Jannah Theme License is not validated, Go to the theme options page to validate the license, You need a single license for each domain name.

Is the yield curve lying?

Good morning. FedEx, which reported earnings yesterday, is struggling to pass along cost inflation. Demand is limp, and it is “flexing costs to protect profitability”, says the chief executive. The shares fell 3 per cent after hours yesterday, following a languid day for stocks. A slow start to the new bull market, if that is indeed what we’re in. Email us: robert.armstrong@ft.com and ethan.wu@ft.com. 

Interrogating the yield curve 

Yesterday we wrote that “we do not like to bet against the three-month/10-year yield curve”. Thinking this over, it seems to us that this simple bet — with or against the recessionary signal sent by the rates curve — has been the defining issue in markets for the past 18 months. Nothing else matters so much, and every other consequential financial and macroeconomic debate is closely related. And because the facts relevant to the bet are changing, the debate is worth revisiting.

Start with the familiar chart or the curve over the past 40 years:

All five inversions in the past four decades predicted recessions, and there have been no false positives. As with every correlation, a compelling explanation is required if causation is to be inferred. And in this case, the explanation is very simple and powerful. 

A curve inversion is nothing else but the Federal Reserve increasing short- term interest rates above the economy’s neutral interest rate (the rate that neither stokes inflation nor increases unemployment). No one knows precisely what the neutral rate is — but long-term interest rates are something like a market approximation of it (adjusted for inflation). Short rates well above the neutral rate slow the economy. Indeed, they slow it until recession occurs, in part because the impact of rate policy on the economy is unpredictable and lagged, so policymakers tend to stay tight for too long. 

On this explanation, the depth of the inversion probably matters. Inasmuch as long rates track the neutral rate, a big gap with short rates indicates tighter policy. So it is worth noting that in 2023 the inversion has become historically deep, as the chart shows. 

Yet various market signals — stock prices, credit spreads, volatility — seem to be shrugging off the inversion. More surprisingly still, parts of the economy itself seem to be mocking the curve with shows of strength. This week’s big surge in housing starts is just the latest example. One has to ask whether this eminently sensible and super-reliable indicator could be wrong this time around.

There are two basic ways to call the yield curve a liar in mid-2023.

The first way is to argue that we are in an unusual economic cycle, in the following sense. There is an amount of tightening that is sufficient to bring down inflation; call that amount X. Then there is an amount of tightening that will force the economy into recession; call that Y. The idea is that, unlike past cycles, X is less than Y, and the Fed will stop at X.

So, why might this cycle be different? One view is that the current bout of inflation is the lagged result of a series of supply shocks that are working their way through the economy. Excess demand from stimulus is not an important contributor. Tightening policy, which is a tool for destroying demand, is therefore almost unnecessary. As such, X is less than Y. I think this is the view of, for example, my colleague Martin Sandbu. 

Another version of the “X is less than Y” story is that the labour market was extremely tight before inflation began, which means that the Fed can destroy quite a bit of demand before a lot of people get fired (and lots of people getting fired is what defines a recession, really). This is the kind of view that Chris Waller of the Federal Reserve took in his famous paper about job openings.

You can make an analogous argument substituting extraordinary household savings for job openings, and resilient consumer spending for resilient employment. All you need is for the savings to last until inflation passes and policy loosens.

Campbell Harvey, a professor at Duke and director of research at Research Affiliates, took the view that X was less than Y, and that the Fed would get the timing right, as recently as January of this year. His view then was that the job market was indeed extraordinarily strong and the financial system was in better shape than before past recessions, particularly 2008. He thought then that the curve was sending a false signal — which is interesting, in part, because he is generally credited with being first to demonstrate the connection between inversions and recessions, back in the 1980s.

Harvey thinks things are different now, though. He thinks inflation is coming down fast, and Fed tightening has exposed weaknesses in the financial system, as demonstrated by this spring’s bank failures. When I spoke to him yesterday he pointed to the “historic” depth of the inversion and argued that more hikes will put the economy in real danger. In short, he believes the curve now. 

The second way to call the yield curve a liar is to suggest that long rates have become hopelessly distorted — are too low, essentially — and therefore the curve no longer sends a reliable signal. It is common to blame the Fed’s bond-buying programs for the distortions. All you really need, though, is an argument that 10-year bonds are badly mispriced and that their yields should be higher (meaning the “real” yield curve is shallower than the apparent one). 

Michael Howell of CrossBorder Capital, for example, points to Treasuries’ disappearing term premium. The term premium is a measure of the difference between long rates and the expected path of short rates. It compensates investors for the risk that the rate outlook changes. The premium is currently negative (on the New York Fed’s estimate) and near an all-time low. Negative term compensation is weird, and Howell says it is evidence that long-dated Treasury prices are not driven by rate fundamentals, but instead by demand for long-term, risk-free securities for use as collateral. The curve is sending bad signals as a result.

Unhedged, on the other hand, thinks that the yield curve is telling the truth, and that a recession is on the way. With all the usual caveats about the folly of short-term economic prognostication, we take this view for three reasons:

  1. The sources of excess demand we currently enjoy, such as high consumer savings and a tight labour market, may well prove transitory. Many economic indicators look very strong right up until the moment a recession starts. In the terms used above, Y tends to be smaller than you think it is, when push comes to shove.

  2. The reason the Fed often over-tightens is that nailing the timing is very hard. Why would it be any easier this time around? 

  3. While we are sympathetic to the idea that 10-year yields are distorted, long-term rates have been coming down steadily for a long time basically everywhere. The neutral rate may be higher than 3.7 per cent (the current yield on the 10-year), but probably not loads higher. 

We think a recession is coming within the next year. With any luck it will be mild. 

Plenty of readers will disagree with both this characterisation of the yield curve and our faith in it. If you are one of them, send an email. 

One good argument

Is active management bunk? Click here for yes. Click here for no.

Read the full article here

Leave a Reply

Your email address will not be published. Required fields are marked *