Steven made a comment about how MFI stocks with a market cap over $100b have done very well. I thought I'd take a look at my database and see how they have fared.
I will also look at two other cuts. Stocks with dividends over 5%, he says they do poorly. And stocks over $3b and yield between 2.5% and 5%, he says they do well. I suspect that will be very true as that is close to my MFI Formula approach.
Simple Analysis
My first cut is simple, just a flag by market cap. I suspect this will not be super robust as my cutoff is $100m, I don't think I get that many stocks over $100b.
| Market Cap | 0 | 1 | Total |
| Big | 9% | 22% | 18% |
| Medium | 5% | 20% | 9% |
| Small | 6% | 10% | 7% |
| Total | 5% | 17% | 8% |
So in this table, Big = 100b+. The 0 and 1 are dividends flags with 1 being 2.6% and greater. Small is under $600m. So cross section of Big + Medium and 1 is really my MFI Formula approach.
In this table, Big doesn't do any better than Medium with the dividend component on... it just tends to have more in the dividend flag on. Here is same table without dividend flag but by year purchased:
| Year | big | medium | small | Total |
| 2006 | 5% | 19% | 6% | 14% |
| 2007 | 0% | -15% | -28% | -21% |
| 2008 | 0% | -6% | -10% | -8% |
| 2009 | 4% | 38% | 41% | 39% |
| 2010 | 7% | 12% | 4% | 8% |
| 2011 | 16% | -4% | 3% | -1% |
| 2012 | 16% | 32% | 20% | 27% |
| 2013 | 30% | 27% | 19% | 24% |
| 2014 | 26% | 5% | -2% | 4% |
| 2015 | -2% | -9% | 4% | -4% |
| 2016 | 21% | 10% | 26% | 17% |
| 2017 | 36% | 11% | 12% | 12% |
| 2018 | 3% | 0% | -12% | -4% |
| Total | 18% | 9% | 7% | 8% |
Then This table shows the counts of stocks by each of these buckets:
| Year | big | medium | small | Total |
| 2006 | 1 | 343 | 239 | 583 |
| 2007 | - | 317 | 272 | 589 |
| 2008 | - | 264 | 336 | 600 |
| 2009 | 4 | 274 | 318 | 596 |
| 2010 | 3 | 286 | 261 | 550 |
| 2011 | 13 | 390 | 246 | 649 |
| 2012 | 15 | 368 | 217 | 600 |
| 2013 | 37 | 372 | 191 | 600 |
| 2014 | 26 | 386 | 188 | 600 |
| 2015 | 36 | 343 | 221 | 600 |
| 2016 | 31 | 337 | 232 | 600 |
| 2017 | 14 | 351 | 235 | 600 |
| 2018 | 10 | 356 | 184 | 550 |
| Total | 190 | 4,387 | 3,140 | 7,717 |
So you can see the number of stocks greater than 100b is pretty small.
More Sophisticated
Instead of using $100b as an absolute definer of "Big", it probably makes more sense to deflate it over time. I will use the Russell 3000 annual gains as deflator. Same tables. Here is what "Big" is going back by year.
| Year | Big |
| 2006 | 39,620 |
| 2007 | 41,241 |
| 2008 | 25,962 |
| 2009 | 34,402 |
| 2010 | 40,724 |
| 2011 | 40,497 |
| 2012 | 47,151 |
| 2013 | 62,715 |
| 2014 | 70,403 |
| 2015 | 70,671 |
| 2016 | 79,505 |
| 2017 | 96,055 |
| 2018 | 100,000 |
| Market Cap | 0 | 1 | Total |
| Big | 6% | 23% | 17% |
| Medium | 5% | 20% | 9% |
| Small | 6% | 10% | 7% |
| Total | 5% | 17% | 8% |
| Year | big | medium | small | Total |
| 2006 | -8% | 20% | 6% | 14% |
| 2007 | -42% | -14% | -28% | -21% |
| 2008 | 0% | -6% | -10% | -8% |
| 2009 | 10% | 38% | 41% | 39% |
| 2010 | 7% | 12% | 4% | 8% |
| 2011 | 10% | -4% | 3% | -1% |
| 2012 | 23% | 32% | 20% | 27% |
| 2013 | 31% | 27% | 19% | 24% |
| 2014 | 28% | 5% | -2% | 4% |
| 2015 | 2% | -9% | 4% | -4% |
| 2016 | 19% | 11% | 26% | 17% |
| 2017 | 34% | 11% | 12% | 12% |
| 2018 | 3% | 0% | -12% | -4% |
| Total | 17% | 9% | 7% | 8% |
| Year | big | medium | small | Total |
| 2006 | 7 | 337 | 239 | 583 |
| 2007 | 3 | 314 | 272 | 589 |
| 2008 | - | 264 | 336 | 600 |
| 2009 | 5 | 273 | 318 | 596 |
| 2010 | 3 | 286 | 261 | 550 |
| 2011 | 17 | 386 | 246 | 649 |
| 2012 | 20 | 363 | 217 | 600 |
| 2013 | 41 | 368 | 191 | 600 |
| 2014 | 28 | 384 | 188 | 600 |
| 2015 | 41 | 338 | 221 | 600 |
| 2016 | 34 | 334 | 232 | 600 |
| 2017 | 15 | 350 | 235 | 600 |
| 2018 | 10 | 356 | 184 | 550 |
| Total | 224 | 4,353 | 3,140 | 7,717 |
Interesting. Didn't really change that much and not that many more stocks reaching 'big", even with deflator.
My conclusion looking at this is that I still have it right. Dividend flag on (2.6% or greater) and market cap over $600m.
Dividend Greater Than 5%
The comment was also made that dividend yield >5% also is a red flag. Now my analysis may be a bit different as I include special dividends.
| Yield Range | Big | Small | Total |
| no dividend | 5.2% | 3.9% | 4.4% |
| Under 1% | 3.9% | 26.0% | 10.6% |
| 1.0% to 2.6% | 8.0% | 6.2% | 7.4% |
| 2.6% to 4% | 19.4% | 11.2% | 16.9% |
| 4% to 5% | 28.1% | 10.0% | 21.5% |
| 5% to 10% | 21.5% | 5.7% | 13.9% |
| over 10% | 33.8% | 10.9% | 19.1% |
| Grand Total | 11.1% | 5.8% | 8.4% |
Here Big and Small is different. Big means top 1/2 of stocks by market cap in any stock/month. Again, I don't see a natural split, over 5% looks just as good to me as 2.6% to 5%. But again, I include special dividends, which probably is a good chunk of stocks over 5%.
Finally, dividends between 2.5% and 5% and market cap over $3b.
| Yield Range | Over 3b | Under 3b | Total |
| no dividend | 3.4% | 4.5% | 4.4% |
| Under 1% | -0.9% | 18.1% | 10.6% |
| 1.0% to 2.6% | 4.6% | 9.6% | 7.4% |
| 2.6% to 4% | 16.4% | 17.4% | 16.9% |
| 4% to 5% | 36.1% | 13.8% | 21.5% |
| 5% to 10% | 38.8% | 7.6% | 13.9% |
| over 10% | 41.5% | 15.1% | 19.1% |
| Grand Total | 11.3% | 7.5% | 8.4% |
Again, in this table it is great to be over $3b and over 2.6% yield. But results don't drop off over 5% yield. They are actually stupendous over 5%. There can be bias in these types of raw figures. Stocks were probably much more likely to be over 5% when really sold off like in 2009.
Good stuff, keep ideas coming.
Posts
4 comments:
Hello, interesting analysis. I think the difference with my analysis is in the basic list of stocks from where your recruit. I take each month the official list and select top 50 stocks with market cap >50M (list 1), market cap >500M (list 2), market cap >1B (list 3) and market cap >3B (list 4). This gives +/- 100 unique stocks each month. As my cutt off is greater than 100M (which you apparently use), I have more stocks with bigger market cap. And those really perform very well: +17,8% in 2016 (S&P +16%); +22% in 2017 (S&P +13,5%) and +5% in 2018 (S&P +1,5%) (but here there are periods not closed yet). I take rolling periods. On average, you only need to buy 4 stocks at the beginning of each month to get these nice results and to avoid the great load of crap that smaller stocks tend to return. All results are exclusive dividends. In a 3 year period, you could have earned an additional 7 to 10% with dividends.
I agree that that is a major difference. But another difference is that you have a very short time period for your analysis.
Yes indeed, I did not yet experience a serious market turndown in my analysis time frame. That's why it will be interesting to see what happens now, if my findings still apply. However, for "running" periods 2018/01 to 2018/12 where I do not yet have full year results, I take the price as of today as preliminary closing price. While S&P500 is in the red for most of 2018 periods, the Mega cap index shows 8 green months and only 4 reds. An outperformance of 4 basis points.
Interesting stuff, isn't it? Keep up the good work, togerther we will beat the market!
Great discussion. I see similar results when I look at my returns between large cap and small cap stocks. The small cap stocks have definitely been a drag over the past few years.
j
Post a Comment