7.4 C
New York
Monday, February 6, 2023


S&P 500 Fair Value: "Have it Your Way"

Courtesy of Doug Short.

When searching for the perfect S&P 500 fair value metric, readers will encounter a wide range of methods for establishing the “correct” fair value. Most of us interested in this pursuit have conflicting views on the accuracy or applicability of these methods when viewed in the context of current state of our economy. Many believe earnings for the S&P 500 will continue to rise; others believe they are destined to fall. Some believe averaging earnings over a period of time helps smooth data.

One of the most common valuation metrics familiar to Advisor Perspectives readers is the one popularized by Yale Professor Robert Shiller. The chart below shows the monthly average of S&P 500 closing prices divided by 10 year average earnings and adjusted for inflation based on the Consumer Price Index (CPI) and earlier pricing data for the period before 1913, when the Bureau of Labor Statistics took on the task of creating this indicator.

Professor Shiller’s latest chart (as of January 23rd) puts the current P/E ratio at 21.14. The Long Term Interest Rates reside in the background and pose no significant contribution to the chart except for general informational purposes. Years ago, however, when I first began dissecting the data, many questions occurred to me. Why adjust for CPI? Who believes the data anyway? Isn’t one of the most important guidelines for a study is to use the same data over the same time period? Knowing that BLS changed the data significantly several times after 1980 means the data may be “corrupt” unless using the same methods prior to 1980. According to John Williams of Shadowstats fame, the current CPI as calculated under the 1980 system would be above 10%.

Regardless of the CPI question, what about simply comparing a chart with and without CPI-adjustment to determine a baseline to let readers decide which to prefer? The following chart removes the long term inflation rate and replaces that line with the nominal (actual) price to earnings ratio using the same calculation Professor Shiller uses for CAPE (cyclically adjusted price to earnings).

Readers may be surprised to see that the inflation-adjusted and nominal versions correlate rather well. Professor Shiller uses the real calculation to determine how much over- or under-valued the S&P index is by comparing the current PE (21.14) to the long term historical PE (16.42). The latest data must be added to the spreadsheet as because Professor Shiller’s data, especially the earnings, isn’t current, and can sometimes be several months old.

Although Q4-11 earnings are not yet “in the books,” Standard and Poor’s Senior Earnings Analyst Howard Silverblatt updates the earnings weekly on the Standard & Poor’s website. We can plug the Silverblatt data into Professor Shiller’s spreadsheet to move our calculations into the “real time” category shown below.

Using current earnings of $89.46 (1 year trailing), the updated spreadsheet (in yellow) reflects a new CAPE of 20.48 as of 12-31-11. Geeky data miners and Excel wizards relish this endeavor (while most probably abhor the process). But if we are to take this data seriously, why not present a chart with the latest information in one place?

That wasn’t too hard, and we now have Professor Shiller’s data updated … and on steroids (the legal kind). Rather than viewing the first chart and trying to determine the “fair value,” this updated chart shows all the relevant information. The blue line clearly depicts the 10 year cyclically adjusted price-to-earnings ratio and the S&P Composite equivalent value. It is easy to see that, based on 10 years of average earnings, we briefly fell below “fair value” back in March of 2009 but now remain well above that level.

What about our thoughts on using inflation-adjusted earnings versus nominal earnings? The chart below shows the comparison.

Both charts use a logarithmic vertical axis to preserve the ratio of change over time. The only difference is that we begin at 1 for the nominal version (since the actual S&P Composite was 5 in 1881) versus 100 for the inflation-adjusted version. The reason is that we’re adjusting to the current dollar value, which raises the earlier price data (to around the 80 level). Interestingly enough, both show about the same current overvaluation (996 versus 991). Those of us who look from 30,000 feet probably don’t care much about a 1% difference when the current value of the index is above 1300. The important item to note is simply that we are overvalued based on both 10 years of cyclically-adjusted data and nominal data.

Whoa! Ten years … what about other timeframes? Yes, I’ve made comparable calculations for 5, 10, 15, 20, and 30 years on both CAPE (the Shiller method) and nominal. The following table lists the results.

The table presents time periods that should suit any individual’s preference. In an effort to create the ultimate “fair value” indicator, the addition of three metrics seemed warranted, a monthly PE value, the use of Earnings Yield, and some sort of “sentiment” indicator.

First, the monthly P/E averages simply use the monthly averages of daily closes (in this case 1243) divided by monthly earnings (89.46) to provide the monthly P/E (13). This number is then averaged for respective time period (i.e. price divided by an average of 5 years of monthly P/Es). Note that the 5-year chart below tracks the S&P relatively well.

Second, many like to use the current earnings yield for valuation purposes. Why not average those in as well. The following chart uses 5 years of earnings divided by current price to arrive at a valuation.

The last indicator used would be something that reflects “sentiment.” Which, for better or worse, the sentiment is where supply meets demand for prices, simply ? the S&P 500 index itself. In the table, sentiment over time exists as the average of monthly closes.

Combining all the indicators over various time periods produces the Combined Fair Market Value, shown below.

Based upon all the indicators and across all time horizons, the S&P 500 index remains overvalued by a substantial margin. My advice is to consider this data simply as information that provides a snapshot of historical relative valuation rather than “tradable” ideas. Investors should understand that knowing a “fair value” target provides an educated backdrop for making “investments” or “trades.” But as the charts and tables above illustrate, the market can remain over- and under-valued for long periods of time. The full report with all the charts is located here. We can revisit this fair value table as earnings season progresses.

Note: For readers unfamiliar with the S&P Composite Index (a splice of historical data different from the S&P Composite 1500), see this article for an explanation.





Notify of
Inline Feedbacks
View all comments

Stay Connected


Latest Articles

Would love your thoughts, please comment.x