As we are near the end of a very long period of ((slow)) economic and ((fast)) stock market growth, we see more and more SA articles predicting a doom-and-gloom forward period for both the economy (recession) and the market (bust or bear).

Many of these forecasts are founded on relatively high market valuations and economic downturns. It makes sense that if the federal reserve indeed created a wealth effect with historically low interest rates for long periods of time and if those low rates and that wealth effect caused growth in consumption and therefore growth in corporate sales and incomes, and thus increased stock prices due to those higher profits, plus higher multiples and stock buybacks, that when it comes time for reversions-to-means, most everything would reverse, causing: higher interest rates, lower consumption, lower profits and ultimately lower stock prices, and a reverse wealth effect and a recession.

But when?

In a recent commentary, David Rosenberg, Chief Economist and Strategist at GluskinScheff, Inc., laid out a possible schedule to answer the “when” questions. And it is premised on the market highs of last fall being this cycle’s market top. To wit:

every stock market peak presaged an economic expansion peak 100% of the time in the past. And the average lead time from the peak in the market to the peak in the business cycle is 7 months (the median is 8)… in a recessionary bear market, the S&P 500 goes down 35% through the piece. And the piece typically lasts 16 months. Ahead of the recession, the stock market is down an average of 10%, where we are now. And the bottom is generally three months before the recession ends… There have been 13 Fed tightening cycles in the post-WWII era and 10 landed the economy in recession.

If indeed the market peaked on 9/20/18 at 2931, and if it falls 35% from this peak, to 1905 on the S&P 500, and the fall lasts 16 months to January 2020; and if the economic cycle peaks 7 to 8 months after the market peaks at around 5/1/2019, and the market bottoms in January 2020 is 3 months before the recession ends in May 2020, the cycle might look something like this:

Source: Image created by author

Here’s another way to forecast what the next down market and economic (recession) cycles might look like.

Every month Jill Mislinski and Doug Short publish scores of informative charts and graphs at dshort.com. One of the graphs is about how over- or undervalued the S&P 500 is relative to its valuation mean using the average of 4 different methodologies from 1900 to present. See below.

In addition, there is a trove of market, valuation, interest rate and more data at multpl.com going back to the late 1800s. I agree with many SA folks that using data going back that far (horse and buggy for transportation, telegraph and Morse code for long distance communication, abacus and slide rules for calculation, no computers, an agrarian economy, etc.) to include in calculations of long-term economic and monetary "means" or averages, distorts meaningful comparisons to the data of today (algo-driven trading in milli-seconds, instantaneous information overload, AI on the way, too many intangibles on balance sheets, etc.).

If the above is true, it is difficult to pick a different date from which to use data to draw more meaningful comparisons. That said, I suggest using the beginnings of the computer era as the starting date, say the mid-1950s, to make comparisons more accurate and meaningful to today’s situation.

To that end, I downloaded p/e and CAPE-10 valuation data to compare valuations from 1956 to today with same from 1871 to today. Here is what I found for S&P 500 valuations:

Mean (Average) | |||

P/E | CAPE 10 P/E | ||

1871-2019 | 15.7 | 16.6 | |

1956-2019 | 19 | 21 | |

Difference | +20.8% | +25.5% | Avg=+23.2% |

Source: Image created by author

Clearly, more modern history evaluates p/e-type valuation ratios to be 23% higher than longer term history.

So, in my personal work I adjust Jill’s valuation results down to give what I think are more accurate comparisons to today’s data. Jill’s graph above suggests the average of the four valuation indicators is an 86% over-valuation to the mean valuation. If I reduce that 86% by 23.2%, I get an overvaluation of 66% to the mean valuation. I am more comfortable with that number.

The S&P 500 closed 1/31/2019 at 2704. If its overvaluation to its valuation mean was 66%, then fair valuation of the index would be 1629 just a little over a week ago.

Now, look at Jill’s graph again. If the market topped in the fall of 2018 and continues down, which implies a correction or bear, then each additional point (one per month) that’s added to the mean over-valuation calculation is still above the mean for a long time. That implies it will raise the mean even though it’s a declining actual number. That premise holds until the valuation falls enough to equal its own mean. Even as that premise holds, each new data point will have an ever-decreasing impact on each new month’s rising mean. Figuring out each month’s impact is difficult.

Approximating the impact is possible, but we’re dealing with two moving parts, each moving in the opposite direction - the overvaluation number moving down and the mean number moving up. The goal is to figure out at what point does the rising valuation-impacted mean equal the declining actual S&P over-valuation level. To determine that, one needs to understand the relationship between how much the mean moves up as the S&P overvaluation moves down.

A few weeks ago, I wrote an article on SA which said that when the S&P 500 approaches or exceeds an overvaluation to its own mean by 2 standard deviations, then when the market reaches its peak it falls hard and fast. Here’s the money graph in that article:

Of the 5 times since 1900 when the S&P approached or exceeded 2 standard deviations above its regression-calculated mean, it did fall hard and fast. Hard, by averaging a bottom valuation of 13% below the mean, and fast, by doing it in about 2 years’ time from the market peak.

SA contributor, Daniel Amerman, CFA, wrote a fascinating article on Feb 10 “A New Great Recession, Once Every 5 Years,” wherein he pointed out several interesting things about recessions:

- Since 1854, the average business cycle was 58 months, including 40 months of economic expansion, followed by 18 months of contraction (recession)
- Modern (since 1946) business cycles last 80 months, including 68 months of expansion, followed by 12 months of recession
- The stand-out exception to the trend in the modern era is the "Great Recession" of 2007 to 2009 which lasted 18 months
- The reason 2007-09 is a stand-out is that the Fed didn’t have the usual bullets then to fight the battle.

I suggest everyone read Dan's article as it contains a lot of valuable information about recessions and the Fed's inability to mount a good fight during the next one.

The bullets the Fed uses are “Lowering interest rates to escape recession.” To fight recessions, the Fed usually lowers rates quickly and low, to about 1%. In previous modern-era times when the Fed needed to lower rates, the starting points from which to begin lowering rates averaged about 8%, so there was plenty of ammo to fight the fight. In 2007, the relevant interest rate was only 5.25%, so the Fed not only lacked firepower it needed a new weapon, QE, to bring the recession to a stop. And that long recession is referred to as “nasty.”

Where are we now? “the current target Fed funds rates range is only 2.25% to 2.50% - which is a complete outlier relative to the situation coming into previous recessions.” And the Fed is on “pause” when a recession may be just around the corner.

The 2007-09 recession was longer and deeper than modern era “normal,” when it started at 5.25% interest. What are the expectations when starting at 2.25%? A long and deep recession, and nasty bear market diving to 10% below its long-term mean?

Who knows.

But if this current market cycle did hit its peak in the fall of 2018, here is what the next 2 years could look like, assuming the market bottomed out at 10% below its mean (remember, the mean will be rising each month during that time until it equals the mean, and then overshoots it by 10%:

Source: Image created by author

There are probably 3rd and 4th alternatives about what will happen when, not if, the reversion to valuation mean begins to occur and the next recession rears its ugly head.

Alternative thoughts?

**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.