Sunday, March 15, 2015

Results By Market Cap Quintile

As my readers know, I have a detailed MFI database that has largely tracked the annual returns of the top 50 stocks > 100m for every month since January 2006.  Every now and then I like to take different slices of this data to see what I see.

I have noted in the past that in my analysis, larger market cap stocks have done better than smaller market cap stocks.  The cut off has broadly been over $600 to $800 million.  But I have also noted that this is on an absolute basis. So I have made no adjustments for the rising valuation of stocks.  I mean a $800m market cap company today on average might have been a $400m market cap company in 2006.

One approach would be to put in an inflationary or stock market index to bring everything to today's value, though I'd have some sort of truncation issues as I have always used the $100m lower bound, when in reality to be apples to apples I should probably be using a $200m lower bound today.

What I decided to do was a relative analysis.  So within each tranche of stocks, which is typically 50, I have split them up into quintiles.  So a "1" represents the 10 largest market cap stocks of the 50 stocks, a "2" is 11-20 and so forth.  Here is a table showing annual returns by quintile and by purchase year:

Year           1           2           3           4          5       Total
2006 25.8% 18.3% 4.8% 4.4% 13.4% 13.9%
2007 -21.8% -13.2% -10.1% -25.4% -33.0% -20.5%
2008 -7.1% -8.0% -9.1% -7.5% -9.4% -8.2%
2009 30.9% 46.0% 49.2% 40.0% 30.4% 39.3%
2010 10.9% 20.8% 2.4% 6.5% -0.8% 8.0%
2011 2.7% -7.3% -4.2% -5.8% 10.6% -0.8%
2012 30.7% 37.5% 26.3% 27.1% 15.7% 27.5%
2013 30.3% 36.8% 28.8% 20.9% 5.6% 24.5%
2014 12.7% 4.3% 11.7% 4.9% -2.5% 6.2%
2015 2.0% 1.4% 1.4% 1.7% -1.6% 1.0%
Grand Total 12.6% 14.6% 10.9% 7.1% 3.3% 9.8%


So the average stock has returned 9.8%, though if we focused on portfolios that have closed, the average is 10.5%.  You can clearly see the pattern I mentioned, the top two quintiles have had an average return of 13.6%, versus the 9.8% overall. So this suggests that a very real formula appraoch (and simple as well) would be to take the top 50 stocks > 100m.  Take the subset of the 20 largest market cap stocks in that list and then randomly select from them.  That would have given you better results in all years except 2008, 2009 and 2011.

To put this into $ terms, if you just picked the top 50 with $100,000 invested equally over first year portfolios you'd have about $189,000 today.  If you used the approach I outlined above, you'd be at $352,000.

Now I should comment that I often list these back test results on my blog.  Just because something worked in the past, there is no assurance it will work in the future.  That is for you, as an investor, to decide whether there is enough evidence to get you comfortable.  I am just a guy with a computer and some data. I am not an investment adviser.

1 comment:

CorpRaider said...

Neat info. Yeah it seems like breaking it down into quintiles takes care of the concerns voiced by others and that's how academics do it.