Yes, it is that cold outside. Another month/year has flown by. And so it is time to look at another monthly tracking portfolio as we hit the end of December 2017. As my faithful readers know, I have been tracking the Magic Formula Stocks as described by Joel Greenblatt in The Little Book That Beats the Stock Market since January 2006. Every month I take the top 50 stocks over $100m market cap from his website and track how that portfolio of stocks fares versus the Russell 3000 for the next 12 months. It has been an uphill struggle as the tracking portfolios have under-performed, driven in part by Chinese reverse merger fiascos, for-profit education stocks and home health care stocks all being proverbial albatrosses.
This also marks the end of a calendar year, so I will also ramble on a bit about my MFI Index, which tracks how MFI has done year by year.
I also encourage new readers to read
Guide To New Readers
MFI Index
Might as well start with this. My MFI Index is a group of 50 MFI stocks that gets reconstituted once a year (like this weekend). It largely runs in parallel with my end of December monthly tracking portfolios. I do treat dividends slightly differently and I do everything manually, so we should not get tripped up by a Yahoo Finance bug... instead it would be my mistake. Here is a table of how my MFI Index has done each year since 2006 compared against the Russell 3000:
| Annual | Inception to Date | |||
| Year | Russell | MFI | Russell ITD | MFI ITD |
| 2006 | 11.40% | 15.03% | 11.40% | 15.03% |
| 2007 | 4.09% | -6.69% | 15.96% | 7.34% |
| 2008 | -37.05% | -37.97% | -27.00% | -33.42% |
| 2009 | 32.51% | 45.18% | -3.27% | -3.34% |
| 2010 | 18.38% | 22.77% | 14.50% | 18.67% |
| 2011 | -0.56% | -10.47% | 13.87% | 6.25% |
| 2012 | 16.43% | 9.70% | 32.57% | 16.56% |
| 2013 | 33.01% | 51.70% | 76.34% | 76.82% |
| 2014 | 12.26% | 12.07% | 97.95% | 98.15% |
| 2015 | 0.38% | -8.95% | 98.33% | 89.37% |
| 2016 | 12.50% | 13.19% | 110.62% | 101.16% |
| 2017 | 20.82% | 4.74% | 133.65% | 105.95% |
So once can see we have 12 years under our belts. MFI has only won 5 of the 12 years (recall my Index is top 50 stocks over $100m). In aggregate, the Russell is up 133.6% against MFI at 106%. You'll see similar numbers in my monthly tracking portfolios later.
People ask, does MFI not work anymore? I am not sure. Perhaps 12 years isn't long enough. Perhaps it works better at a higher market cap cut off (I believe this to be true). I think too many stocks make the screen over the years by "mistake"... that is they are not true value/cheap stocks. In 2011, we had a bunch of Chinese Reverse Merger stocks that hurt that year and were out and out frauds. Then we have also had a number of stocks that have a one time spike in earnings that fools the formula. Frankly, it might do better if instead of a trailing 12 month income look, it used a trailing 24 month look on income, to filter out the one time spikes.
But even with that being said, Value stocks have not exactly been in favor the past couple of years.
Onward
Fearless readers will recall we had a 29 month "losing streak".. Then we had a 13 month winning streak. This month, MFI lost to the benchmark by almost 16 percentage points.
I went back and looked and the 15.8% shellacking the tracking portfolio took was the worst of all my closed portfolios. Un-fricking-believable.
And 9 of the 11 open portfolios are trailing, by about 7 points on average, Pardon me while I puke.
Here are the 50 stocks in descending performance order (from the portfolio that just hit a year):
| Stock | Initial Price | End Price | Percent Change | Mkt Cap |
| MYGN | 16.67 | 34.35 | 106.0% | 1,140 |
| ACHI | 2.25 | 4.41 | 96.0% | 240 |
| MBUU | 19.08 | 29.73 | 55.8% | 338 |
| MCFT | 14.58 | 22.22 | 52.4% | 272 |
| MDCA | 6.55 | 9.75 | 48.9% | 345 |
| AAPL | 113.99 | 169.23 | 48.5% | 618,035 |
| HPQ | 14.42 | 21.01 | 45.7% | 25,406 |
| TDC | 27.17 | 38.46 | 41.6% | 3,524 |
| IRMD | 11.10 | 15.15 | 36.5% | 119 |
| SCMP | 13.55 | 17.95 | 32.5% | 621 |
| CSCO | 29.18 | 38.30 | 31.3% | 151,825 |
| VEC | 23.85 | 30.85 | 29.4% | 256 |
| PDLI | 2.12 | 2.74 | 29.2% | 351 |
| TRNC | 13.87 | 17.59 | 26.8% | 505 |
| RPXC | 10.76 | 13.44 | 24.9% | 531 |
| AVID | 4.40 | 5.39 | 22.5% | 179 |
| SSNI | 13.31 | 16.24 | 22.0% | 691 |
| USNA | 61.20 | 74.05 | 21.0% | 1,490 |
| HRB | 22.18 | 26.22 | 18.2% | 4,862 |
| SYNT | 19.79 | 22.99 | 16.2% | 1,664 |
| MCK | 139.30 | 155.95 | 11.9% | 31,742 |
| TGNA | 21.39 | 23.69 | 10.8% | 4,586 |
| GHC | 507.37 | 558.35 | 10.0% | 2,874 |
| DLX | 70.39 | 76.84 | 9.2% | 3,479 |
| RGR | 51.41 | 55.85 | 8.6% | 1,000 |
| ESRX | 68.79 | 74.64 | 8.5% | 43,302 |
| CA | 30.78 | 33.28 | 8.1% | 13,126 |
| AMCX | 52.34 | 54.08 | 3.3% | 3,702 |
| UTHR | 143.43 | 147.95 | 3.2% | 6,520 |
| GILD | 69.58 | 71.64 | 3.0% | 94,668 |
| BBSI | 63.01 | 64.49 | 2.4% | 464 |
| NSR | 33.40 | 33.50 | 0.3% | 1,823 |
| BPT | 20.43 | 20.10 | -1.6% | 508 |
| MSGN | 21.50 | 20.25 | -5.8% | 1,614 |
| MPAA | 26.92 | 24.99 | -7.2% | 532 |
| VIAB | 34.26 | 30.81 | -10.1% | 13,928 |
| IDCC | 90.30 | 76.15 | -15.7% | 3,132 |
| PBI | 14.30 | 11.18 | -21.8% | 2,820 |
| GME | 23.51 | 17.95 | -23.7% | 2,592 |
| AWRE | 6.10 | 4.50 | -26.2% | 138 |
| NHTC | 23.31 | 15.19 | -34.8% | 280 |
| AGX | 69.50 | 45.00 | -35.3% | 1,078 |
| SWHC | 21.08 | 12.84 | -39.1% | 1,186 |
| LFVN | 8.15 | 4.76 | -41.6% | 115 |
| UIS | 14.95 | 8.15 | -45.5% | 749 |
| FPRX | 50.11 | 21.92 | -56.3% | 1,425 |
| AGTC | 9.35 | 3.60 | -61.5% | 169 |
| CXRX | 2.12 | 0.66 | -69.0% | 108 |
| DHX | 6.25 | 1.90 | -69.6% | 313 |
| ICON | 9.34 | 1.29 | -86.2% | 525 |
I do go on and on about the importance of avoiding "stinker" stocks, those that drop over 30%. 10 of them this year. For a year when the markets went up 20%, to have 10 of 50 stocks DROP by more than 30% is a problem.
Here is a listing of every portfolio I have tracked:
| Date | MFI | R3K | Lead |
| 1/6/2006 | 16.0% | 10.9% | 1 |
| 2/17/2006 | 21.2% | 14.6% | 1 |
| 3/29/2006 | 13.0% | 9.6% | 1 |
| 4/7/2006 | 10.3% | 12.1% | 0 |
| 5/12/2006 | 20.4% | 18.6% | 1 |
| 5/31/2006 | 29.2% | 23.3% | 1 |
| 6/30/2006 | 22.4% | 20.0% | 1 |
| 7/31/2006 | 19.7% | 17.3% | 1 |
| 8/31/2006 | 13.0% | 13.3% | 0 |
| 9/28/2006 | 12.7% | 14.6% | 0 |
| 10/27/2006 | 10.3% | 12.0% | 0 |
| 11/29/2006 | -0.3% | 4.8% | 0 |
| 12/28/2006 | -6.9% | 3.4% | 0 |
| 1/26/2007 | -10.2% | -6.6% | 0 |
| 2/27/2007 | -3.7% | -1.0% | 0 |
| 3/26/2007 | -9.8% | -5.5% | 0 |
| 4/27/2007 | -10.9% | -5.0% | 0 |
| 5/29/2007 | -11.5% | -6.3% | 0 |
| 7/3/2007 | -30.0% | -15.6% | 0 |
| 7/30/2007 | -19.9% | -11.5% | 0 |
| 8/30/2007 | -12.5% | -8.7% | 0 |
| 9/27/2007 | -19.0% | -18.2% | 0 |
| 11/2/2007 | -40.4% | -34.3% | 0 |
| 11/28/2007 | -40.1% | -38.3% | 0 |
| 12/28/2007 | -36.3% | -40.0% | 1 |
| 1/25/2008 | -36.4% | -35.9% | 0 |
| 2/26/2008 | -51.7% | -41.5% | 0 |
| 3/24/2008 | -40.9% | -36.8% | 0 |
| 4/25/2008 | -25.6% | -31.0% | 1 |
| 5/28/2008 | -22.2% | -33.6% | 1 |
| 7/2/2008 | -11.7% | -25.3% | 1 |
| 7/29/2008 | -10.5% | -20.9% | 1 |
| 8/29/2008 | -13.8% | -17.9% | 1 |
| 9/26/2008 | -4.3% | -10.0% | 1 |
| 10/31/2008 | 18.7% | 13.9% | 1 |
| 11/26/2008 | 50.9% | 27.7% | 1 |
| 12/26/2008 | 48.9% | 32.3% | 1 |
| 1/23/2009 | 59.3% | 36.4% | 1 |
| 2/27/2009 | 92.8% | 55.6% | 1 |
| 3/27/2009 | 85.8% | 48.1% | 1 |
| 4/24/2009 | 69.7% | 45.8% | 1 |
| 5/29/2009 | 31.8% | 22.8% | 1 |
| 6/29/2009 | 21.3% | 24.0% | 0 |
| 7/29/2009 | 19.5% | 15.9% | 1 |
| 8/28/2009 | 7.4% | 8.8% | 0 |
| 9/25/2009 | 12.6% | 12.4% | 1 |
| 10/30/2009 | 22.7% | 18.3% | 1 |
| 11/27/2009 | 24.3% | 13.6% | 1 |
| 12/31/2009 | 23.7% | 18.1% | 1 |
| 1/22/2010 | 19.0% | 20.6% | 0 |
| 2/26/2010 | 18.6% | 23.6% | 0 |
| 3/25/2010 | 10.0% | 15.4% | 0 |
| 4/23/2010 | 7.1% | 11.4% | 0 |
| 5/28/2010 | 19.3% | 25.4% | 0 |
| 6/29/2010 | 16.7% | 25.7% | 0 |
| 7/29/2010 | 5.4% | 20.1% | 0 |
| 9/2/2010 | 7.3% | 10.1% | 0 |
| 9/24/2010 | -4.3% | 0.3% | 0 |
| 10/29/2010 | -2.9% | 10.4% | 0 |
| 11/26/2010 | -8.5% | 1.4% | 0 |
| 1/3/2011 | -11.4% | 0.1% | 0 |
| 1/28/2011 | -7.6% | 4.9% | 0 |
| 2/25/2011 | -5.5% | 5.0% | 0 |
| 3/24/2011 | -4.4% | 7.4% | 0 |
| 4/21/2011 | -16.0% | 3.2% | 0 |
| 5/27/2011 | -12.0% | -0.4% | 0 |
| 6/24/2011 | -9.5% | 5.0% | 0 |
| 7/29/2011 | -4.3% | 8.1% | 0 |
| 8/26/2011 | 12.0% | 21.6% | 0 |
| 9/30/2011 | 23.5% | 29.6% | 0 |
| 10/28/2011 | 0.4% | 11.5% | 0 |
| 11/25/2011 | 13.9% | 24.0% | 0 |
| 12/29/2011 | 9.9% | 15.9% | 0 |
| 1/27/2012 | 7.4% | 16.6% | 0 |
| 2/24/2012 | 7.8% | 13.3% | 0 |
| 3/23/2012 | 9.1% | 15.2% | 0 |
| 4/27/2012 | 10.5% | 15.3% | 0 |
| 5/25/2012 | 23.4% | 27.5% | 0 |
| 6/22/2012 | 26.1% | 24.6% | 1 |
| 7/27/2012 | 30.4% | 25.7% | 1 |
| 8/24/2012 | 26.0% | 19.8% | 1 |
| 9/28/2012 | 39.3% | 22.1% | 1 |
| 10/26/2012 | 48.3% | 28.9% | 1 |
| 11/23/2012 | 45.7% | 32.1% | 1 |
| 12/31/2012 | 52.3% | 33.0% | 1 |
| 1/25/2013 | 39.5% | 22.0% | 1 |
| 2/22/2013 | 46.5% | 26.4% | 1 |
| 3/28/2013 | 40.1% | 21.2% | 1 |
| 4/26/2013 | 35.0% | 20.5% | 1 |
| 5/24/2013 | 20.4% | 19.2% | 1 |
| 6/28/2013 | 26.9% | 24.1% | 1 |
| 7/29/2013 | 17.8% | 19.0% | 0 |
| 8/30/2013 | 17.5% | 24.6% | 0 |
| 9/27/2013 | 12.9% | 17.7% | 0 |
| 10/25/2013 | 11.8% | 15.4% | 0 |
| 11/29/2013 | 12.3% | 15.6% | 0 |
| 12/31/2013 | 12.7% | 11.9% | 1 |
| 1/31/2014 | 14.2% | 13.0% | 1 |
| 2/28/2014 | 15.4% | 13.9% | 1 |
| 3/28/2014 | 6.1% | 12.3% | 0 |
| 4/25/2014 | 8.6% | 15.8% | 0 |
| 5/30/2014 | 5.9% | 11.7% | 0 |
| 6/27/2014 | 2.5% | 9.2% | 0 |
| 7/25/2014 | 3.3% | 8.7% | 0 |
| 8/29/2014 | -2.2% | 1.0% | 0 |
| 9/26/2014 | -0.6% | -0.9% | 1 |
| 10/31/2014 | 4.3% | 4.4% | 0 |
| 11/28/2014 | -2.9% | 2.9% | 0 |
| 12/31/2014 | -9.3% | 0.3% | 0 |
| 1/30/2015 | -12.4% | -2.8% | 0 |
| 2/27/2015 | -19.0% | -7.3% | 0 |
| 3/27/2015 | -17.1% | -1.9% | 0 |
| 4/24/2015 | -15.4% | -0.8% | 0 |
| 5/29/2015 | -14.1% | 0.1% | 0 |
| 6/26/2015 | -8.7% | 0.4% | 0 |
| 7/30/2015 | 3.96% | 4.02% | 0 |
| 8/28/2015 | 5.6% | 10.2% | 0 |
| 9/25/2015 | 6.5% | 13.8% | 0 |
| 10/30/2015 | 2.0% | 4.0% | 0 |
| 11/27/2015 | 8.9% | 8.4% | 1 |
| 12/31/2015 | 13.7% | 12.6% | 1 |
| 1/29/2016 | 22.6% | 22.4% | 1 |
| 2/26/2016 | 27.8% | 25.4% | 1 |
| 3/24/2016 | 26.5% | 19.6% | 1 |
| 4/22/2016 | 13.7% | 14.8% | 0 |
| 5/27/2016 | 14.3% | 15.6% | 0 |
| 6/30/2016 | 23.2% | 16.1% | 1 |
| 7/29/2016 | 14.4% | 16.0% | 0 |
| 8/26/2016 | 10.6% | 14.6% | 0 |
| 9/30/2016 | 18.6% | 18.5% | 1 |
| 10/28/2016 | 18.3% | 24.1% | 0 |
| 11/25/2016 | 7.4% | 19.2% | 0 |
| 12/30/2016 | 4.7% | 20.5% | 0 |
| 1/27/2017 | 4.1% | 17.6% | 0 |
| 2/24/2017 | 4.0% | 13.9% | 0 |
| 3/31/2017 | 2.2% | 14.0% | 0 |
| 4/28/2017 | -1.5% | 12.9% | 0 |
| 5/26/2017 | 2.8% | 11.5% | 0 |
| 6/29/2017 | 2.1% | 10.9% | 0 |
| 7/28/2017 | 2.7% | 8.6% | 0 |
| 8/25/2017 | 9.6% | 9.9% | 0 |
| 9/29/2017 | 0.1% | 6.0% | 0 |
| 10/27/2017 | 5.5% | 3.5% | 1 |
| 11/24/2017 | 3.6% | 2.3% | 1 |
| Open 11 | 3.2% | 10.1% |
Wow, April 28th is actually negative. Probably explains why my May 6th real money tranche is struggling. That was just a bad time to start.
Dividend Subset
I also have historically tracked 3 subsets of the larger portfolios: (1) dividend stocks (those with a yield per Yahoo of at least 2.6%), (2) new stocks (those new to the tracking portfolios in past 12 months and (3) Dogs. It is like the dogs of the Dow in that you take the worst five performing stocks from the prior year that renew or all those that dropped more that 29%.
I have officially stopping the dogs and new subsets. It takes time and they have never really shown any advantage. The dividend approach has proven merit. In this portfolio, the dividend subset did better than standard MFI. (up 12.3%). And the 11 open dividend portfolios are up 9.6%, while the open portfolios (all stocks) are up a meager 3.2%. That implies that for all my open portfolios, stocks that do not pay a dividend of at least 2.6% are up just 1.1%. Hopefully readers can see why I strongly prefer stocks on the MFI list that pay a dividend. (in my view, by the way, it is not the fact that they pay a dividend that is important. It is the fact that stocks that masquerade as MFI stocks via a one time/temporary earnings boost generally don't pay dividends... so we weed those out).
I also have historically tracked 3 subsets of the larger portfolios: (1) dividend stocks (those with a yield per Yahoo of at least 2.6%), (2) new stocks (those new to the tracking portfolios in past 12 months and (3) Dogs. It is like the dogs of the Dow in that you take the worst five performing stocks from the prior year that renew or all those that dropped more that 29%.
I have officially stopping the dogs and new subsets. It takes time and they have never really shown any advantage. The dividend approach has proven merit. In this portfolio, the dividend subset did better than standard MFI. (up 12.3%). And the 11 open dividend portfolios are up 9.6%, while the open portfolios (all stocks) are up a meager 3.2%. That implies that for all my open portfolios, stocks that do not pay a dividend of at least 2.6% are up just 1.1%. Hopefully readers can see why I strongly prefer stocks on the MFI list that pay a dividend. (in my view, by the way, it is not the fact that they pay a dividend that is important. It is the fact that stocks that masquerade as MFI stocks via a one time/temporary earnings boost generally don't pay dividends... so we weed those out).
Finally, I like to track cash as it is more "honest" than percentages as if you go up by 25% and then down by 25%, that is a different result than up by 5% then down by 5%. So I track what you would have today if you had spread $100,000 over the first twelve portfolios evenly. I think it is obvious why my formula approach is a function of dividend subset.
| Category | Value |
| Total | 209,354 |
| Total Russell 3K | 260,329 |
| Dividend | 485,345 |
Ok. Hopefully people found this almost interesting. Snowing here again. Luckily not at NH Condo, where it was -15 this morning. Yikes! Just 3 degrees here is morning.
Posts
3 comments:
Where in NH is your condo? I grew up in Southern NH but am now living in NC. I miss the skiing but thats about it :)
Just a cynical thought about MFI if I may. What if there was never really an anomaly there?
What if Greenblatt looked at a rather limited data set, found a model that fit which showed excess returns during that period, published a book and made millions off of us suckers? I am not implying Greenblat did this on purpose, rather, he may have "discovered" something that really was never there to begin with and we all got too excited about it. I always find it suspect when these amazing strategies work great almost every year within the in-sample data and then as soon as we start applying them to out-of sample data you have to ask if its just really bad luck for the past 10+ years or maybe something else is going like bad data analysis to begin with.
On a side note, have you read What Works on Wall St? I would be happy to send you my copy if you want to borrow it. I really enjoyed it and some of MFI like strategies were tested and seemed to do well (in-sample). My favorite idea is combing value with momentum where you essentially screen for value stocks but only buy ones going up over the past 6-12 months. This should eliminate some of the really bad stocks that just seem to go down from the start.
-Brad
Brad - my condo is in the White Mountains, Jackson. Just gorgeous there - like a cheaper Colorado. I am thinking about going next week.
MFI is working for me. I don't think Greenblatt gamed the system. He had an unbelievable run as a hedge fund manager and doesn't need our millions. I just think it needs to be tweaked a bit as the universe has a bunch more one hit bio techs (like MNKD) that make the screen.
I will check out the Wall Street book. Thanks for offer to sned, but I can grab a copy easily.
M
I see that Dividend > 2.6% has performed well, but do you happen to have any data or even have a sense how Dividend > 0 would have performed? Love the blog - been a long-time follower.
Post a Comment