Resolving The Contradictory Signal Between The Yield Curve And Housing: Watch Corporate Profits

|
Includes: BBRE, DDM, DIA, DOG, DRN, DRV, DXD, EEH, EPS, EQL, FEX, FLAT, FREL, FRI, FTY, FWDD, HUSV, IARAX, ICF, IVV, IWL, IWM, IYR, JHML, JKD, JRS, KBWY, LRET, NRO, OTPIX, PPTY, PSQ, PSR, QID, QLD, QQEW, QQQ, QQQE, QQXT, REK, RFI, RIF, RNP, RORE, RQI, RSP, RWM, RWR, RYARX, RYRSX, SCAP, SCHH, SCHX, SDOW, SDS, SFLA, SH, SMLL, SPDN, SPLX, SPUU, SPXE, SPXL, SPXN, SPXS, SPXT, SPXU, SPXV, SPY, SQQQ, SRS, SRTY, SSO, STPP, SYE, TNA, TQQQ, TWM, TZA, UDOW, UDPIX, UPRO, URE, URTY, USRT, UWM, VFINX, VNQ, VOO, VTWO, VV, XLRE
by: New Deal Democrat
Summary

Virtually all measures of the yield curve are signaling recession ahead.

By contrast, the rebound in new home sales suggests no recession in sight.

Neither indicator, while very good, has been perfect.

Another long leading fundamental measure of the economy, corporate profits deflated by producer prices, can help separate out the true signal.

Q1 corporate profits will be reported by the BEA Thursday, and meaningful tax payments by corporations for Q2 will begin in several weeks.

Introduction

Is your favored metric housing or the yield curve? Depending on which one you prefer, the economy is either careening towards recession or else “no worries.” Because the two very important leading metrics are giving completely different signals this spring. Is there a way to break the tie? I think there is, and I’ll explore it below.

The yield curve has not been a perfect indicator

Let’s start with the yield curve. At present everything between the Fed Funds rate and the three-year-bond is inverted (and largely has been for almost 6 months). The 5-, 7-, and 10-year treasuries are also inverted as against all maturities shorter than one year. On the contrary, the 20- and 30-year maturities are not inverted vs. short-term maturities, and from the 5-year maturity out, the yield curve is normal. I’ve depicted these several maturities in the graph below:

I want to emphasize something important: none of these individual yield curve metrics is perfect. For example, the long end inverted for almost the last 1/2 of the entire 1960s, but only inverted for about one month before the “Great Recession”:

The 10-year minus three-month spread also inverted slightly in 1966, but just as importantly failed to invert at all before the 1990-91 recession:

Short-Term maturities also inverted in 1966 and several times in the 1980s and 1990s:

Indeed, most maturities registered a false positive in both 1966 and 1998.

Housing has not been a perfect indicator either

Now let’s compare that with housing. Last week Bill McBride a/k/a Calculated Risk posted the below graph of the 3-month average of YoY new home sales:

along with the statement that “New Home Sales appears to be an excellent leading indicator, and currently new home sales (and housing starts) are up solidly year-over-year, and this suggests there is no recession in sight.”

Indeed, helped by a significant decline in mortgage rates, and a sharp deceleration (and in the case of new home sales, decline) in prices, most housing sales metrics appear to have bottomed.

But housing isn’t perfect either. Just turning to Bill McBride’s graph again, new home sales declined severely in 1966 and also 1995 with no recession occurring. Contrarily, new home sales declined less in 1999 than they did in 2018, and rebounded by about the same percentage in 2000 as they have this year so far, and yet a recession followed within 12 months.

A fundamentals based approach including corporate profits helps identify the correct signal

So I think relying on housing for an “all clear” signal is problematic as well. More broadly, relying on any one metric on an ad hoc basis amounts to cherry-picking data to suit a pre-conceived conclusion, and why making use of a model with a constellation of data metrics, looking for a widespread if not universal movement in one direction is so helpful.

In the case of the “indicator death match” between housing and the yield curve, I think another acknowledged long leading indicator based on the fundamentals of the economy helps us out: corporate profits.

Along with housing permits and two financial-based indicators, real M2 and corporate bond yields, corporate profits are one of Prof. Geoffrey Moore’s long leading indicators. They have an established track record of giving us fair warning about how the producer side of the economy is doing. I have found that they perform even better when deflated by the producer price index for final demand vs. deflated by unit labor costs (Moore’s metric). Here’s what they look like going back over 70 years:

Notice that they barely dipped in 1966 and the other times when housing sales gave us false negatives. They also declined sharply when the dot-com bubble popped, telling us that coupled with even the shallow decline in housing in 1999 enough damage had been done to bring on a recession.

Let me present this information in two ways. First, here is the YoY% change in corporate profits deflated by producer prices vs. single-family housing permits:

Usually, but not always, both have been negative on a YoY basis before a recession began. And there have been several brief false positives measured this way.

But even more accurate is to measure each from its decline from its business cycle peak. Here’s the table, showing not just each decline from the peak as of the month a recession began, the total declines by the end of the recession and also other important declines in each that did not result in a recession:

Expansion

S/F HousingPermits

Corporateearnings

Combined %Decline

Declines at

Recession bottom

1961-70

-20.5

-16.2

-36.7

-34.1 / -20.5

1971-74

-39.9

None

-39.9

-54.2 / -24.9

1975-79

-38.5

-9.3

-47.8

-62.8 / -22.9

1980-81

-38.4

-6.6

-45.0

-54.5 / -24.7

1982-90

-35.1

-4.4

-39.6

-50.9 / -6.2

1991-2001

-6.5 [-12.5*]

-13.2

-19.7

-10.0 / -21.0

2001-07

-58.7

-19.3

-78.0

-81.3 / -53.7

——-

—-

—-

—-

1966-67

-41.0

-4.0

-45.0

1978-79

-26.9

-1.9

-28.8

1985

-26.4

-1.6

-28.0

1986

-11.1

-21.5

-32.6

1989

-28.0

-11.1

-39.1

1995

-21.3

n/a

-21.3

1998

-n/a

-16.4

-16.4

2010

-29.3

-12.3

-41.6

2015-16

-n/a

-12.1

-12.1

2009-19**

-11.7

-1.8

-13.5

*bottom reached before recession began

**to date

First, I call your attention to the fact that even sharp declines in housing, without significant declines in corporate profits, have typically failed to produce recessions. With the exception of the shallow 2001 recession (in which the 9/11 terrorist attacks played a role), typically it has taken a combined decline of nearly 40% in the two metrics for a recession to occur. And with the exception of 1990-91, a decline in corporate profits of over 20% by the time they bottomed has been an earmark of a recession. In general, each % decline in corporate profits seems to be about twice as important as the equivalent % decline in housing.

Second, the “near misses” typically involved important government policy reversals. In the case of 1966, LBJ’s “guns and butter” firehose of military and social program government spending simply overwhelmed everything else. In the case of the several slowdowns in the 1980s and 1990s, the Fed quickly lowered interest rates. by contrast, renewed fiscal stimulus in the next 18 months is almost impossible to envision, and for now the Fed has indicated it is maintaining its “patient” wait and see attitude about interest rates.

Conclusion: No recession signal yet, but stay tuned. Q1 corporate profits will be reported tomorrow

For now, the final line in the above table shows us that, so far, neither fundamentals-based measure — housing and profits — has declined anywhere near enough as of their last reports for a recession to be imminent, based on historical measures.

But tomorrow, the BEA will report corporate profits as part of its second estimate of Q1 GDP. If S&P earnings have been correct, these will decline significantly (note my “placeholder” series of proprietors’ income was already reported to have declined in the first estimate).

Additionally, as shown in the graph below, it appears that corporate tax receipts reported to the Department of the Treasury do a good job in real time of forecasting the trend in corporate profits (note the three big YoY divergences in 1981, 2001 and 2018 coincide with corporate tax cuts):

The drawback here is that almost all of the payments (outside of April 15) are made during the last three weeks of the quarter. So we won’t be able to see how these are running in any realistic sense for Q2 for several more weeks.

If corporate profits, especially in Q2, rebound, that will be potent evidence that the rebound in housing sales is the correct signal. If they decline, beware the inverted yield curve!

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.