WALL STREET'S BULLISH BIAS

On Wall Street, money is king!  "Winning" is making more money than ever before and more than your peers.  Wall Street in aggregate makes much more money during bull markets than bear markets. As a result, the industry has a bullish bias; it tends to inflate earnings expectations and misapply the historical average price to earnings ratio ("P/E ratio") with the result that the market appears undervalued most of the time.  Stress on the word "appears."

For example, last year at this time, Wall Street's consensus 2012 operating earnings forecast for the Standard and Poor's 500 Index ("the Index") was around 107, up 10% from 2011. The actual figure for 2012 will come in at around 100.  That seven percentage point overestimate, given historical annual earnings growth of 7%, is a big miss.

Furthermore, there has been little notice given to the fact that third quarter Index earnings were actually DOWN 3% from last year's earnings.  Since aggregate revenues of the Index were up in the third quarter,  lower earnings imply a slight DECLINE in profit margins, which have been close to all time highs of late. 

Has this recent earnings decline dampened Wall Street's consensus operating earnings forecast for 2013?  Nope! The estimate is 113, up 13% from 2012.  How can that be with the confluence of the following: the U.S. economy is currently growing nominally at around 5% annually, profit margins  have at least temporarily peaked and may be declining,  there will  probably be modest austerity measures put into place to start deleveraging at the federal government level, and the EU is at best going to grow moderately in 2013.

 Granted, housing starts, an important engine of growth during normal times, have bottomed and have rapidly increased recently, albeit from a very low base.  And car and truck sales are recovering nicely.  Finally, recent statistics support the notion that China's growth rate will not decline as rapidly as previously thought.  Netting these considerations out, I would be surprised if earnings were to exceed 107 in 2013.

I have tried to show that Wall Street tends to inflate earnings prospects.  What about the P/E ratio? The historical average P/E for the Index is 15.  Keep in mind that there have been 47 recessions in this country's history, or, on average, one every five years.  (I do not remember an occasion where the  consensus Index operating earnings forecast for the next year was down from the previous year.)  Unfortunately, many strategists tend to forget that the historical average P/E ratio  should be applied to mid-cycle, or normalized, earnings not to bottoming or peaking earnings. A lower than average P/E should be applied to peak earnings and vice versa for trough earnings.  Index  earnings can fluctuate violently over a business cycle.  Consider recent cycles: earnings peaked at 57 in 2000, bottomed at 39 in 2001, peaked at 92 in 2007, bottomed at 40 in 2009, and are at 100 currently.  A Wall Street strategist who applies a 15 P/E to peaking earnings is likely to miss the next bear phase.  For example, even multiplying a more rational  earnings of 107 for 2012 times a 15 multiple results in a price target during the next twelve months of 1605, a 12% appreciation from current levels. With that approach, the Index appears  undervalued at this time. How does one know when earnings are peaking?  One doesn't.  But with  profit margins near record highs  and one quarter of lower margins having already occurred, one might have some concern that the peaking process has already started.


How To Adjust Index Fair Value For Earnings Cycles

To seek a better way to consider whether the Index is undervalued or not, perhaps one should visit academia, where money is important but truth more important.  Yale professor Robert Shiller, whom I highly respect, has devised what he calls CAPE, an acronym for "cyclically adjusted price earnings ratio".  He takes the average twelve month  reported Index earnings over the last ten years and divides that by the Index, producing a P/E ratio.  He then compares that with the historical average P/E to determine over, under, or fair valuation.  This should be more accurate in that it takes actual earnings rather than forecast earnings, thus removing that bullish bias, and accounts for earnings over many cycles, thus removing the tendency to misapply the average P/E to peak earnings.  According to this model, as of yesterday, the Index was 27% overvalued. 

My approach has similar objectives, but uses a somewhat different technique. I use actual operating earnings instead of reported earnings because reported earnings are after writeoffs which can distort earnings trends.  Based on a least squares technique, I try to establish an earnings trendline with an upward slope of 7% annually, which represents normalized earnings growth.  Any point on that trendline could be deemed "normalized", or "mid-cycle", or "trendline" earnings.  Normalized earnings will be below peak and above bottom earnings.  Currently it is 79.  I then multiply that by the average historical P/E of 15 to arrive at 1185, the Index fair value.  According to this approach, the Index, at 1428 is 20% overvalued.

One caveat: In my approach and presumably in Professor Shiller's,  I assume not only normalized  earnings but normalized interest rates as well.  An inflation rate of 2% annually has usually meant a ten year U.S. government treasury note yield of 5%.  That is the number I use in my dividend discount model to arrive at the 15 P/E ratio.  Currently, that yield is 1.65%.  All things remaining equal, a lower than normalized interest rate should mean a higher P/E and a higher theoretical or fair value for the Index.  I stick with my normalized method because the Federal Reserve has  lowered interest rates to stimulate the economy and once our economy resumes normalized growth (reaches escape velocity) the yield on the ten year government should approach 5%.  (That would mean a horrific loss of principal in long maturity fixed income instruments which would dwarf the interest income earned.  As I have said on several posts, the bond market is in a bubble.)

Even with the Index's  being overvalued by 20%, the Index (inclusive of dividends) should return pretax 7% a year for the next ten years with a regression to fair or normalized value during that period.  That compares favorably with the meager 1.65% return on government notes.  Over a five year period the Index generates a total return of 5% a year if a regression to normal valuation were to occur within that period.

The Index is significantly more attractive than bonds over the next five to ten years.  This reminds me of the joke about the man who was asked to eulogize an acquaintance who was not a good guy.  The eulogizer couldn't think of anything positive to say without prevaricating, so he went to the podium and said "His brother was worse!"  Both government bonds and the Index are overpriced, but bonds are worse.

The sentiment indicator I follow is still neutral.  Neither an extreme of bullishness nor bearishness has occurred.  I remain 30-35% long equities (my core position) with the remainder in cash equivalents.  If the Index goes to or below fair value, I shall increase my equity exposure by as much as 100% depending on the extent of the undervaluation.