Thursday, February 28, 2019

Rhyme Time at the San Francisco Federal Reserve Bank: "Did the Yield Curve Flip? Will the Economy Dip?"

The curve is not yet important. We (possibly delusionally) think we'll be able to tell you when it is.

From the SF Fed blog:
As the yield curve has continued to flatten, worries about recession risk have increased. However, according to conventional metrics, the current yield curve is not inverted and therefore does not predict a recession in the near future.

Perhaps the single most reliable way to see into the economic future is the yield curve—the relationship between maturities and interest rates on government bonds. The yield curve captures the cost of borrowing money to finance consumption, investment, or government spending and thus is of central importance to the entire economy. Its shape turns out to be an excellent predictor of future economic activity.

Yield curves generally come in three different shapes—normal, flat, and inverted—which are characterized by long-term interest rates being above, similar to, or below short-term interest rates. In the United States and many other countries, an inverted yield curve invariably signals a future economic slowdown. In fact, every U.S. recession in the last 60 years was preceded by an inversion (Bauer and Mertens 2018a). While different theories could account for this empirical phenomenon, part of the explanation is undoubtedly the forward-looking nature of long-term interest rates, which incorporate investor expectations about the future economic outlook.

Recent shifts in the Treasury yield curve have prompted extensive discussion in the news about the possibility of a recession. Figure 1 shows the evolution of three different Treasury yield spreads since the beginning of 2018. The gradual downward trend in spreads turned into a more sudden decline since last November, as long-term Treasury yields fell due to downward revisions of the global growth outlook, uncertainty over global trade, and a perceived shift in communication from Fed officials, among other factors. In December, some yield spreads turned negative, including the difference between the five-year and one-year yields, which has remained between –5 and –10 basis points (hundredths of a percentage point) for most of 2019. Some commentators have viewed this as the beginning of a yield curve inversion and a signal of an impending recession.



Figure 1
Different yield spreads are sending mixed signals
Figure 1 shows the evolution of three different Treasury yield spreads since the beginning of 2018.

A yield curve inversion is defined as an episode when short-term interest rates, including the federal funds rate, rise above long-term rates. However, today’s yield curve still slopes upward, if only slightly so, with the ten-year Treasury yield about 20 to 30 basis points above the three-month yield and the federal funds rate.

Visual inspection may be the single best way to judge the shape of the yield curve. Figure 2 plots the recent Treasury yields in comparison to the levels in the beginning of 2018. The yield curve has shifted up, flattened somewhat, and now has a slight “dent” at maturities from one to five years, where yields decline with maturity. But it currently is not inverted according to the usual definition.



Figure 2
Yield curve has shifted in the past year
Figure 2 plots the recent Treasury yields in comparison to the levels in the beginning of 2018.

What single number best summarizes the shape of the yield curve? That is, which term spread (difference between interest rates of different “terms” or maturities) should we focus on?...
...MORE
Related:
St. Louis Fed: Does the Yield Curve Really Forecast Recession?
Professor Damodaran Looks At Yield Curves

Finally, as noted in the outro from December 7's "Blackstone's Byron Wien Seems Chipper, Almost Jolly".
Remember early December? Everyone was talking yield curve around the clock. And so were we. Here's the outro:
note: the above was written before the 2's/5's or 3's/5's or whatever the fashionable bit of the curve that makes the inversion case, inverted.
At the moment the curve doesn't matter, if and when we think it does, you can rest assured we'll go "curve, curve, curve" 24-7.

For now, as a commenter—I forget where—said: "I think the pundits just like saying 'inversion'".
Inversion. It's almost soothing, in a paradoxical kind of way.
Personally, I like paradoxical. More consonants.