Why is There Such a Large Discrepancy Between The S&P’s Forward and Trailing Multiple?

Depending on how you’re looking at the S&P 500’s current earnings multiple, there’s a good chance that you could walk away with a pretty divergent view of the how attractive equities are from a valuation standpoint.

If you’re looking at the S&P’s multiple on a trailing basis, you might see that the index is trading at somewhere around 19x earnings, which is a pretty hefty price.  On the other hand, if you look on a forward basis, you might find that the index is trading at around 14x earnings, which could reasonably lead you to conclude that stocks are cheap.  The difference between the multiples is about 25%, which means that if the forward multiple traded up to the current trailing multiple, the index would be above 2000.

Since the numerator of the P/E’s is the same, the difference is obviously a function of the denominator.  In 2012 the S&P 500 earned $90 and in 2013 the published estimates have the $SPX earning $110.  Are S&P earnings really going to grow by >20% this year? And if so, where is the growth coming from?

The index level of the S&P 500 is 1658, but that number and the related earnings estimates are really pretty arbitrary, based on an index level set in 1957.  The real total market value of the S&P 500’s constituents is $15.75 Trillion.  These companies earned $866 Billion last year on a reported basis, which is the number reflected by $90 in index earnings.  The current index estimate of $110 implies that the aggregate earnings number would have to get to around $1.03 Trillion next year, suggesting growth of $162 B.

By looking at the bottom up EPS estimates for the S&P 500’s constituents we can get a better sense of where this $162 B is coming from.  The majority of the difference is due to the fact that the trailing number represents as reported GAAP earnings, while the $110 number reflects operating earnings before one time charge-offs. For instance, HPQ lost $12 B last year on a GAAP basis (thanks to the autonomy charge-off), although it generated $9 B in EBIT.  Based on analyst estimates HPQ’s 2013 earnings will narrow the gap between trailing 2012 and 2013 operating earnings by $19 B .  Below is a list of companies that are expected to contribute more than $2 B in incremental earnings to the S&P 500.  60% of the growth is coming from 20 companies and 90% is coming from 50.

From the HPQ example it should be more clear that a lot of the discrepancy between trailing and forward PE ratios may be driven by the fluctuations in GAAP earnings from one time expenses rather than true earnings growth.  Perhaps that’s an argument in favor of using the forward estimates as a proxy for more normalized earnings.  However, you could make the argument that in the aggregate, one time expenses are actually recurring.  In 2012 35 S&P 500 companies lost money on a GAAP basis; in 2013 analysts only expect 7 to lose money on an operating basis.

GAAP Earnings vs Estimates

 Source: Avondale Estimates, Compustat Data

S&P 500 Positive on 64% of Days in 2013

In terms of volatility 2013 has so far been one of the best years in the last 20.  That’s because the $SPX has been up on 64% of all trading days so far this year, which is a couple of percent better than the next best year, 1995, when the index was higher 62% of the time.  The amazing thing is that the 20 year reference period encompasses the 1990’s bull market, so the fact that this year stands out on top is perhaps even more impressive.

For the record, the S&P 500 closed higher on 53% of all trading days in the past 20 years.  If the index were to finish the year in line with the average, that implies that between now and the end of the year the market would only close higher 47% of the time.

Percent Positive Days by Year

How Likely is A Continued Market Rise?

One week after I wrote a post about the S&P 500 being up 14.5% year to date, the index is now up 16.5%!  As I wrote last week, this is a spectacular increase even when judged on an annual basis, and our year to date rise already puts 2013 in the top 1/3 of all years in $SPX history.

As the previous statistic would imply, at least 1/3 of all years show greater increases than we have had so far this year, so one might think that a continued rise isn’t impossible.  However, if you take the market’s initial valuation into account a continued rise would be a more rare event.

Using the Shiller Cyclically Adjusted PE ratio, the market started the year at 21.9x earnings.  Since 1900 there have been 33 years that the market has begun the year with a Shiller PE ratio greater than 20x.  Only 7 times has it risen more than 16% in those years (using the $DJIA).  All of those occurrences happened within the last 20 years, most of them in the 90’s.

Many of the older investors that I speak to who have been investing for 20+ years are some of the most bullish people that I know.  It’s likely that part of the bullishness is because these investors have seen the illogical heights of valuation that securities can be pushed to in a bull market.  Still, to the extent that one believes that the 90’s were an anomaly and that we have spent the last 12 years unwinding a stock market bubble, a return to higher valuation is unlikely.  Within that framework it seems similarly unlikely that we will have an annual close much higher than we are today.

Dow Jones Annual Return vs. Shiller PE

Source: FRED, multipl

(Note: Yes, there are plenty of flaws with the Shiller PE, but it does help smooth the effect of abnormal spikes that occur with a traditional PE ratio like in 2009  and it has been calculated back to 1900.  I also realize that it’s not a perfect comparison to use CAPE and Dow, but it’s close enough)

8th Longest Economic Expansion of All Time

The current economic expansion is now in its 47th month going back to June of 2009.  That means that since 1857 there have only been seven periods of expansion that have lasted longer than the current one.  Out of the 33 total cycles in that period, the average expansion lasted 38 months.  Since World War II the average has been 58 months in 11 cycles.

As the market continues to climb, we’re entering that crucial 50 month period where two secular bear market expansions stalled (1933-37 and 1974-1980).  On the other hand, it’s totally possible that we’re on a path more like  decade long expansions of the 60’s, 80’s and 90’s.  So which will it be Mr. Market?  The answer has pretty big implications for whether one thinks that the market can continue to advance from here.

Economic Expansion DurationSource: NBER


Stay For May?

Tomorrow will be the last trading day of April, which means that many investors will be bracing for the wrath of rhyme time by preparing to “sell in May and go away.”  As I noted back in December, May has been a difficult month for four of the last five years, but that doesn’t mean it’s always bad.

Look on the bright side!  May is positive not only some of the time, but in fact most of the time (51% positive since 1900), and has had some really good years in the past 113.  Below are the top 10 Mays in Dow history.  Several even occurred following strong annual starts, although in most of those years the May mark was near the high point for the year.

Best Mays Ever

Signs of Investor Skittishness

After today’s 127 point gain,  the Dow has been up, down and then up again by more than 100 points for three days in a row.  Below is a chart of the other times that the Dow has had alternating triple digit moves (either down, up, down or up, down, up) since 1999.  In case we have a 100 point down day tomorrow, I also highlighted the times that the Dow has displayed this behavior for four days in a row (either up, down, up, down or down, up, down, up).

The reason I did this analysis is that my hypothesis was that these large swings were indicative of market indecision and possibly high levels of stress that could be signals of market turning points.  Eyeballing the data below doesn’t seem to disprove the hypothesis.  The three day swings may be slightly less conclusive; however, the four day swings, which are less frequent, tend to cluster around important market crossroads including the March 09 low, the October 07 top and the 02/03 bottom.  One observation that may be significant is that these type of swings hardly happened at all in the 03-07 bull market, but have occurred frequently in the most recent period.  Perhaps this is evidence that the whole bull run has been characterized by higher than typical skittishness and volatility.

Alternating Consecutive Triple Digit Dow Moves

How Often Does the S&P 500 Decline by 1%?

Through February 20th there’s only been one day this year that the S&P 500 has fallen by 1% (Update: well, make that two).  That marks the 3rd year in a row that the year has gotten off to an extremely mild start.  In 2012 it took until March to have a 1% down day and that ended up being the only one in the first quarter.  At this point in 2011 there had only been two big daily declines.  On average, the S&P 500 falls by 1% 26 times per year, a little over 6 times per quarter.

1 Percent Declines by Year.jpg



Where do Growth Stocks Peak?

Over the course of the recent bull market there have been a few growth stocks that have hit extreme levels only to come crashing down.  Although some of these have recovered slightly, those that come to mind include: NFLX, GMCR, OPEN, MNST and CMG.

As a post-mortem on these stocks, below is a chart of the price to sales multiples that they hit at their highest levels.  For comparison I included the peak price to sales multiples of five stocks that had similar sentiment (judged subjectively) at the 07 peak and five from the dot com era.  Also included are the current multiples of six growth stocks that haven’t slowed since 2009.

momentum price to sales peak
Note: red bar is the average multiple for the group

S&P Annual Performance After a Big January

This is an update to a post that I first wrote last year, the last time that the S&P 500 had a big rise in the first month of the year.

When the S&P 500 has a good first month, it has statistically been followed by a really good year.  The index has risen by more than 4% in January 18 times in its 56 year history.  In those years it has averaged a 21.1% return for the full year, and it has been up double digits in every one of those years except for 1987 (which was a good year up until the October crash).

The S&P 500 has never been negative in a year with a big January, but it’s worth noting that if a similar analysis is performed on the Dow, which has a 118 year history, there are five years (out of 28) that the index was up more than 4% in January and ended negative for the year.  Many of those years were significantly negative too: the average loss was 18.4% and the list includes 1914, 1929 and 1930.  The index ended those years down 30.7%, 17.2% and 33.8% after being up 5.1%, 5.8% and 7.5% in January respectively.

Weird eerie coincidence, the Dow has had a daily crash three times in its history: in 1914, 1929 and 1987.  All three years had big Januaries.

Number of Days Since Last 3% Down Day

2012 was a pretty mild year as far as volatility is concerned.  There were two periods of correction, but both were relatively light and there wasn’t a single day that the S&P 500 was down 3% or more.

Markets have calmed down to the extent that it’s actually been 448 days since the last time that the S&P 500 has fallen by 3% or more in a single day.  At today’s level on the Dow that would be a 400 point decline.  For comparison, since 2008 we had grown accustomed to getting a decline that large once every 32 days on average.
Looking at S&P history since 1957, the current 448 day streak is better than average, but not quite at the best levels that the index has ever seen.  Over that period, a 3%+ daily decline happens about once every 217 days.  However there are several long periods without them. There was no such decline for 11 years between 1962-1973.  Even recently there wasn’t a 3% decline for nearly 1500 days between 2003-2007. That streak was broken on February 27, 2007.