As a a person with a degree in mathematics, I always enjoy watching movies where they are in a professor's classroom and they want to put some impressive formulae on the blackboard. Always a lot of greek notation, integrals (which always look cryptic) and exponents.
I have been involved with MFI now for almost seven years. Stunning! While I have enjoyed keeping this blog, I can honestly say I have not had much of a financial payout from TLBTBTSM. Hard to say if we have just been in an unusual bad streak or whether the formula as published doesn't work.
I have kept a data base that is essentially what JG does in his book where he back-tested for 17 years. My database has been real time (not back testing) where I track 50 MFI stocks every month and hold for a year. Now this is not a full-time job for me and I do not have a classroom of MBA students who can help with some of the heavy lifting. So I am sure my analysis is not perfect, though I do believe it is directionally correct. The biggest risk in my analysis is when something material happens to a company during the one year holding period, such as a spin-off, extraordinary dividend or a split. I do look at stocks that perform poorly to try to spot such instances, but it is highly probably I have missed a few (my approach with Yahoo would catch the extraordinary dividend, assuming they catch it).
Results by Market Cap Percentile
Overall, the average MFI stock has gone up 5.2% per year. In the same time period, the Russell 3000 has arithmetically averaged 6.1% per year. I took all the stocks in my list and grouped them by market cap decile. Now I have done this on a raw basis (as I am a bit lazy). The appropriate way to do it would be to somehow inflation-adjust it as over time you would expect market caps to increase.
| Percentile | Average of Percent Change | Min of Mkt Cap | Max of Mkt Cap2 |
| 1 | 9.2% | 6,012 | 262,943 |
| 2 | 5.7% | 2,660 | 6,953 |
| 3 | 6.4% | 1,419 | 2,870 |
| 4 | 9.2% | 947 | 1,573 |
| 5 | 3.0% | 696 | 1,021 |
| 6 | 2.9% | 484 | 781 |
| 7 | 2.8% | 349 | 558 |
| 8 | 4.3% | 257 | 397 |
| 9 | 6.4% | 169 | 289 |
| 10 | 3.0% | 25 | 204 |
| Grand Total | 5.2% | 25 | 262,943 |
So the table is read in that a "1" are the largest 10% of market cap stocks in my data, running from 6b to Microsoft or Apple. The obvious conclusion looking at this table is that if you only picked from the top 50 greater than 100m list (my starting point) and then went with market cap > 947m, you would juice your return to 7.6%. This why in my re-boot of MFI in August of 2012, I am focusing on $500m+ stocks.
Results by Country
As readers/followers know, China has been problematic for MFI. I have attempted to pull in stock by country. Again, as I had thousands of entries and some stocks that do not exist any more, I did the best I could.
| Row Labels | Average of Percent Change | Count of Stock |
| Bermuda | 1.3% | 1 |
| British Virgin Islands | 26.6% | 2 |
| Canada | 10.1% | 176 |
| Cayman Islands | 54.2% | 26 |
| China | -28.1% | 104 |
| Ireland | -1.4% | 28 |
| Israel | 22.2% | 30 |
| Netherlands | 42.2% | 18 |
| Russia | 152.5% | 9 |
| South Africa | -14.1% | 14 |
| Switzerland | 16.6% | 18 |
| Taiwan | -27.5% | 24 |
| United Kingdom | 12.7% | 8 |
| USA | 5.1% | 3,709 |
| Grand Total | 5.2% | 4,167 |
China has been very poor. However, it is (to me) very interesting to note that if you juts went with US stocks, you'd actually do worse than average. So the maximum approach would be to exclude China, Taiwan and South Africa. That would take you to 6.3%, which is frankly just average. So my conclusion is that country code is not a great way to differentiate.
By First Letter
Obviously, it is silly to think the first letter is an predictor of results. But here it is for grins:
| Row Labels | Average of Percent Change | Count of Stock |
| A | -3.3% | 315 |
| B | 5.0% | 120 |
| C | 2.8% | 425 |
| D | 23.3% | 196 |
| E | 5.7% | 219 |
| F | 16.7% | 155 |
| G | 5.0% | 175 |
| H | 6.0% | 169 |
| I | 8.2% | 152 |
| J | -6.3% | 69 |
| K | 2.2% | 163 |
| L | -0.9% | 124 |
| M | 11.3% | 260 |
| N | 4.9% | 153 |
| o | -6.4% | 60 |
| P | 4.1% | 351 |
| Q | 61.3% | 33 |
| R | 0.8% | 101 |
| S | -0.7% | 207 |
| T | 10.7% | 221 |
| U | -0.4% | 214 |
| V | 4.6% | 209 |
| W | 0.1% | 55 |
| X | -29.8% | 14 |
| Y | -97.2% | 1 |
| Z | 15.2% | 6 |
| Grand Total | 5.2% | 4,167 |
D has been a great pick!
I will look at letters in ticker and by dividend yield in my next post! Cheers.
Posts
No comments:
Post a Comment