S&P 500 Valuation Analysis: Near Bottom

 |  About: SPDR S&P 500 Trust ETF (SPY)
by: Denis Ouellet, CFA


Absolute PE ratios on the S&P 500 Index averaged 15 since 1927 with one standard deviation ranging between 10 and 20. Although useful in determining if absolute PEs are high or low, this is not the definitive buy-low-sell-high approach.

Being a crude discounting factor, PE multiples vary considerably depending on inflation rates. Look at the volatility of PEs and inflation between 1927 and 1950. This period was characterized by unusual rapid fluctuations in PE multiples as corporate profits and investor sentiment swung along with deflation and very high inflation rates.

click to enlarge images



Long time investors will remember the Rule of 20 which was used to assess PE multiples given various inflation environments. The chart below plots the sum of trailing PE multiples with yoy inflation rates on a monthly basis since 1927. We can observe that since about the mid-fifties PE + Inflation have fluctuated around 20.


The Rule of 20 has been a useful rule in the post-war era but it should have been called the Rule of 15 before the 1950s when PE multiples were generally lower. The Depression and a severe deflationary period between 1930 and 1933, and another milder one in the late 1930s, resulted in a dismal decade for corporate profits. Then came World War II with more unusual economic conditions, sharply rising prices throughout the 1940s and continued depressed corporate profits.

The CPI was 17.70 in December 1926, dropped 29% to a low on 12.60 in May 1933 and was only 15.50 at the time of Pearl Harbor in December 1941. Prices took off during and after the war and the CPI peaked at 24.50 in September 1948, a 58% jump in 7 years, before deflating mildly until June 1950 and spiking again briefly until mid-1951 (Korean war).

S&P 500 profits were $1.23 in January 1927, troughed at $0.41 in January 1933 and saw $1.23 again only in early 1947. Twenty years later!

Profits doubled during the following 3 years, only to stay essentially flat for another 5 years.

The S&P 500 Index (as per Robert Shiller) peaked in September 1929 at 31, went below 5 in June 1932 and saw 31 again in September 1954, a quarter of a century later. Aren’t we happy not to have been investors during that period?



It is thus not very useful to look at the early part of the 20th century for clues on stock market valuation. Depression, deflation, inflation, recession, war, inflation, deflation, war, inflation. Nothing to foster a sustained economy nor build investors confidence, quite the opposite.

The following chart begins in 1950 and plots PEs and inflation (inversed scale) to show the inverse correlation between PE multiples and inflation. Note the recent dichotomy highlighting the disinflationary effect of the economic crisis and the significant decline in PE multiples reflecting investor uneasiness towards where all this is heading.


This looks like a major buying opportunity! According to the Rule of 20, with inflation near zero, PE multiples should be anywhere between 15 and 25. Using the low end of the range, given S&P 500 earnings estimates of $40-45 for 2009 and $60-65 for 2010 we get the following range of values and returns for the next 18-24 months:

------------ ----PE----- ------------ ----------- RETURNS ------------
15 20 25 current 685 15 20 25
40 600 800 1000 40 -12% 17% 46%
45 675 900 1125 45 -1% 31% 64%
50 750 1000 1250 50 9% 46% 82%
55 825 1100 1375 55 20% 61% 101%
60 900 1200 1500 60 31% 75% 119%
65 975 1300 1625 65 42% 90% 137%
Click to enlarge

The above table uses expected “reported earnings” which include “provisions” and “write-downs” which can be considered non-recurring “one-time events” and thus not represent the operating earnings base. Goldman Sachs estimates “operating” EPS of $63 and $71 for the S&P 500 Index for 2009 and 2010 respectively. This approach makes sense in crisis periods since investors should try to focus on companies’ earnings potential rather than crisis-induced profits. The drawbacks are that, one, to a large extent, the “non-recurring” losses are designated as such by management, two, we could be in a period of recurring “non-recurrings”, and, three, historical PEs used in the analysis are generally based on reported earnings.

Using a range of “recurring” earnings estimates around Goldman’s mid point of operating EPS, equities are excessively cheap at current level.

------------ ----PE----- ------------ ----------- RETURNS ------------
15 20 25 current 685 15 20 25
55 825 1100 1375 65 42% 90% 137%
57 855 1140 1425 67 47% 96% 145%
59 885 1180 1475 69 51% 101% 152%
61 915 1220 1525 71 55% 107% 159%
63 945 1260 1575 73 60% 113% 166%
65 975 1300 1625 75 64% 119% 174%
Click to enlarge

Using Value Line database, I found that 76% of the 1396 companies with positive earnings trade below 15 x trailing EPS and 50% trade at less than 10 x trailing.

Risks to the forecast:

  • Earnings in 2009 could be lower than forecast. We will be watching the Leading Economic Indicators.


  • Deflation, bringing a long period of very low corporate profits. Tough to prevent that other than stimulating demand.
  • Rising fears of coming high inflation, depressing expected PE multiples. Inflation is not in the cards yet.


These are not insignificant risk factors which likely explain the current low market levels…providing what could be a phenomenal buying opportunity.


Robert Shiller proposes the cyclically adjusted PE approach which uses average EPS on a trailing ten-year basis which are currently $61.76. At 685, the S&P 500 now trades at 11 times cyclically adjusted EPS, very much near its absolute historical low levels. In case you did not notice it on the chart below, the cyclical PE was 4.5 in June 1932.



Following is Goldman Sachs’ look at the Fed Model:

The Fed Model approach at its core compares the S&P 500 earnings yield (inverse of P/E using NTM (next 12 months) earnings) and the 10-year Treasury yield. The rationale for the model is that investors ultimately choose whether to invest in equities or bonds, and that the yield differential of the two assets should be roughly comparable over time. We calculate the S&P 500 upside or downside from fair value by assuming the spread between the 10-year Treasury yield and the earnings yield reverts to the 10 year trailing average spread with a price change by the S&P 500 that closes half the gap.

The Fed model has been criticized by some investors who claim the concept of the model is too simplistic and some assumptions are flawed. A primary critique is that the Fed model compares the Treasury bond yield, which is nominal, with the S&P 500 earnings yield, which is real. The consequence of comparing a nominal to a real measure is that the Treasury yield will fluctuate based on inflation expectations while the earnings yield should not.

We base our Fed Model valuation on an average of two versions of the model: the traditional model comparing the S&P 500 earnings yield to 10-year Treasury yield and one which substitutes 10-year BBB corporate bond yields for US Treasuries. In this modified version, investors choose whether to invest in the equity or credit of a corporation. The yield differential between the S&P 500 and the BBB corporate bonds is typically narrower than the spread between S&P 500 equity yield and US Treasuries.

The average of our Fed Models using our top-down earnings estimates ($40 for 2009) suggests the S&P 500 current fair value is 785. Using the Fed Model and our top-down earnings estimates suggests the year-end 2009 fair value for the S&P 500 is 1210. We assume the yield gap closes by half through the price appreciation of the S&P 500, and the other half closes by changes in bond yields.


Here again, I will use Goldman Sachs’ approach:

The DDM is our preferred method to gauge the valuation of the S&P 500 and it offers several advantages versus other valuation techniques. The DDM is not a relative measure so it does not rely on historical comparisons to calculate a fair value. A common
criticism of relative valuation measures such as a P/E analysis is that the conclusion is based on the assumption that the relative comparison is fairly valued, whether the relative
comparison is a historical average or another asset such as bonds. The DDM relies less on comparable measures, although the Treasury yield is an important input into the model.

The Goldman Sachs Global ECS DDM is a standard 4-stage dividend discount model that can be used to compare valuations across markets. Based on our updated topdown earnings forecasts, the Global ECS DDM suggests a current fair value of 1020. The Global ECS DDM assumes a longer-term payout ratio of 32%, the average payout ratio over the past 5 years. The US Portfolio Strategy DDM assumes a lower longterm payout ratio of 26%, fuelled by slower growth in Financials dividends relative to recent years. Other long-term assumptions standardize the Global ECS DDM across equity markets, but differ from the US Portfolio Strategy DDM assumptions.



Never mind earnings. What is the intrinsic value of equities. The current S&P 500 book value, per Goldman Sachs, is $525 which means that the Index is now selling at 1.3 x BV. Goldman says that, since 1973 (the earliest available), bear markets have troughed at 1.3 x BV (Average P/BV since 1973 is 1.9 x), although the chart below from Ned Davis Research shows P/BV troughing at 1.00 in 1974 and 1982, both brutal bear markets characterized by high inflation and interest rates.

Furthermore, the most recent official figure for book value is $529 for year-end 2007. Given S&P 500 earnings of nearly $50, year-end 2008 BV is likely closer to $550.

In addition, RBC Capital Markets has this chart showing that 1/3rd of index equities trade below book value in both Canada and the US, levels only seen in 1990.


Using Value Line database, I found that 44% of the 1700 securities listed with positive book values currently trade below book value.


This is a financial crisis and financial stocks have led the way down and continue to scare investors. It is unlikely that the market will turn definitely unless and until financials turn around. When might that be? If history is any guide, this RBC Capital Markets chart says we could be right there.



Given the complexity and severity of the crisis, investors will need some encouragement to believe that the worst is passed before committing seriously to equities again. Buying too early can be rewarding but also costly if considered over a 6 months period:


As usual, equities will take off well before the economy actually turns as investors begin to anticipate the end of the crisis and the beginning of the recovery. This is true of all bears, even the mama bears:


The key resides in the housing market. House prices must stop declining before banks can begin to breathe and for investors to begin to discount the end of the economic and financial death spiral. Other things needed are a reasonably active consumer achieving an adequate balance between spending and saving, and China keeping economic growth above 7%.

Disclosure: no positions