Friday, November 16, 2012

Small Cap Valuations: Market Cap Bias and the Dangers of Headline P/E Ratios

       Using headline valuation metrics as a barometer can be particularly dangerous and misleading, especially for the Russell 2000 Index.  In addition to distorting the index, a market cap "blind spot" has created persistent inefficiencies (and opportunities) at the smaller end of the small cap universe.  Our small cap strategy, with a current median market cap of roughly $350 million, concentrates on smaller companies because this is precisely where thorough research matters the most.  
  
      During our recent conversations with institutional investors and consultants, we have frequently heard that "small caps look expensive."  Indeed, the Russell 2000 currently sports a headline P/E ratio of nearly 30x trailing earnings compared to around 14x for the S&P 500.  Is this conclusive evidence that small caps are overvalued -- especially relative to large caps?  

      We have been somewhat puzzled by this conclusion, as it contradicts what we have been seeing "on the ground."  As bottom up investors, we spend our time researching one company at a time.  While we are inherently intimidated by high valuations, we continue to come across plenty of small, well-run enterprises which appear to be both undervalued and out of favor.  Likewise, many of the quarterly earnings conference calls we tune in to have continued to garner muted interest at best (with some having no Q&A participants at all)--not exactly a sign of "irrational exuberance" for small caps.  Both of these observations are confirmed by the most recent additions to our small cap portfolio which are valued well below the Russell 2000 Index by most metrics.  So...what gives?

      One part of the explanation lies in the structural market bias we have previously referred to as the "market cap blind spot."  A majority of sell side analysts and professional money managers concentrates on the largest companies in the small cap universe and ignores the smallest ones.  This stubborn reality is reflected in the holdings data for buy-siders:  the average small cap mutual fund in the Bloomberg universe has a median market cap of $1.8 billion - a full three and a half times the $491 million median of its Russell 2000 benchmark.  In addition to creating persistent inefficiencies in the small cap marketplace, this bias has distorted valuations within the index itself.  An average of the valuation metrics presented indicates that the largest quintile of the Russell 2000 is 20% more expensive than the smallest.  This relationship underpins a counterintuitive dynamic in small caps, namely, the size of the company--not necessarily the merits of the business--is often the determining factor of its valuation.  This contrasts sharply with large caps, where size generally plays no discernible role in company valuations.

Evidence of the Blind Spot

 Russell 2000 Median Valuations by Quintile (10/31/2012)*

Market Cap (Millions)
Trailing P/E
Forward P/E
Trailing EV/EBITDA
# of Analysts
1 (Largest)
1,642
18.6
15.2
13.5
10
2
854
16.7
14.1
12.4
8
3
491
16.3
13.9
13.3
7
4
269
16.4
13.6
11.8
5
5 (Smallest)
150
14.9
13.2
11.3
4

 S&P 500 Median Valuations by Quintile (10/31/2012)*
  

Market Cap (Millions)
Trailing P/E
Forward P/E 
Trailing EV/EBITDA
# of Analysts
1 (Largest)
56,433
15.2
13.2
12.3
28
2
19,999
14.7
13.8
12.4
24
3
7,722
16.4
13.8
12.2
19
4
7,668
16.8
14.5
12.2
19
5 (Smallest)
3,932
15.5
12.8
11.5
16

*We chose to use median data to eliminate the effect of extreme numbers and also believe it is more representative of company valuations found in each quintile. Data assembled from Bloomberg, L.P.

      Because of market cap bias, the Russell 2000's headline P/E (a weighted average) is inflated, as the most expensive companies have the greatest representation in the index.  This dynamic is further compounded by the fact that the Russell 2000 has more than twice the proportion of companies with exceptionally high P/E ratios* (often caused by marginal recent profitability).  In our view, median data produces a far more meaningful comparison because it is less affected be these distortions.  We also believe that median data is more indicative of the valuations of a "typical" company in the index and also more representative of the characteristics which an active manager is most likely to uncover during his or her research.

*10% of Russell 2000 names have a P/E ratio over 40 compared to only 5% of S&P 500 names.     

Median Data Comparison (10/31/2012)


Market Cap (Millions)
Trailing P/E
Forward P/E
Trailing EV/EBITDA
# of Analysts
Sales Growth
Earnings Growth
Median R2K Company
491
16.9
14.2
8.9
7
5.5%
7.0%
Median S&P 500 Company
12,495
15.9
13.6
9.0
23
2.4%
4.5

      When examined on a median basis, the valuation disparity narrows substantially: small caps are only 4% more expensive than their large cap counterparts despite appearing twice as expensive on a headline P/E basis.  We think that this slight valuation premium for small caps is more than warranted considering their stronger sales and earnings growth characteristics (in red).