S&P 500 P/E Ratio at Troughs: A Detailed Analysis of the Past 80 Years

There is a great debate among investors and academics about “trough” PE multiples, i.e. what is the appropriate multiple to apply to S&P 500 earnings at market trough. There is clearly a need for thorough objective analysis. The following charts and observations, going back to 1927, should help in this debate.

CONCLUSIONS

• Using historical absolute PE lows to assess the potential downside to the S&P 500 Index is simplistic and based on superficial, non-rigorous analysis. The absolute historical lows used by the bears, while strictly accurate, were attained in high inflation periods, not comparable to the present.
• Using the Rule of 20 to assess PE multiples takes into account the inflation environment and is thus a better tool to value equities in general.
• Using this method, and assuming inflation rates in the 0-2% range, trough PE multiples should be 12-14 times trailing earnings.
• The current financial crisis is substantially distorting S&P 500 earnings, both reported and operating, in an unprecedented way. Using trailing earnings, reported and operating, can result in a meaningful underestimation of Index earnings in the present environment.
• Macro earnings estimates are more appropriate in the current exceptional circumstances. Goldman Sachs’ $63 estimate for 2009 appears conservative in light of historical evidence. A low probability worst case scenario would take earnings down to the $43 level.
• Trough valuation analysis shows trough S&P 500 Index levels at 720 using 2009 estimates (which are lower than trailing), with a low probability downside risk to between 516 and 602.

ABSOLUTE PE MULTIPLES

A large number of economists, analysts and investors use absolute PE multiples to assess the downside risk left in equity markets, an approach which has the merit of being simple and straightforward but, as we will see, is in fact too simplistic and superficial.

• The absolute lowest S&P 500 PE multiple at month-end has been registered in September 1974 at 7.0x trailing EPS.
• Single-digit PEs have been recorded in four other instances: July 1982 (7.7), April 1942 (7.7), February 1948 (8.4) and June 1932 (9.4).
• Other market lows have been reached at higher multiples as seen in the chart below, some in the low teens, some in the high teens although bears would presumably point out that none of these (except 2002) were MAJOR bear markets.

(sorry chart missing)

The absolute PE approach is too simplistic since it does not consider the financial environment of each of the periods reviewed. PE multiples are crude discount factors that need to account for, among many factors, interest rates and/or inflation rates over the forecast periods.

The chart below plots US yoy inflation rates. Arrows indicate when PEs troughed and the red circles highlight periods that witnessed the four single-digit trough PEs (excluding 1932 when deflation was raging). Every single-digit PE year occurred during one of the four double-digit inflation periods that the US experienced during the 20th century. In each case, inflation rates were above 6% at the trough month (1942: 12.5%, 1948: 9.3%, 1974: 12.0%, 1982: 6.4%).

(sorry chart missing)

THE RULE OF 20

In my March 2 post S&P 500 Valuation Analysis: Near Bottom, I explained how “The Rule of 20” incorporates inflation in the PE equation (and why it was more like “The Rule of 15” before 1950). 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” has been around 20, generally oscillating between 15 and 25.

Using The Rule of 20, PE + Inflation has been the lowest in 1982 at 14.1(excluding the 1932 deflation-ravaged –0.6). It was 14.6 in 1935, after inflation turned positive, and 14.9 in 1957 when inflation was 3%. Other market troughs were above PE + Inflation = 17.

(sorry chart missing)

THE PROBLEMS WITH INFLATION

The last bout of high inflation rates was extinguished by Paul Volcker and Ronald Reagan in the early 1980’s. Many investors thus fail to understand its negative impact on PE multiples. Here is why high inflation rates compress PE multiples:

• High inflation rates are, next to deflation, the worst economic evil. Sharply rising consumer prices destroy the value of money and, if not checked in time, self-feed into an upward wage-price spiral that eventually forces monetary and fiscal authorities to take action through higher short-term interest rates and budget contractions to create a recession to kill inflation and inflationary expectations.
• Equities face intensifying competition from fixed income investments, particularly short-term investments, as interest rates rise and the economic outlook deteriorates. Hard assets such as gold also attract more investment money seeking shelter from rising inflation.
• Corporate profit margins get squeezed as costs, primarily wages, rise sooner and faster than selling prices.
• The quality of corporate profits declines as inflation produces inventory profits (under FIFO accounting) that investors know are not sustainable.

(sorry chart missing)

During inflationary periods, equities are understandably less appealing to investors which demand a higher discount factor to account for higher perceived economic and corporate risks and lower earnings quality.

It seems that investors begin to worry about inflation when it exceeds 6% which may be a level where fears of it becoming uncontrolled are rising, leading to anticipation of forthcoming strong policy actions causing a recession to stop it.

Deflation is even worse because once consumers become convinced that prices will be lower next month, they naturally and logically stop buying (like the housing market currently). The collapse in demand forces prices further downward which results in an economic death spiral that is virtually immune from policy actions. Japan lived through such a period in the 1990’s.

THE DIFFERENCE BETWEEN MARKET TROUGHS AND PROFIT TROUGHS

Equity markets are discounting machines; they generally anticipate economic turns 6-9 months before the actual turn. Corporate profits generally begin to improve only AFTER the economy has turned. It follows that equity markets generally begin to recover while economic news remain grim and well before profits even begin to stabilize.

As the chart below shows, with few exceptions (market declines during economic expansion), equity markets have troughed 5 to 14 months before profits turned up, and the turn in profits occurred at levels that were 4%-31% lower than their level at market trough.

(sorry chart missing)

That means that waiting for the economy and/or earnings to trough before buying equities can be very costly. In the 9 instances when equities troughed before earnings, the S&P 500 Index had appreciated between 11% and 37% by the time earnings actually troughed, the average and median gain both being 26%.

REPORTED VS OPERATING EARNINGS

Since the mid-1980’s companies have been showing 2 sets of earnings: “reported” GAAP earnings and “operating” earnings or GAAP earnings excluding non-operating profits and losses and write-offs. Companies which record “non-recurring” gains or losses or which book write-downs or write-offs in asset values in a given year also provide “operating” earnings so that investors get a “more accurate” picture of a corporation’s “true” profits from continuing operations and its realistic earnings power going forward.

The pros and cons of “operating” earnings are that:

• It makes sense for corporations to segregate from operating earnings those gains or losses which are deemed outside of the normal operations of a company (e.g. asset sales, discontinued operations). This is also useful to investors who can better appreciate the recurring profit generating capabilities of companies.
• Write-offs and write-downs are more debatable since they reflect a loss in value of certain specific assets which management or auditors deem inflated in a company’s books given present and expected circumstances. On the one hand, such losses in values reflect the inability of management to extract acceptable or appropriate returns from these assets at their current book values, justifying their inclusion in earnings. On the other hand, to the extent these write-offs or write-downs are one-time non-recurring events, excluding them from “operating” earnings provide a more accurate picture of the company’s earnings power under its new adjusted asset base.
• There is admittedly a degree of discretion for corporations to classify gains or losses as “recurring” or “non-recurring”. Some companies have used this discretion in such a way that they report “non-recurrings” so often that they become recurring non-recurrings.

Reported vs operating earnings has only been a big factor in market valuation in early 2000, when the tech bubble burst, and …now.

(sorry charts missing)

Also, historical “reported” earnings are not as “clean” as one may think. Accounting and accounting standards have evolved considerably over a century. As a young analyst in the mid-1970s, I had to thoroughly comb financial statements and the notes to uncover “unusual” accounting treatments and calculate true earnings and returns for some of the more liberal companies. I can only imagine how earnings statements could be tweaked in decades previous. Interestingly, the tendency then was generally to enhance earnings through the more liberal use of accounting rules (and other tricks for some even more liberal companies)! If widespread, this would have resulted in historical PE multiples actually being deflated by creatively enhanced profits.

THE CURRENT PROBLEM WITH HUGE LOSSES, OPERATING OR REPORTED

Whatever one thinks of the reported vs operating earnings debate, it is fair to say that the current financial crisis has brought the debate to another level because of the sheer size of some companies’ losses and the way Standard & Poor’s accounts for earnings.

As Wharton professor Jeremy Siegel wrote in the Feb. 25 WSJ (see articlehere), it is important to understand how S&P calculates “aggregate” earnings for the S&P 500 Index:

(…) Unlike their calculation of returns, S&P adds together, dollar for dollar, the large losses of a few firms to the profits of healthy firms without any regard to the market weight of the firm in the S&P 500. (…)

As the fourth-quarter earnings season draws to a close, there are an estimated 80 companies in the S&P 500 with 2008 losses totaling about $240 billion. Under S&P’s methodology, these firms are subtracting more than $27 per share from index earnings although they represent only 6.4% of weight in the index. S&P’s unweighted methodology produces a dismal estimate of $39.73 for aggregate earnings last year.

If one applies market weights to each firm’s earnings using the same procedure that S&P employs to compute returns, the results yield a more accurate view of the current profit picture. Market weights produce a reported earnings estimate of $71.10 for 2008 — nearly 80% higher than the unweighted procedure. The reason for this stark difference is that the firms with huge losses generally have extremely low market values and hence have a much smaller impact on the total earnings in the index.

Similarly, operating earnings (essentially, earnings before write-offs), of the S&P 500 are boosted to $81.94 per share when earnings are weighted by market value, yielding a P/E ratio of about 9.4 for the market, instead of S&P’s $61.80, which yields a P/E ratio of 12.5 when firm profits are simply added. Even the negative earnings for the fourth quarter disappear when market weights are accounted for, as fourth-quarter GAAP earnings on the S&P 500 Index total $7.44 per share and operating earnings reach $14.40.(…)

At the extreme, and we are admittedly in an extreme period, a large company with a tiny market capitalization could incur losses so large as to wipe out most of the S&P 500 earnings (AIG lost over $60 billion last quarter alone). As a result, the Index PE would skyrocket even though the other 499 stocks’ valuation would actually not change at all. In effect, a casual or superficial observer looking at the Index would conclude that equities are expensive or overvalued when, in fact, 499 stocks would be cheap or undervalued.

With the release of its Q4 2008 results, AIG subtracted $5.13 to S&P 500 Index operating earnings and $7.10 to reported earnings in the December quarter. These losses will negatively impact the S&P 500 Index earnings throughout 2009. Yet, AIG is 0.02% of the S&P 500 Index so its market value has literally no meaning to the overall Index. Were the US government to completely nationalize AIG tomorrow, its removal from the Index would make no difference to the Index value but the removal of its losses, operating and reported, would immediately boost Index earnings by 7.8% for operating and 17% for reported.

In all, there is no easy way out of the problem and, as usual, hard work, thorough analysis and common sense must prevail.

MACRO EARNINGS ESTIMATES

For the reasons mentioned above, it may be more appropriate and useful at this time to use earnings estimates derived using a macro approach. Goldman Sachs does this exercise to forecast $63 for 2009 earnings assuming that sales decline 6% and margins bottom at 5.8%. Goldman assumes that peak-to-trough margins will decline 248 basis points this cycle, a reasonably conservative assumption given previous cycles experience (see table below). Using 4% margins like in 1990, earnings would be $43.

(sorry chart missing)

S&P 500 INDEX VALUATION

• Valuation using the Rule of 20 method give “trough” valuation of 791-923 for the S&P Index using current trailing earnings, 3% to 20% above current levels.
• Using 2009 operating earnings estimates, “trough” valuation would be 720-840.
• The worst case scenario, using the $43 estimate would bring trough valuation of 516-602.

REMAINING RISKS

• The economic and financial environment deteriorates much further and 2009 earnings are materially below current estimates of $40 (reported) and $60 (operating).
• Deflation, which would adversely impact both earnings and PE multiples.