FALSE START! Inverted Yield Curve has not Predicted a Recession

For the past few months there have been swarms of PhDs, economists, and strategists buzzing to a hypersensitive media about the dire straits portended from the inverted yield curve. It only seems appropriate we share more of our thoughts and observations on the matter.

As a primer, a bond is a loan that is to be repaid at prespecified date when the bond/loan matures. The borrower/bond issuer pays interest to the lender/bond holder as compensation. For the purposes of this general essay, the terms interest rate and yield are used interchangeably, and inflation is assumed to remain subdued.

What is the yield curve?

[endif]--The yield curve is a chart of the interest rates that are paid to borrow money for various amounts of time. Normally, a borrower pays a higher interest rate for a longer maturity. The general shape might look like this.

An inverted yield curve occurs when the yield on a longer-term bond is less than a shorter-term bond. The general shape might look like this.

The buzz today is about how the yield on a 10-year US Treasury (UST) was lower than a 2-year UST and even the 3-month UST. As we will demonstrate, the nature of the borrower determines the implications of the shape of the yield curve, and we will discuss the yield curves of four types of borrowers: The Starving Author, Mediocre-or-worse Companies, High Quality Companies, and the US Treasury.

In general terms, when the borrowers are individuals and companies, the interest rate is a measured reflection of the chance of the bond holder being repaid, either because enough cash was generated, the borrower was able to refinance, or otherwise there was sufficient collateral.

The Starving Author

Just to be clear, there is no market for starving author loans and our thoughts are only conceptual. Imagine that a bright, creative, passionate, and determined person comes to you and says they have an important book to write. They have no money and need a $37,000 two-year loan to cover $1000/month for an apartment, $100/week for food and cigarettes, and $2600 for a computer and printing supplies. If you extend this two-year loan to the starving author, there is little chance that cash from book sales will happen fast enough to pay back the loan, so the bet is that the author will attract attention and someone (else) will want to finance the author’s future when the two years is up. Obviously, loans to starving authors are very risky and should bear a high interest rate. However, the author’s risk profile would dramatically improve with a successful book, plus a career change may increase the chance that the loan would be repaid. So, it may arguably be less risky to lend the starving author $37,000 for ten years than for two years, and the interest rate would be lower for a 10 vs a 2-year loan. Therefore, the hypothetical yield curve of the starving author seems naturally inverted to us.

Mediocre-or-worse Companies

When the economy is growing, the results from mediocre companies tend to be sufficient and the rising economic tide lifts all boats. Bond holders would be less concerned about being repaid in the coming year or two than in the next 10 years, so the interest rates are only a bit higher than for high quality companies in the short run but the differential increases as the length of the loan increases. When a recession is on the horizon, the interest rates on bonds of mediocre-or-worse companies can skyrocket, reflecting the diminishing chances of the loan being repaid in full. Indeed, an oft cited statistic is the spread between high yield bonds and USTs, also known as the high yield spread. A spike in the high yield spread is seen as a decent predictor of recessions, though in our experience the spread widening is a bit more of a concurrent indicator than a leading indicator of a recession. Further, a recession does not always accompany a wide high yield spread, such as in 2011. Like the starving author, the interest rates mediocre-or-worse companies pay is overwhelmingly driven by the odds of the company generating enough cash to repay the bond or otherwise being able to refinance it. These are largely company specific derived interest rates, and the yield curve is upward sloping in good times and steeply upwardly sloping in a recession.

High Quality Companies

Interest rates for high quality companies are a blend of company specific evaluations and the overall economic environment. Company specific evaluations are generally only negative developments that may cause the company to become mediocre-or-worse. So, the majority of the interest rate setting mechanism involves an assessment of the economic environment for the company’s business sector. The changes in interest rates paid by high quality companies will generally mirror the changes in the risk-free rate of US Treasuries.

US Treasury

Ask US financial professionals and they will tell you the only investment that can legally be guaranteed are the bonds issued by the US Treasury. Despite what you may perceive, the Unites States and its government are very stable!!! Worldwide, the United States is generally presumed not to have credit risk and USTs are considered risk free assets. So, considerations that impact UST interest rates principally involve economic growth and inflation, and for this overly general essay we will assume inflation remains subdued. When economic growth is expected, the farther one looks out the greater the effects of compound growth, and risk-free interest rates increase for each subsequent year. The yield curve has an upward slope to it, which is normal in the United States.

However, if the steady growth expectations are interrupted by expectations for a recession, several things are expected to happen with varying effects. First, risk free interest rates are expected to be reduced during a recession to help incentivize activity and spur a resumption of growth. Financial speculators and economists look at the 10-year UST as a series of 10 sequential 1-year US Treasuries. Because of a recession, the average (it is not technically an average) expected rate for the next ten individual years can be lower than for a 2-year UST, so the 10-year UST yield may be lower than the 2-year UST, in which case the yield curve is inverted. Secondly, investors may sell risk assets that were originally intended to be held for the long term and run for safety of a risk free 10-year or other long dated UST. This is a market driven force that can be materially accentuated by the increased demand of financial speculators, foreign government and foreign investors. Therefore, inverted UST yield curves are appropriately associated with recessions and that is why there has been so much attention paid to the topic.

All this is well and good, but it does not give any indication as to what to do or when. So, we will take a look back and explore the history. A tremendous thank you goes to strategists of Deutsche Bank (DB) who wrote a very detailed report titled Yield Curve 101 published December 11, 2018.

In 1965 Reuben Kessel, a professor at the University of Chicago’s Graduate School of Business published a paper on the cyclical behavior of the interest rate curve1. But Kessel did not note that an inverted yield curve was a leading indicator. Computing power started a transformative revolution in the early 1970s and by the mid-1980s dazzling computations and analysis were available. In 1986 Harvey Campbell, a professor at Duke University’s Fuqua School of Business, furthered the work of Kessel and wrote a thesis highlighting the inversion of the 2-year UST versus the 10-year UST as a leading indicator of economic recessions2. Recently, on the heels of the day when the UST yield curve last inverted, in a March 22, 2019 Barron’s article3 Harvey Campbell was quoted as saying “I have been analyzing the yield curve for more than 30 years. Importantly, my model argues that a yield curve inversion must be realized for a full quarter—not merely a few days.”

The DB team’s report was a nerdy deep dive and further dissected UST relationships of the 2-year vs the 10-year, 3-year vs the 5-year, 3-month vs the 10-year, and the 3-month vs the 5-year. The report confirmed that the 2/10 was indeed the best pair for predicting recessions, in our opinion.

The most actionable thing that jumped off the pages of the DB report involved what we would call false starts. In this case, a false start occurs when the yield curve briefly inverts, and time passes before the yield curve inverts and stays inverted for a full quarter. The day the yield curve first inverts does not mark the start of the recessionary signal.

The recessions starting from 1960 to 1990 had only 1 false start. Let’s face it, the business cycle was different back then. In those days, inventory levels played a much larger part in driving recessions. Improving business conditions lead to overly optimistic production of stuff, there was a recession and workers were laid off until the excess inventory was sold, and eventually they were rehired to make stuff again. It seems that booms and busts were far easier to predict back then.

We have written about this before and are happy to reshare our thoughts upon request, but as the economy shifted from and industrial production economy to the service-based economy we have now, cycles of the past either morphed or their cyclicality ceased. The last two recessions started in 2001 and 2008 and both had significant false starts. According to the DB report, the 1998 false start occurred almost three years before the 2001 recession, though the true inversion signal occurred 13 months before the recession. The 2005 false start occurred two years before the 2008 recession and the true inversion signal occurred 18 months before the recession.

The bottom line for all this highly technical yield curve mumbo jumbo is that one should expect a recession to start before 2024. But did you really need hypersensitive and overly dramatic media coverage of sound bight filled swarming intellectuals to tell you that?! It is plain to see that recessions come around every so often, the precipitating events are often unpredictable, and we are more or less due for a recession in the coming years. We continue to be on alert!

And to put first things first and highlight the actionable proper perspective, in our opinion, the table below is the approximate level of the S&P 500 and Dow Jones Industrial Average at the time of each first inversions over the last thirty years.

The two major cautionary takeaways from this table are:

1) There is exponentially greater opportunity cost to not being invested in equities, than temporary bear market downside! The chance for a tenfold increase in value is well worth enduring a few bear markets, in our opinion.

2) You must be able to endure the bear markets! The shorter your time horizon, the more important it is to be able to live off dividends and interest and have a rainy-day cash fund.

We would love to meet for a discussion about your asset allocation and to revisit your long-term financial plan. Please call us!

1. https://www.nber.org/chapters/c1662

2. https://faculty.fuqua.duke.edu/~charvey/Term_structure/Harvey.pdf

3. https://www.barrons.com/articles/what-the-yield-curve-inversion-really-means-campbell-harvey-51553283550

Redmond Asset Management, LLC April 2019


Recent Posts