A Study on the Flattening Yield Curve
by Ron Griess
Proprietor, The Chart Store
www.thechartstore.com
December 13, 2005

Originally published in The Chart Store Weekly Chart Blog for the Week Ending 12/9/05

The shape of the U.S. Treasury yield curve and related spread analysis has been a “hot” topic in recent months. We decided to try to draw a few new charts to portray the so-called “conundrum” with specific emphasis as to its effect on the stock market.

What we have used for our analysis and chart renditions since July 1954 are:

The weekly S&P 500 Composite
The weekly average effective fed funds rate reported by the Federal Reserve
The weekly average 10 year U.S. Treasury yield reported by the Federal Reserve
The spread ratio calculated by dividing the 10 year yield by the effective fed funds rate
We have only shown the yield spread ratio up to 5 since weekly settlements often caused distortions in the fed funds rate in the early years of the study
We have used green color coding to represent those weeks when the spread ratio is greater than 1, and red color coding to represent those weeks when the spread ratio is equal to or less than 1.

The first chart show the entire history of our 51-year study:

Chart 1

Here’s what we see in Chart 1:

All recessions since 1970 were preceded by the yield spread ratio going below 1 (going inverted)
All but one of those recessions (1990) started with the ratio below one (inverted)
The number of weeks of inversion before the onset of each recession is:
1) January, 1970 to November, 1970 – Beginning April 12, 1968, all but 7 of 90 weeks prior to January, 1970

2) December, 1973 to March, 1975 - Beginning February 23, 1973, all 41 weeks prior to December, 1973

3) February, 1980 to July 1980 - Beginning September 29, 1978, all 70 weeks prior to February, 1980

4) August, 1981 to November, 1982 - Beginning October 3, 1980, all 44 weeks prior to August, 1981

5) August, 1990 to March, 1991 – Beginning January 20, 1989, all 52 weeks through January 12, 1990. For the next 28 weeks prior to the beginning of the recession, the ratio averaged 1.0411.

6) April, 2001 to November, 2001 - Beginning May 26, 2000, all 45 weeks prior to April, 2001

We will now turn our attention to the effect that yield curve inversions have on the S&P. We have drawn charts for each decade since 1950.

Chart 2

Comments for Chart 2:

The yield curve inverted briefly during the middle of the recession. The S&P had already bottomed.

Chart 3

Comments for Chart 3:

There was no yield curve inversion preceding or during the 1962 "bear" market. Yield curve inversion did take place during the 1966 "bear" market and continued well into the bounce off the 1966 low. Inversion started again in 1968 and lasted into the recession of 1970. The S&P was up initially on the inversion, but then declined into the middle of the recession.

Chart 4

Comments for Chart 4:

The yield curve was inverted for almost all of the 1973-74 "bear" market. Inversion also took place for most of the time period from late 1978 until the summer of 1982. The S&P was choppy during this period between 92 and 136 before putting in the August, 1982 bottom at 101.44.

Chart 5

Comments for Chart 5:

The final bottom in August, 1982 ushered in a period where the yield curve did not invert in earnest again until early 1989 (the 3 occasions in 1986-87 appear to be settlement spikes). This inversion was over by the time the market collapsed into the October, 1990 bottom.

Chart 6

Comments for Chart 6:

From the October, 1990 bottom in the S&P, inversions took place for only 7 weeks in 1995, 1996 and early 1998. The sustained inversion from July, 1998 until November, 1998 was coincident with the S&P downturn during that period.

Chart 7

Comments for Chart 7:

The yield curve inversion which began in April, 2000 was coincident with the initial drop from the retest high of the S&P in mid-2000 through the onset of the recession.

Our generalized conclusions are:

Recessions have been preceded by yield curve inversions since 1970. The lead time averages over 40 weeks.
The S&P 500 does not do well when the yield curve is inverted and one measures performance over the entire span of the inversion.