Message of the Yield Curve: No “There” There

There’s a lot of angst swirling around the yield curve these days. The yield curve is what you get when you plot the yield of Treasury bonds of various maturities. When short-term Treasuries have a lower yield than long-term ones, this is a “normal” curve, such as today’s: 2-year Treasuries yield 2.81%; 5-years 2.95%; and 10-years 3.07%. When the curve “inverts” (when 2-year Treasury yields become higher than 10-year yields), this is said to be a bad omen for the economy, and for stocks. Bears are pointing to today’s relatively flat “twos-tens” yield curve (10-year minus 2-year) of 0.26% as a big reason to sell. Our take: The yield curve has no significance now.

This chart shows stocks (S&P 500, green line), the yield curve (twos-tens, blue line), and recessions (pink shaded areas). At first glance, yield curve inversions do seem to give a good warning of pending recessions. We’re not convinced (there are very few data points to work with, and a few near inversions in the 1990s that had no impact), but let’s set that aside and focus on the stock market. No one likes recessions, but there’s nothing we can do about them. On the other hand we can always sell stocks, and the question is, is the yield curve telling us something of value today?

First let’s look at the track record of the curve as a sell signal, namely how we’d have done if we’d sold as soon as the curve inverted, and then waited for a major sell-off before buying again. There have been five inversions since 1975, including the long mostly-inverted period of 1978-1982 (counted as one), and the very brief inversion in mid-1998. During that first 4-year inversion period, the S&P 500 gained 8% from the beginning of the inversion to the low of the next sell-off—not a great return, but a gain nonetheless. The market went up after the signal more than it went down.

Next the curve inverted at the end of 1988, and apart from a short-lived correction in mid-1990, the S&P did just fine. In fact, if you’d sold at the beginning of the inversion, you would have missed a 32% rally leading up to the peak, and the subsequent 18% sell-off would still have left you with a substantial gain. Ironically, the 1988 inversion sell signal came 14 months after the great crash of October 1987—a big down move the yield curve completely overlooked. 

Ten years later the brief inversion of 1998 had no value as a signal—stocks were on a tear, and if you’d sold, the 29% gain before the next signal would have passed you by. Finally the curve got one right when it inverted in early 2000, eight months prior to the Tech Wreck bear market. It got another hit at the tail end of 2005, 22 months before stocks topped out in October 2007.

All in all, not much of a track record. Tallying up the first four signals, you’d have been much better off doing nothing than selling. The gains following the signal and the false signal of 1998 would have been left on the table, and in all likelihood you’d have missed buying the bear market lows to boot. Only the 2005 signal prior to the 2008 bear market would have saved you money overall.

Let’s move on to forward-looking questions: Will the curve invert, and if so, will it matter? There’s no reason to think the curve must invert, a point lost on a lot of analysts these days. Just because we’re close to the zero line doesn’t mean we’ll breach it. Case in point: The curve averaged 0.35% for the entire period from 1995-2000 and only inverted once, for a few short weeks in 1998. There’s no manifest destiny when it comes to yield curves. It might invert. It might not.

We don’t think it will. The Fed has raised rates many times to get us to the current Fed funds rate of 2%, and the 2-year yield is pricing in several more hikes. Core inflation is 2.2%, and the Fed’s favorite inflation indicator, the PCE, is right at 2%. We think it’s unlikely that short-term rates will rise significantly above inflation. At the same time, with QE now in reverse and the federal deficit rising quickly, the growing supply of longer-term bonds should push long rates higher, steepening the curve. 

But really, does it matter? Would a further flattening of, say, 0.30% leading to a 0.10% inversion really change things? It’s hard to see why it would. The implicit reasoning behind the yield curve signal is that an inverted curve is bad, that the Fed raises rates to cool off an overheating economy, and that eventually short rates get too high. The cost of borrowing gets to be out of reach of businesses and banks, leading to a recession and bear market.

The twos-tens yield curve is made up of two interest rates. What if this time it’s not that two-year rates are too high (bad)—what if it’s that ten-year rates are too low (good)? After all, the Fed spent $3.5 trillion in QE to buy bonds during the financial crisis with the explicit purpose of lowering long-term rates and boosting the economy and has only now begun to sell them off. 

At the time of past inversions rates were always much higher than the current 2.75-3% zone. The last time the curve inverted at the end of 2005 rates were 4.4%; during the second half of the 1990s when we had two inversions and a near miss rates were around 6%; during the 1989 inversion rates were above 8%, and during the long inversion from 1978-1982 rates got above 15%. Those rate environments were worlds away from today’s—it’s apples and oranges. This difference alone makes it difficult at best to squeeze out meaningful information from the curve.

The chart at left shows an important relationship that underscores today’s benign rate environment. The blue line represents 10-year Treasury rates from 1975, while the green line shows inflation. Apart from some noisy periods during the inflationary ‘70s and the financial crisis, the 10-year Treasury yield has always been higher than inflation. This makes sense. Investors don’t want to receive an interest rate that would be completely wiped out by rising prices. The Fed’s QE bond buying that began in 2009 has at least temporarily changed that relationship. The chart shows the gap has collapsed.

Since borrowing rates are linked to Treasury yields, it’s cheap to borrow money. The so-called real cost of borrowing (the interest rate minus inflation) is still rock-bottom despite the Fed’s continuous rate hikes. These low rates are a positive, not a negative. Rather than a half-empty curve that says short rates are very high, we see a half-full curve that says long rates are very low.

To be clear, though, we’re not bullish on stocks. We think the market is rangebound now, and that we’re closer to the top of the range than the bottom. From a fundamental perspective the market positives and negatives strike us as being in balance. But what little weight we give to the curve today we score as a plus, not a minus. We’re much more focused on the gap between rates and inflation than we are on the shape of the curve, and we’re looking for that gap to widen. That’s one signal we don’t want to miss.

Stay tuned.

Stephen J. Morton, CFA

Wei Peng, Ph.D, CFA

 Inyo Capital Management LP’s material contained herein represents the opinions of Inyo’s investment team and does not constitute a solicitation or recommendation for the purchase or sale of any commodities, securities, or investments. Although the information compiled herein is considered reliable, its accuracy is not guaranteed.