On Normalized P/E Ratios Over Six Decades
After my last post, Bill Rempel (among others) inquired about the difference in normalized P/E ratios over different decades. He felt a standard applied across 70+ years might not be particularly useful today, because so much has changed (increased participation in equity markets, different monetary policies, etc.) As a result, it might be that while “relatively cheap” is better than “relatively expensive” within each time period, it is inappropriate to compare years from dissimilar decades as if the same standards applied.
I agree. So, over the next two posts, I’ll try to give you an idea of what the compound point growth in the Dow looked like following the low normalized P/E years and high normalized P/E years within each decade. In other words, I’ll look at low normalized P/E years and high normalized P/E years relative to other years in the same decade.
In this post, I’ll simply show you the results of two different comparisons across decades.
The first comparison consists of a group of the five lowest normalized P/E years from each decade vs. the five highest normalized P/E years from each decade. The second comparison consists of a group of the three lowest normalized P/E years from each decade vs. the three highest normalized P/E years from each decade.
Although this is still a comparison across many decades (for this post, I’m only using 1940-1999, because I want to use only “complete” decades), it should give you some idea of whether the normalized P/E effect is simply a result of a few years in just one or two particular decades, or whether the effect tends to hold up over many different decades.
In my next post, I will go a step further, and actually break down the results decade by decade.
The Least You Need to Know
If you can’t be bothered to read yet another post on normalized P/E ratios (I’ve written quite a few lately), here’s the least you need to know:
1. The five lowest normalized P/E years from each decade from the 1940s through the 1990s saw higher compound point growth in the Dow over the subsequent 1, 3, 5, 10, 15, 20, 25, and 30 years than the group of the five highest normalized P/E years from each decade.
2. The three lowest normalized P/E years from each decade from the 1940s through the 1990s saw higher compound point growth in the Dow over the subsequent 1, 3, 5, 10, 15, 20, 25, and 30 years than the group of the three highest normalized P/E years from each decade.
Obviously, there’s a lot more nuance to the results than is suggested by the above two statements; but, the most important point to take away from this post is simply that the combined low normalized P/E group (drawing equally from all decades from the 1940s through the 1990s) beat its high normalized P/E counterpart over all the holding periods I measured.
Remember, these are the combined groups. This post doesn’t address the normalized P/E effect within specific decades; I’ll discuss that in my next post.
As you look over the compound point growth data, also remember that low normalized P/E years are not necessarily low P/E years. This is reflected somewhat in the data. For instance, in these first two groups, the average P/E ratio is 12.81 for the low normalized P/E group and 14.61 for the high normalized P/E group. The difference in 15-year normalized P/E ratios is quite a bit larger – 11.30 for the low normalized P/E group and 16.02 for the high normalized P/E group.
The Results
If you group the five lowest normalized P/E years from each decade and the five highest normalized P/E years from each decade, this is what the subsequent compound point growth looks like:
1 Year
Low Normalized P/E Group: 9.88%
High Normalized P/E Group: 6.78%
3 Year
Low Normalized P/E Group: 9.84%
High Normalized P/E Group: 5.99%
5 Year
Low Normalized P/E Group: 9.93%
High Normalized P/E Group: 5.46%
10 Year
Low Normalized P/E Group: 8.69%
High Normalized P/E Group: 6.61%
15 Year
Low Normalized P/E Group: 8.63%
High Normalized P/E Group: 6.08%
20 Year
Low Normalized P/E Group: 8.19%
High Normalized P/E Group: 5.48%
25 Year
Low Normalized P/E Group: 7.21%
High Normalized P/E Group: 5.61%
30 Year
Low Normalized P/E Group: 6.22%
High Normalized P/E Group: 5.94%
The difference between the mean and median returns within each group for each span of years is usually very small. However, there is one exception: 1-year point growth for the high normalized P/E group.
The mean is 6.78%; the median is 3.21%. This wide disparity between the mean and median occurs only over a one-year period – by three years, the mean and median are more or less in agreement.
In these next two groups, the average P/E ratio is 12.73 for the low normalized P/E group and 15.07 for the high normalized P/E group. The difference in 15-year normalized P/E ratios is quite a bit larger – 10.80 for the low normalized P/E group and 17.31 for the high normalized P/E group.
If you group the three lowest normalized P/E years from each decade and the three highest normalized P/E years from each decade, this is what the subsequent compound point growth looks like:
1 Year
Low Normalized P/E Group: 6.95%
High Normalized P/E Group: 3.55%
3 Year
Low Normalized P/E Group: 8.89%
High Normalized P/E Group: 3.01%
5 Year
Low Normalized P/E Group: 9.64%
High Normalized P/E Group: 3.77%
10 Year
Low Normalized P/E Group: 9.24%
High Normalized P/E Group: 6.12%
15 Year
Low Normalized P/E Group: 8.69%
High Normalized P/E Group: 5.49%
20 Year
Low Normalized P/E Group: 8.70%
High Normalized P/E Group: 4.98%
25 Year
Low Normalized P/E Group: 7.38%
High Normalized P/E Group: 5.58%
30 Year
Low Normalized P/E Group: 5.98%
High Normalized P/E Group: 5.88%
The difference between the mean and median returns within each group for each span of years is usually very small. However, there is one exception: 1-year point growth for the high normalized P/E group.
The mean is 3.55%; the median is 1.84%. So, if we were comparing median point growth instead of mean point growth, the low normalized P/E group would outperform the high normalized P/E group by 7.10% vs. 1.84% instead of 6.95% vs. 3.55% over a one-year period.
I’m not saying that the median is a better measure than the mean; I’m just saying that considering the variability within the group, anyone making a bet on the direction of a high normalized P/E market over a one-year time span needs to know that the rather tame mean return for this group obscures the fact that unusually big up or down moves are actually quite common in such years.
Related Reading
On 15-Year Normalized P/E Ratios for the Dow
On Normalized P/E Ratios and the Election Cycle
On Normalized P/E Ratios and the Election Cycle (Again)
On Normalized P/E Effects Over Time
On Calculating Normalized P/E Ratios
On the Difference Between Actual Earnings and Normalized Earnings
On the Dow’s Normalized Earnings Yields for 1935-2006
In Defense of Extraordinary Claims