the next logical tightening step to start shedding some longer-dated balance sheet assets rather than another increase in the FFR? Also, when do we stop subsidizing excess reserves? Can’t we do both?

  2. Gravatar of Ironman Ironman
    21. November 2017 at 08:58

    “Research suggests that it would need to be substantially inverted to signal a recession is more than 50-50.”

    Not necessarily – some of the better yield-curve-as-recession-predictor research that we’ve seen over the years also factors in the level of the Federal Funds Rate, where if that rate is high, it can lead to an indication of a >50% recession probability, even with a positive spread between long and short term treasuries in the yield curve.

    Back in 2007, we developed a chart to make it easier to see how the treasury yield curve and FFR interact to indicate a given probability of recession. (Last updated on 9 November 2017).

  3. Gravatar of Russ Abbott Russ Abbott
    21. November 2017 at 09:15

    I’ve always been bothered by discussions that talk about recessions as if they were acts of nature, like earthquakes. Presumably recessions are the result of some economic problem, like inventory build-up. Why don’t we talk about the likely causes of the next recession rather than just when it might occur?

  4. Gravatar of Kevin Erdmann Kevin Erdmann
    21. November 2017 at 10:05

    Ironman,

    I am confused about your chart. It looks like if the Fed Funds rate was 1% and the 10 year rate was 0%, it would still only predict a 15% recession probability. How am I reading it wrong?

  5. Gravatar of ssumner ssumner
    21. November 2017 at 12:15

    Brian, I agree.

    Ironman, The simpler the better. Adding more variables reduces statistical significance.

    Russ, The cause of most recessions is tight money.

  6. Gravatar of B Cole B Cole
    21. November 2017 at 16:15

    Conventional macroeconomists say unemployment is low.

    On the other hand there are about 1.2 to 1.3 people looking for a job for every job opening in the United States, presently. The Fed is in squeamish hysterics over labor markets.

    Yet unit labor costs are falling. People appear presently to be drawn into the labor force and labor force participation rates are rising.

    Has the Fed been fighting the wrong source of inflation? I think so.

    Is most inflation today because of tightly zoned housing markets and not by “tight” labor markets?

    Well, check house prices against wages last 10 years.

  7. Gravatar of Viking Viking
    21. November 2017 at 17:14

    “Longer term bond yields tend to track expected long-term NGDP growth, although of course other factors also play a role. ”

    Should long term yields track total NGDP growth, or per capita NGDP growth?

    In other words, let us assume 2 scenarios, each with 3% NGDP growth:

    1. Population grows 3% per year, and total NGDP only matches population growth, so GDP per capita stays constant.
    2. Population is constant, total NGDP and NGDP/capita grow 3%.

    Should we expect the second scenario has higher long term rates?

  8. Gravatar of Cameron Cameron
    21. November 2017 at 17:46

    Scott,

    If you knew the US was headed for recession next year, how well do you think the current fed would handle it? Are we better off because of what they’ve learned, or are we worse off because we’re starting at a lower funds rate and trend NGDP growth?

  9. Gravatar of ssumner ssumner
    21. November 2017 at 18:20

    Viking, I’ve thought a lot about that question and don’t really have an answer. I believe that both factors influence interest rates, but per capita GDP is probably more important.

    Cameron, Good question, with no easy answer. I’d start with the fact that if we were headed for recession that would be a sign that the market has no faith in the current Fed, indicating that problem of low rates is worse than the benefit of what they have learned. One can think of a big drop in AD as a sort of “prediction of central bank incompetence”.

    On the other hand, I don’t expect a recession, which is a good sign I suppose.

  10. Gravatar of HL HL
    21. November 2017 at 22:35

    What is happening in Turkey (Erdogan suddenly attacking the central bank for keeping interest rates and inflation too high) might amuse you a bit, especially because you once asked the rhetorical question (my rough memory), “what would happen if Turkey wants to keep rates at zero due to religious reasons”? 10 year bond yields spiking whereas it is falling everywhere else.

  11. Gravatar of Alec Fahrin Alec Fahrin
    22. November 2017 at 06:37

    Scott,
    May you explain these sentences.

    “That mandate calls for slightly above 2% inflation during periods of high unemployment, and slightly below 2% inflation during periods of low unemployment. And right now unemployment is well below average. The dual mandate calls for sub-2% inflation at this point in time.”

    It makes sense in a way, yet why then does the entire Fed board seem perplexed that inflation is lower than 2% and has been since 2012. I keep hearing the Philip’s Curve argument from them.

  12. Gravatar of Neil Neil
    22. November 2017 at 08:45

    I guess what is surprising is that despite the acceleration in nominal GDP growth this year, ten year yields have not budged.

  13. Gravatar of ssumner ssumner
    22. November 2017 at 09:09

    HL. Yes, the world’s most famous NeoFisherian. :)

    Alec, That’s because the mandate also includes employment, so you want policy to be a bit more expansionary when employment is low, and vice versa.

    Yes, the Fed’s big mistake is relying on the (highly unreliable) Phillips curve model.

    There’s no mystery to low inflation (it’s slow NGDP growth, obviously.)

  14. Gravatar of ssumner ssumner
    22. November 2017 at 09:14

    Neil, Yes, but what matters is expected future NGDP growth, which remains stuck at close to 4%, as it’s been for 8 years.

  15. Gravatar of Mark Mark
    22. November 2017 at 11:01

    When one says the yield curve predicts a recession with greater than 50% probability, what timeframe is assumed? Within the next year? And relatedly, what’s the consensus (if there is one) on what the approximate period is for the business cycle?

  16. Gravatar of Ironman Ironman
    22. November 2017 at 11:59

    Kevin Erdmann asked:

    I am confused about your chart. It looks like if the Fed Funds rate was 1% and the 10 year rate was 0%, it would still only predict a 15% recession probability. How am I reading it wrong?

    The vertical axis is showing the spread between the 10 Year and 3-Month constant maturity treasuries – not the level of either.

    What the chart is communicating is that according to Wright’s recession forecasting model, the NBER would be unlikely to declare that a national recession will have begun in the U.S. within 12 months of the date of observation if the FFR were averaging 1% and the spread between the 10-Year and 3-Month treasuries were 0% (or rather, when the 10-Year and 3-Month treasuries have identical yields), where the odds of it ever doing so are just 15%.

  17. Gravatar of Geoff Orwell Geoff Orwell
    22. November 2017 at 14:43

    Buried in the fed minutes today, is this a significant development?

    “In view of the persistent shortfall of inflation from the Committee’s 2 percent objective and questions about whether longer-term inflation expectations were consistent with achievement of that objective, a couple of participants discussed the possibility that potential alternative frameworks for the conduct of monetary policy could be helpful in fulfilling the Committee’s statutory mandate. One question, for example, was whether a framework that generally sought to keep the price level close to a gradually rising path—rather than the current approach in which the Committee does not seek to make up for past deviations of inflation from the 2 percent goal—might be more effective in fostering the Commit-tee’s objectives if the neutral level of the federal funds rate remains low.”

  18. Gravatar of Geoff Orwell Geoff Orwell
    22. November 2017 at 14:46

    and also, replying to Alex above, presumably the Fed are still determined to hike despite falling market based IE and flattening curves (falling inflation compensation, as they call it) BECAUSE they see it as normal that realized inflation is low during the expansion/high employment, which fits with Scott’s framework.

  19. Gravatar of Geoff Orwell Geoff Orwell
    22. November 2017 at 14:47

    and also, replying to Alec* above, presumably the Fed are still determined to hike despite falling market based IE and flattening curves (falling inflation compensation, as they call it) BECAUSE they see it as normal that realized inflation is low during the expansion/high employment, which fits with Scott’s framework.

  20. Gravatar of Pyrmonter Pyrmonter
    22. November 2017 at 14:57

    ‘Longer term bond yields tend to track expected long-term NGDP growth, although of course other factors also play a role’

    How is that different to saying they track expectations of the price level? Why would a bond-holder be interested in the return on the bond measured as a share of output, rather than its real value?

  21. Gravatar of flow5 flow5
    22. November 2017 at 15:08

    “We don’t know, and won’t know until the next recession.”

    Wrong. We already know. Nothing’s changed in over 100 years. In contradistinction to Janet Yellen’s “Inflation Shortfall Is a ‘Mystery’ ”, there is no such conundrum. Janet Yellen’s rate hikes were prescient.

    Monetary policy objectives should be formulated in terms of desired rates-of-change, RoC’s, in monetary flows, M*Vt (volume X’s velocity), relative to RoC’s in R-gDp.

    parse: dt; real-output; inflation:

    1/1/2017 ,,,,, 0.13 ,,,,, 0.19
    2/1/2017 ,,,,, 0.08 ,,,,, 0.16
    3/1/2017 ,,,,, 0.06 ,,,,, 0.13
    4/1/2017 ,,,,, 0.08 ,,,,, 0.18
    5/1/2017 ,,,,, 0.09 ,,,,, 0.23
    6/1/2017 ,,,,, 0.08 ,,,,, 0.21
    7/1/2017 ,,,,, 0.11 ,,,,, 0.20
    8/1/2017 ,,,,, 0.09 ,,,,, 0.23
    9/1/2017 ,,,,, 0.08 ,,,,, 0.25
    10/1/2017 ,,,,, 0.03 ,,,,, 0.23
    11/1/2017 ,,,,, 0.08 ,,,,, 0.24
    12/1/2017 ,,,,, 0.10 ,,,,, 0.16
    1/1/2018 ,,,,, 0.06 ,,,,, 0.22
    2/1/2018 ,,,,, 0.04 ,,,,, 0.23
    3/1/2018 ,,,,, 0.03 ,,,,, 0.19
    4/1/2018 ,,,,, 0.01 ,,,,, 0.16
    5/1/2018 ,,,,, 0.02 ,,,,, 0.16
    6/1/2018 ,,,,, 0.01 ,,,,, 0.14

    The trajectory after the holidays is significantly lower. First real-output falls in the first qtr, second inflation falls in the first half.

    With “sticky deposit rates”, Vi, which has risen during the 2nd, 3rd, and 4th qtrs. of 2017, will reverse and fall going into 2018.

  22. Gravatar of Scott Sumner Scott Sumner
    23. November 2017 at 07:36

    Mark, One year.

    Thanks Geoff. Interesting.

    Pyrmonter, You asked:

    “Why would a bond-holder be interested in the return on the bond measured as a share of output, rather than its real value?”

    Because you are comparing it with alternative investments whose rate of return is correlated with NGDP growth.

  23. Gravatar of msgkings msgkings
    23. November 2017 at 17:30

    @flow5:

    These posts of yours have been interesting to me because of the flat mathematical certainty you project, so I’ve been watching. A year ago (or so) you said the real output drop would start in summer 2017. So are you one of those permabears who just keeps crying ‘recession’ until you are right?

  24. Gravatar of Benjamin Cole Benjamin Cole
    24. November 2017 at 03:09

    Hey, the Japan Times and Tokyo stock brokers get it right:

    “The FOMC minutes were taken as a hint that U.S. interest rate hikes would continue at a slow pace next year, pushing down long-term rates and inducing dollar sales, brokers said.”

    https://www.japantimes.co.jp/news/2017/11/24/business/financial-markets/nikkei-ekes-gains-backed-buying-dips/#.Whf9KyN95jc

  25. Gravatar of flow5 flow5
    29. November 2017 at 09:29

    @ msgkings:

    Fuck you. You’re a deliberate liar. Those #’s have always indicated that the economy would rebound beginning in the summer. Given the fact that you don’t know what those #s mean, you are interpreting them to suit your own wishes (that I would be wrong).

  26. Gravatar of flow5 flow5
    29. November 2017 at 09:42

    And you can remind me that I got money velocity right too.

  27. Gravatar of flow5 flow5
    30. November 2017 at 10:14

    The general level of prices is determined by (1) the quantity of money; (2) the velocity of money; and (3) the physical volume of trade.

    Money velocity, Vt (the transactions velocity of circulation), based on historical variabilities, has been 3 times as important a factor in determining aggregate monetary purchasing power, AD (or money X’s velocity), vs. that of our means-of-payment money. This mathematical weight is not reflected by income velocity, Vi, a contrived metric (which sometimes moves in the opposite direction as Vt). Vt is a much higher figure, encompassing, e.g., intermediate goods. Thus, even a minute increase in Vt translates into a large increase in AD.

    Of the 3 figures on income velocity published in the FRED database by the Federal Reserve Bank of St. Louis, the path of MZM velocity is the closest to representing the previous trajectory, its relative direction, of transactions velocity (debits to deposit accounts).

    http://bit.ly/1A9bYH1

    American Yale Professor Irving Fisher’s “equation of exchange” can be used as an analytical tool, even though some the truism’s variables are unsolvable [i.e., “P” is not a price index, it is not a relative, but rather an absolute term – and cannot be calculated because of the item “T”. It is obviously impossible to add up quantities of things: – pounds of sugar, yards of cloth, and hours of work, cannot be combined except by reducing these elements to a common denominator]. M*Vt, can nevertheless give us valuable insights into economic horizons.

    The importance of Vt in formulating or appraising monetary policy derives from the obvious fact that it is not the volume of money which determines prices and inflation rates, but rather the volume of money flows, relative to the volume of goods and services offered in exchange. A dollar bill that turns over five times can do the same “work” as a five dollar bill which turns over only once.

    For our means-of-payment money to have effective purchasing power it must be used; it must turnover, i.e., pass from buyer to seller. The equation of exchange further reveals that prices will not necessarily rise even if total effective purchasing power increases. An increase in the numerator of the equation could, and sometimes is, offset by an increase in the denominator. It is possible for an increase in purchasing power to be offset by an increase in the total output of goods and services, or in the rapidity with which goods turns over. The equation of exchange is adaptable to both a long, and a short-run approach.

    However, there is no evidence that the FOMC takes either Vt, or Vi, into account in formulating or executing monetary policy. And the FOMC is no longer required by the Humphrey-Hawkins Act, c. 2000, to set target ranges for any money metric.

    The scope of this irresponsible decision is demonstrated by the fact that virtually nothing on the Federal Reserve Bank of New York’s website on “The Money Supply” is palpable.

    https://www.newyorkfed.org/aboutthefed/fedpoint/fed49.html

    To wit: “Thus, in July 1993, when the economy had been growing for more than two years, Fed Chairman Alan Greenspan remarked in Congressional testimony that “if the historical relationships between M2 and nominal income had remained intact, the behavior of M2 *in recent years* would have been consistent with an economy in severe contraction.”

    Chairman Greenspan added, “The historical relationships between money and income, and between money and the price level have largely broken down, depriving the aggregates of much of their usefulness as guides to policy. At least for the time being, M2 has been downgraded as a reliable indicator of financial conditions in the economy, and no single variable has yet been identified to take its place.”
    Not so. Money flows, AD, rebounded from a negative figure:

    http://monetaryflows.blogspot.com/

    In fact, such conclusions were literally inane: “Interest rates were at the lowest levels in more than three decades, prompting some savers to move funds out of the savings and time deposits that are part of M2 into stock and bond mutual funds, which are not included in any of the money supply measures.”

    As any competent economist knows (those who can read the tea leaves), funds transferred through non-bank conduits never leave the payment’s system.

    2017 was a very interesting year. Contrary to Janet Yellen: “[t]he biggest surprise in the US economy this year has been inflation”, no, contrary forces were at work, raising the remuneration rate 3 times destroyed money velocity, and the 2 year rate-of-change in money flows acted in the opposite direction, raising money velocity. 2018 will most likely reflect the opposite situation, Vt will decline in conjunction with the drop in inflation and the jacking up of sticky deposit rates.

Leave a Reply