Recessions and Yield Curves

NOT déjà vu.

If we are going into another round of recession, we aren’t going the same way as before.

In the late 90s, I was part of a group of cyber-buds who debated markets, economic theory and policy in a stimulating endless argument.  A few of them (George Ure, Tom Drake, Rick Ackerman) have carried the tradition to this day, adding their voices to the tree frog chorus of economic thought on the web.

Just to be clear, I love tree frogs and their nighttime racket.

I have my Tree Frog Beer T-shirt from freshman year in college, much the worse for wear, but you can still recognize the all-knowing face of the Checkered Demon.  For those who know and appreciate the greatness of the Demon and S.Clay’s other characters, you might want to stop by and donate to help him with uncovered medical expenses.

But, I digress.  What I really came to do today was to talk about yield curves predicting recessions.

I would post a link to the long piece on yield curve inversion I did for George, Tom, Rick and other readers in the Colorado State University Longwave (Kondratyev) discussion group if I could. Turns out that stuff is so old that the web doesn’t seem to have any copies bouncing around any more.

So, let me start with the definition of an inverted yield curve that can actually predict a recession.

In later posts I’ll answer the most frequent objection — that inverted yield curves make false recession predictions,or that other indicators are better predictors.

I’ll also ask the musical question:  Has Big Ben and his QE’s disrupted the predictive role our yield curve plays?  I’ll admit in advance that I don’t know the answer to that question.

Yet another post (or even a book), could lay to rest some of the True Believer myths about gold and recessions, inflation and recessions, and government policy and recessions.

To start at the beginning, an inverted yield curve is a condition that Baby Boomers think is unusual.  If we lived to be 150 or 200 years old, then there would be plenty of us around who knew that inverted curves are just about as common as upwardly sloping curves.

A yield curve is called ‘inverted” when a shorter maturity credit from a given borrower is a higher yield than a longer maturity credit from that same borrower.  The most common reference credit is Treasuries.  Some people define the steepness of the curve as the difference between the two year Treasury and the ten year.  We call that the 2’s-to-10’s, and today, that difference is 2.56% – 0.29%, or 227 basis points.  That’s a pretty steep curve.  An inverted curve exists when the “steepness” is negative.

I prefer to use the 3-month T-bill and the 10-year T-note as my measure of the steepness or inversion of the Treasury curve.  With today’s T-bill close of 0.01% (yes, that is ONE basis point!) and the Note trading at 2.56%, my measure of steepness is +255 basis points.

My reasoning for that choice is that the 3 month bill is the basis that we use to measure short term bank credit in the form of Eurodollar contracts (AKA 3-month LIBOR). That makes 3-months the most common term for “short” money all over the world, and the basis for every swapped corporate bond out there.

The 10-year is the most common term before refinance for commercial real estate, and the most common term for investment grade corporate issuers.  We could also look at the five-year point on the curve, since that’s today’s most common tenor for Credit Default Swaps.  Since the CDS market is still basically opaque even if it is huge, I can’t yet consider it important enough to drive the basic predictive power of the market.

A little “eyeball” estimating of curve inversions that preceded prior recessions tells me I need to see 50 basis points or more inversion to get a valid signal.  I need to add two more conditions to make the inverted curve predictive without false positives.

The first additional condition is that the BBB (investment grade) corporate bond yield curve remains positively sloped.  The second condition is that the spread of those BBB bonds is widening at the time the Treasury curve inversion appears.

When I see those conditions, I feel confident predicting a recession beginning roughly six months later.

“Why the confidence?”, you might ask.

First, because the inversion itself is a sign the market believes that short rates will decline significantly in the relatively near future.  The most frequent reason this happens is that the Fed Open Market Committee lowers their target for Fed Funds to stimulate growth.  We’re all Pavlov’s dogs to Alan Greenspan’s rate cuts, so we salivate over coming cheap money, knowing that recessions result in Fed easing.

It’s those other conditions that make me confident in the prediction.  If the market knows the risk of performance for investment grade corporates requires more yield spread over longer maturities than short, and more yield spread in the market week by week (at the same time the Treasuries are moving into an inversion), then the market is telling me that the corporates are under stress, and that the odds of default are rising even while Fed policy is lowering rates.  Ergo, cyclical recession is coming.

By adding those extra conditions, I eliminate the conditions that create inverted yield curves for corporates and Treasuries at the same time.  That keeps me from predicting recessions frequently during a secular bull market in bonds, which can go on for several business cycles.

[Note for those who missed the bond market shorthand above:  A secular bull market in bonds is an extended period during which interest rates steadily fall to lower and lower levels, subject to ordinary market movements up and down, just like a secular bear market in stocks is a long period when stock prices keep moving down in spite of counter-trend rallies.]

[Second, more succinct note for those who live outside the bond world:  When yields go up, bond prices go down.  When yields go down, bond prices go up.  One of my best friends from outside the financial world reminded me that I should repeat this mantra every so often for the non-mavens, because it feels so topsy-turvy to most people.]

When I used these conditions to predict the shallow recession of 2001 the for the CSU Longwaves group in late 2000, those particular yield curve conditions had not existed for a decade. By the way, I predicted a shallow recession because the inversion and corporate spread widening were both relatively small (on the order of 50 basis points).

The fascinating part of that prediction was that so many of us in that group had been so sure the good times would end many times before, but the good times kept on rolling.

The 2001 recession didn’t come until ten years of continuous economic expansion had passed, and it was preceded by the yield curve inversion for Treasuries accompanied by the steepener/widener in the corporate yield curve.

When you listen carefully to what the Treasury and corporate bond yield curves are telling you, you can hear the recession train coming a couple of quarters before it arrives.

hh