Why I’m not paying too much attention to the flattening yield curve.

July 9th, 2018 at 3:17 pm

As Nick Timiraos ably describes, there’s a debate afoot about how seriously to take the flattening and possible future inversion of the yield curve. I got into this a bit last week, pointing out that the signal from the yield curve is a lot more ambiguous than usual (my conclusion was that we should worry a lot more about how we’re going to offset the next recession versus when it’s coming, which is not reliably knowable).

One reason for this ambiguity is the very low term premium on long-term bond yields (see figure). Longer-term interest rates, like the yield on the 10-year Treasury, can be broken up into the expectation of the average of future short-term rates and the term premium, or the extra yield investors require to lock up their money for the term of the loan. Since it’s thought to be the first part — expected rates — that correlates with future downturns, it makes sense to net out the term premium from the model. As the next figure shows, that significantly lowers the curve recession probability (see these Fed papers for details). As economist David Mericle recently put it, low term premia imply that “an inversion…no longer signals that current interest rates are nearly as far above expected average future short rates as in the past.”

Source: Mericle, GS


Sources: Fed, NBER

But for this quick post, I want to get a bit more into what an inversion might mean for Fed policy. Would recessionary signals from an inversion lead them to pause in their rate hike campaign?

I doubt it, and agree with Jan Hatzius: “…yield curve inversion does not cause recession, but is merely indicative of the types of conditions (i.e. overheating) that are often followed by recession. This sounds like a technical distinction but implies, crucially, that the Fed cannot lower the risk of recession simply by refusing to deliver hikes that would invert the curve. This would worsen the overheating and could ultimately lead to an even bigger inversion and an even higher risk of recession.”

We can and should have good debates about the extent of overheating in the current economy, about which I’m pretty dovish. Yes, inflation’s up a bit, but a) it should be at this point in the expansion, b) core PCE just hit 2%–the Fed’s target—after being below target for years, c) most importantly, inflation expectations remain as well-anchored as ever.

So, based on extraneous factors flattening the curve and the low likelihood that an inversion may not much alter the Fed’s normalization plans, I’d pretty heavily discount the yield curve. Of course, that doesn’t mean a recession isn’t out there somewhere—it is. We still don’t know where, but unless it’s unusually mild, we can be pretty confident that neither monetary nor especially fiscal policy will be poised to do enough to offset it.

Print Friendly, PDF & Email

2 comments in reply to "Why I’m not paying too much attention to the flattening yield curve."

  1. Fred Donaldson says:

    With the European countries’ reluctant to raise interest rates, declining cash in consumer pockets due to privatize-motivated healthcare hikes, and the impact of tariffs until manufacturing rebounds, only a reduction in the labor force by Visa controls for middle class workers can maintain wages and consumer confidence, let alone lead to excessive inflation. We are faced with artificial pricing (eye glasses that cost $7 to produce overseas and sell for $200 here, for example) and a tax system that restricts net earnings for seniors as caps like the $32k “test” are not COLA adjusted.for the past 30+ years. As seniors with defined pensions are replaced by those with low-yield 401ks, the economy’s dependence on retired spenders will diminish, not to mention the negatives of housing price hikes that discourage purchasing by the young, who also face rising mortgage interest rates. The Fed might best “cool it” by doing nothing, or even backtracking its quest for more return on stalled cash.

  2. dwb says:

    “an inversion…no longer signals that current interest rates are nearly as far above expected average future short rates as in the past”

    Sounds to me exactly like the confirmation bias I heard right before housing crash. Nope this time is different, right. “This time is different” will lead the Fed to raise rates too far, a recession, then economists who denied the inverted yield curve will deny it was cause by raising rate too far.

    As I read Mericle’s Figure three, there is roughly a 20% change of a recession.

    I think that is about right: There is a good chance that the curve inverts later this year or Q1,2019… and the recession would ensue 12-18 months after that. So I would not expect a model that only looks 4 quarters ahead to pick up a significant probability. Monetary policy acts with a significant lag, so by the time a high recession probability shows up in these 4-quarter ahead models, its too late!

    The probability of a recession in the next 24-26 months is closer to 50%, but mainly due to Fed policy error and peevish focus on inflation. Same as it ever was.