Economist who helped discover predictive powers of bond market says there’s no sign of recession right now

  • Arturo Estrella, a former economist at the New York Federal Reserve, says the yield curve between the 10-year note and the 3-month T-bill is not flat enough to predict a recession next year.
  • The economist co-authored several important research papers on the predictive powers of the yield curve.
An employee welds a specialized metal product in the shop at the Amuneal Manufacturing plant in Philadelphia, Pennsylvania.
Paul Taggart | Bloomberg | Getty Images
An employee welds a specialized metal product in the shop at the Amuneal Manufacturing plant in Philadelphia, Pennsylvania.

Call it "the unbearable flatness of yields" or "the trouble with the curve." However one cares to describe it, the slope of the Treasury yield curve has been the subject of much concern of late, as regards the future direction of the economy.

The yield curve — the relationship of short to long-term interest rates — can tell us much about the bond market's view of economic growth prospects.

During an economic recovery, the slope is positive, with long-term interest rates a couple percentage points above short-term rates. That's because bond market investors want more compensation, over the long haul, for lending money to government.

The reason for that is that as the economy accelerates, so too does inflation, which reduces the purchasing power of longer-term bonds. With more inflation investors demand higher yields as a means of protecting the value of their bond investments.

As the economy accelerates, the Federal Reserve, typically, begins raising interest rates to counter-act rising inflation, often to the point of tipping the economy into recession. At that juncture, long rates begin to fall as short-rates rise.

Bond investors, who are often quite prescient, fear the Fed will raise rates too much, not only killing off inflation, but also the economic growth that comes with it. Hence, the yield curve flattens and, ultimately, "inverts" to the point where long rates fall below short rates in anticipation of a recession and disinflation.

That's where we are today. The spread between 10-year Treasury notes and two-year notes, at 36 basis points, is the narrowest it has been since 2007.

Looking at wrong curve

A yield curve this flat has raised concern among economists, bond-watchers (myself included) that this unbearable flatness of yields portends a meaningful slowdown in growth, possibly even a recession, 9 months to 15 months from now.

However, I wanted to make sure I was viewing the yield curve through the proper lens, so I called Arturo Estrella, a former economist at the New York Federal Reserve, who co-authored several important research papers on the predictive powers of the curve.

Estrella agreed that the flattening of the yield curve, and, more specifically, an inverted curve, has accurately preceeded every recession the U.S. has experienced since 1968, without fail. The average time from a completely flat, or inverted, curve to a recession is 12 months … in this case, that would suggest a recession by next June.

The problem, he told me, is that economists today are looking at the wrong curve.

In his research, the curve that was most important involved the spread between the 10-year note and the 3-month T-bill. And while that curve has also flattened, today it's at 98 basis points; it is NOT flat enough to predict a recession next year. He said this is not a red alert … at least not yet.

Irrespective of all other concerns, and the behavior of all other markets, historically, the 10-year to 3-month spread was the single best indicator of a recession anywhere from 9 months to 15 months down the road.

The real recession risk

Estrella remains concerned, however, that, as has been typically the case, the Fed will tighten too much and tip that curve into recession territory sometime next year. That would still give us another year to prepare for a recession.

If the Fed were to hike rates another four times in this cycle, he believes the yield curve that he watches will flash a serious warning sign.

The main determinant of the spread is Fed policy, he says. Excess supply, other non-Fed factors are used to explain away the power of the curve, but remain excuses, not causes.

However, he warned that a trade war, if one does occur, might speed the economy toward recession, as well.

Estrella has been surprised that very few professional economists, including those at the Fed, rarely factor a flat-to-inverted yield curve into their future forecasts, despite its historical reliability.

He has been heartened of late, though, that incoming New York Fed president, John Williams, and James Bullard, president of the St. Louis Fed, of late, have been warning of the danger signs being flashed by flattening curves.

He pointed to a paper co-written by Williams in 2008 that warned about ignoring the curve.

If they hold sway in Fed deliberations, he told me, the Fed may not go too far in raising rates and ultimately causing a recession.

Estrella, who is a professor of economics at Rensselaer Polytechnic Institute, remains concerned, however, that this time will not be any different than prior cycles and that the curve will flatten to zero and eventually invert, predicting with 100 percent accuracy an impending recession.

He also warned me to make sure not just that we see the curve that is coming at us, but also to know which curve is the right one to take a swing at.