The Great Valuation Debate: Navigating a 'New Era' Market - Part I
Below is a series of valuation indicators this week, including three separate views of the Price-to-Earnings ratio, two approaches to the Price-to-Dividend ratio, the Price-to-Book ratio, the Price-to-Cash Flow and then a couple composite models that were designed to provide an overall reading of market valuation.
Since this is a fairly large amount of data to review, the research is broken up into several parts this week. Without further ado, here is a review of the P/E ratios...
The P/E ratio is the granddaddy of all valuation metrics as most investors believe that earnings are the lifeblood of the market. As such, measuring the ratio of current prices to corporate earnings would seem to be a worthwhile endeavor as this tells us how much investors are willing to pay for $1 of earnings at any given time.
However, it is important to recognize the following:
Corporate earnings can be viewed in a multitude of ways
GAAP (Generally Accepted Accounting Principles) Earnings are the best measure of earnings
Companies like to use "Operating Earnings" which tend to exclude a lot of bad stuff that purportedly (key word) occurs on a one-time basis (despite the fact that the bad stuff tends to occur quarter after quarter)
Earnings tend to dive during recessions, which dramatically skews the P/E ratio
- The P/E ratio is an absolutely miserable "timing" indicator for buying and selling decisions
With that said, here are some P/E ratios...
GAAP P/E Ratio
As of 6/30, the current reading of the P/E ratio using GAAP earnings on the S&P 500 was 17.48. The good news is the 50-year average is 19.20 and the 25-year average is 24.80. The key here is that the credit crisis caused the P/E to spike to about 125.0 during 2008-09, which obviously skews the data.
Prior to that, the highest level occurred during the tech bubble bear market when the GAAP P/E hit the high 40s. Next, we should note that the average since 1926 is 16.97.
The key to this analysis is that prior to the mid-1990s, this P/E ratio stayed within a range of about seven and 28. Since 2000, that range has changed dramatically as the low has been about 13 with the high of 125. And if we throw out the 2008 debacle, the range is more like 13 to 45.
If you look at the full history of the indicator going back to 1926, the average GAAP P/E - aka "fair value" - is 16.5. And since the "expensive," (or overvalued) zone historically has been about 20.0, it can be concluded that the current reading of 17.48 for the S&P 500 is "fairly valued" and would become overvalued at about 1835.
However, the bulls will argue that in the "new era" of the stock market (since the mid-1980s), the current reading is undervalued. Thus, the heroes in horns suggest that there is plenty of room to the upside using this indicator.
Operating Earnings P/E Ratio
Okay, let the games begin. In short, operating earnings are considered "financially engineered" results as earnings wind up being whatever the company would like them to be. And the bottom line is that most companies wind up "beating" the Street's consensus estimates by a penny or two on a regular basis.
The idea of excluding one-time items such as the cost of restructuring or the acquisition of another firm is valid. Therefore, the way to look at operating earnings is by removing "write-offs." This approach includes the benefits of true one-time items but avoids the other accounting shenanigans.
Since 1926, the "norm" for this P/E has been 15.2. Over the past 50 years, the norm, or average, has been 16.08. And over the past 25 years, the average has been 18.77. Once again, the 2000-02 and 2008 periods clearly skewed the historical P/E's to the upside.
For example, the high during the Great Depression was just above 26, which was a level that was not approached again until 1999 when the tech bubble truly got out of hand. The readings over the next four years waffled between 30 and 22, and then spiked back up to 28 during the credit crisis.
So, the current reading of 16.17 is, like the results of the GAAP P/E, a little above the historical average of 15.2 -- but on the low end of the readings seen since the mid-1980s (when "operating earnings" began to become widely used).
Looking at the chart, if one were to "throw out" the extremes seen in 2000-02 and 2008, it appears that an "expensive" or overvalued reading would be in the 18 zone, while an undervalued level would be about 12. Therefore, the current reading of 16.17 seems to fall in the "fairly valued" zone.
"Normalized" P/E Ratio
Another way to play the P/E game is to "normalize" the GAAP earnings over a period of time via the use of a linear regression. This approach attempts to put the extreme readings into perspective by comparing the current readings to the long-term trend.
From this perspective, this is an elegant solution to the problem of the dislocations that can occur during times like 2000-02 and 2008-09. However, it is important to recognize that this approach is best used as a "from 30,000 feet" type of indicator.
Using a linear regression from 1927, the "normalized" GAAP earnings on the S&P 500 is $78.6. Thus, at Friday's closing price of 1680, the current normalized P/E is 21.37. Unfortunately for the bulls, this is an elevated reading.
Once again though, despite the fact that we are using a linear regression approach here, there are distortions to what would otherwise be a fairly definable range. The problems are, as you might suspect, the 1928-30 and 1999-2002 periods.
However, using this approach, the S&P tends to be "overvalued" when the normalized P/E is above 19 and "undervalued" when below 10.
If we toss out the extreme periods, it looks like the S&P would be overvalued when above the 16 zone and undervalued when below 11. As such, the current reading on 21.37 can only be viewed as being overvalued.
Yet at the same time, there is a caveat here. We must understand the math involved with a linear regression approach. Remember that earnings cratered to just $14.88 per share on the S&P 500 during the 2008 crisis. This was from a reading of $81.51 in 2006 and $66.18 in 2007.
After the crisis, reported earnings rebounded to $50.97 in 2009, $77.35 in 2010, $86.95 in 2011, and $86.51 in 2012. Thus, it is safe to say that the current linear regression reading of 78.6 is artificially low due to the $14.88 reading seen in 2008.
In sum, the primary takeaway from our review of the various P/E ratios is that the extreme readings seen during 2000-02 and 2008-09 have greatly distorted the data. The bottom line appears to be that one must make a decision as to how to interpret the information provided.
One approach is to simply look at the current readings in relation to the historical data and employ an "it is what it is" view. This approach suggests that stocks are currently "fairly valued."
Another way to play is to assume that the market is in a "new era" (since the mid-1980s) and that we are dealing with a new valuation range. While there were clearly extremes seen during this period, the bottom line here is that the S&P appears to be modestly undervalued.
And finally, the other approach is to "throw out" the extreme periods and try to rework the ranges. Using this approach, we see that stocks are somewhere between fairly valued and slightly overvalued.
Regardless of how you choose to play this game, it is vital to understand that valuations tend to fall into the category of, "Things don't matter until they do - and then they matter a lot!" So, until traders begin to worry about valuations, simply check in on these indicators every month/quarter to ensure that you don't get surprised by an extreme reading.
Read more on Benzinga:
The Great Valuation Debate: Are Stocks Overvalued? - Part 2
- The Great Valuation Debate: How to Spot Fairly Valued Stocks - Part 3
© 2017 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.