CAPE, or Cyclically Adjusted PE Ratio, is a commonly used metric to gauge the valuation of the S&P 500. The use of CAPE generally involves taking the current CAPE reading and comparing it to history. Like any PE metric, the CAPE has a numerator equal to current S&P 500 price, but the denominator is equal to average of the last 10 years of earnings, adjusted to present levels based on changes in CPI. Many investors compare the current CAPE reading of 33 to the long-term average of 17 and draw conclusions that the valuations are very high for stocks, or worse, that the stock market is near 100% overvalued and needs nearly a 50% crash just to be at fair value. The comparison of the current reading to long-term averages has significant data issues and likely shouldn't be made.
Buybacks lead to a permanent reduction in share count in excess of the share count that would exist without buybacks and that allows earnings to grow much faster than inflation. Yes, it is true that many companies keep issuing shares each year, but having a buyback will make the share count grow less than otherwise, or if big enough as many are, actually reduce the share count. Therefore, on a relative basis, buybacks constitute a permanent reduction in share count. However, CAPE only adjusts older earnings to bring them to present day level based on inflation and misses how the older earnings (per share) would be greater if they occurred in the present due to reductions in share count over time from buybacks. The issue arises from the fact that buybacks only really existed after 1982 when SEC provided a Safe Harbor for companies to repurchase their shares, so earnings grew faster after 1982 and accelerated after 2000 when buybacks surged, yet older earnings are not being adjusted for this growth.
The same issue arises for the CPI. The CPI nature has changed over time, due to the substitution effect and hedonistic adjustments. Most economists agree (especially from the Boskin Commission) that these effects have led to the CPI reporting lower levels of inflation than say before these changes around the mid-1990s, so CPI inflation readings before the mid-1990s will be greater. Therefore, more recent CAPE readings may not have the earnings of last 10 years increased as much as they would have under pre-1990s CPI, which leads to lower relative CAPE denominator and greater CAPE reading. This makes comparisons of CAPE across time, specifically after the mid-1990s, and before, problematic as they are not like-to-like.
The current level of long-term interest rates has a tremendous effect on valuations of stocks, and in fact all assets. The average 10-year note yield from 1861 to present was 4.7%, yet it currently is barely 3.0% as of September 17, 2018. Therefore, it is only reasonable that valuations be higher now for stocks than in the past, on average. For example, in March 1982, the S&P 500 was at 7.5 times earnings which results in an earnings yield of 13.3%, and it occurred because interest rates were sky high, with 10-year note yield at 13.9%, so at that time, the average highly rated S&P company would pay likely around 15.0% to borrow, so the stock price was remotely reasonable given borrowing cost. But if the same low 7.5 PE ratio existed now in September 2018, and companies can borrow at say 4.0%, (assume 1.0% spread on 3.0% 10 year note), then without contemplating tax advantages of debt, the accretion of borrowing at 4.0% to retire shares with earnings yield of 13.3% (PE of 7.5) is nearly 1,000 basis points. Meaning there is massive earnings "arbitrage" available to borrow cheaply and buy back stock, which would create a huge bid under stocks and likely drive them higher than a 7.5 PE ratio. This is why stocks don't currently trade with a PE Ratio of just 7.5, since long term interest rates are 3% for Treasurys.
The CAPE completely misses fundamental shifts that affect earnings, like tax cuts. The CAPE reading for September 2018 will use earnings from last 120 months adjusted for inflation to form the CAPE denominator. But the earnings for 112 of the 120 months reflect a corporate tax rate of 35% instead of the going forward 21% rate. The tax cut was said to boost S&P earning an 8% per Credit Suisse, so the fact that nearly all of the older earnings of the last 10 years don't reflect the prospective 8% boost to earnings power means the CAPE currently will understate realistic going forward earnings power to the tune of around 8%.
Overall Impact of Issues
If buyback run rate has averaged 3% of market cap for more recent years and essentially 0% before 1982, then older earnings are lacking a 3% annual upward adjustment to be on the present share count basis. We may assume CPI is adjusting inflation about 60 basis points less than levels after the mid-1990s measures. Baskin Commission economists agree estimated that older year's CPI was adjusted about 60 basis points annually more than that after the mid-1990s.
As discussed above, in the current CAPE reading, older earnings before 2018 need to be adjusted up by 8%.
This does not affect Earnings but it affects what becomes a reasonable PE. So, it doesn't impact how CAPE is calculated but impacts our assessment of a reasonable CAPE. So we need to adjust the last 10 years of earnings by 3% a year for additional buyback EPS growth that didn't occur pre-1982 and for 0.6% annually for CPI being less now. Also provide a one-time 8% boost for Tax reform for 2018-forward.
After these adjustments the Current CAPE drops from 33 to just 26. And 26 is above long-term average of 16, but given current interest rate is below long-term average of 4.7%, the higher valuation is not entirely unreasonable and doesn't warrant saying the stock market requires a 50% drop to get to fair value.
The long term average CAPE of 16.9 is a ridiculous number for valuation. That implies the S&P fair value is about 1500. At S&P 1500, that would mean that the forward earnings yield based on forward earnings of 172, is 11.4%, versus a 10 year note yield of barely 3.0%. This implies an over 8% equity risk yield premium, but the total return risk premium also includes any growth in nominal aggregate Earnings. If we assume nominal earnings grow at nominal GDP then 4% is a roughly conservative estimate assuming a 2% inflation figure, which is the fed target, and 2% real GDP growth. Therefore, if the S&P 500 was 1500, the long term equity returns would include 4% earnings growth plus 11.4% earnings yield or roughly 15.4% versus a 10 year note yield of just 3.0%. An equity risk premium of 12% in excess of the 10 year note is way more than the historical average of about 4.5%. Unless there is a crisis, such a wide risk premium of over 12% versus the historical average of 4.5%, seems unwarranted.
Therefore, the indicated valuation of S&P 500 being at 1500 from CAPE appears unreasonable. It appears prudent to ignore CAPE comparisons as the metric is distortive, just as other metrics such as the Buffet Indicator also have substantial issues. When looking at the forward operating PE of 16.5 for an earnings yield of about 6% compared to a 10 year note of 3%, stocks continue to look at least reasonably valued, while CAPE without any adjustments indicates the S&P fair value is about 1500 which appears unreasonably low compared to interest rates.
If you enjoyed reading this article, and would like to read similar content in the future, please hit the follow button above at the top of the article to follow releases of new articles!
Additional Disclosure: The information contained in this article is an opinion and constitutes neither actionable investment advice nor a recommendation to trade any security.
Disclosure: I am/we are long IVV.
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.