Afraid of the yield curve? You’re look­ing at the wrong one

When bond yields flat­ten to cur­rent lev­els be­fore a re­ces­sion, the S&P of­ten posts gains

Waterloo Region Record - - Business - JAMES MACK­IN­TOSH

Share­hold­ers con­stantly live up to their rep­u­ta­tion as flighty ir­ra­tional be­ings prone to panic and wild aban­don. What is odd is that so many have come to view their coun­ter­parts in the bond mar­ket as stolid and sen­si­ble fore­cast­ers on whose views they can rely.

The ker­fuf­fle about the yield curve is a pri­mary ex­am­ple of the di­vi­sion. This week the gap be­tween long- and short-dated U.S. Trea­surys reached its low­est in more than a decade, with one ver­sion – the five-year mi­nus two-year yield – turn­ing neg­a­tive. Stocks have sold off hard, as in­vestors fear such so-called in­ver­sions of the yield curve presage re­ces­sions (ev­ery re­ces­sion since the 1950s was fore­shad­owed by an in­verted curve).

For in­vestors there are two im­por­tant ques­tions. First, is the yield curve re­ally telling us that re­ces­sion is loom­ing? Sec­ond, what should we do about it?

The case for be­ing scared of yield curves ap­pears strong in the U.S. be­cause they have sent few false re­ces­sion sig­nals and lots of cor­rect warn­ings. Still, it is un­clear why the curve should mat­ter, or which gap mat­ters most.

Un­til now no one paid at­ten­tion to the five mi­nus two, with mar­kets watch­ing the 10-year mi­nus 2-year yield and aca­demics tend­ing to fol­low the 10-year mi­nus the 3-month Trea­sury bill yield.

A model de­vel­oped by the New York Fed based on the 10-year mi­nus 3-month yield puts the re­ces­sion prob­a­bil­ity over the next year at roughly 15%, higher than in the past few years but still low.

The mar­ket’s fa­vored 10year-2-year spread isn’t flash­ing red, ei­ther.

With a gap of 0.11 per­cent­age points on Wed­nes­day, it was last at these lev­els and fall­ing in Novem­ber 2005 – two years be­fore the re­ces­sion, and a great time to own stocks, at least for a while.

That doesn’t mean the yield curve is telling us noth­ing. Longterm yields of­fer a rough guide to what in­vestors think will hap­pen to long-term growth and in­fla­tion. If short-term yields are higher, one in­ter­pre­ta­tion is the Fed’s about to make a mis­take by rais­ing in­ter­est rates too much and cause a re­ces­sion. An in­verted yield curve may also push lenders away from long-term loans to­ward more-prof­itable short-term debt, re­strict­ing ac­cess to fi­nance and im­pact­ing the econ­omy.

At the mo­ment the yield curve tells us that in­vestors think we are in the late stage of the eco­nomic cy­cle, and in­ter­est rates aren’t all that far from peak­ing. This isn’t a sur­prise, as al­most ev­ery­one agrees, in­clud­ing the Fed.

Some at the Fed and else­where ar­gue that this time is dif­fer­ent, be­cause long-term bond yields are sup­pressed by the Fed’s postcri­sis bond buy­ing pro­gram, known as quan­ti­ta­tive eas­ing, or QE. This doesn’t stack up as an ar­gu­ment, though.

It is true that when the ef­fects of QE are stripped out by us­ing NY Fed es­ti­mates of the so­called “term premium” to split off cer­tain sup­ply and de­mand fac­tors, the un­der­ly­ing long­dated bond yield would be higher and so the yield curve steeper. But it turns out that with­out the term premium, the yield curve of­fered no warn­ing of most re­ces­sions, and was in­verted for more than half of the past half­cen­tury. In other words, it doesn’t re­ally tell you any­thing use­ful about im­pend­ing re­ces­sions.

The in­vest­ment im­pli­ca­tions of a flat or even an in­verted curve aren’t ob­vi­ous.

The time to re­ces­sion from in­ver­sion in the past has var­ied from a few months to more than two years. Buy­ing the S&P 500 on the day of the first pre­re­ces­sion in­ver­sion of the 10-year-3month spread led to 12-month price re­turns vary­ing from a loss of 21% after Fe­bru­ary 1973 to a gain of 37% from Septem­ber 1998.

When the yield curve flat­tened to the cur­rent, still not in­verted, level of the 10-year-3-month spread be­fore each of the past seven re­ces­sions, the S&P went on to gain over the next year in five cases, and lost in only two. The curve was also this flat a few times where no in­ver­sion or re­ces­sion fol­lowed and stocks did well, such as in the mid-1990s.

Past ex­pe­ri­ence is no guar­an­tee of the fu­ture. And Japan’s ex­am­ple sug­gests that in a low­in­ter­est-rate world, re­ces­sions might hit with­out any warn­ing from the yield curve. Still, in­vestors shouldn’t panic purely on the ba­sis of what the yield curve is telling them.

That said, it makes sense, as it has for a while , grad­u­ally to shift to a more de­fen­sive stance as the eco­nomic cy­cle ma­tures. That might mean get­ting only part of fur­ther gains be­fore re­ces­sion even­tu­ally hits, but it also means be­ing bet­ter pre­pared when the big drop hap­pens.

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.