Most of you know that I cut my teeth on AAII's Stock Screens. I have been following them for several years and I love their simplicity. Moneysuckle (of the complicated poker post) has put together some information on why he has chosen just a few screens among the several that AAII shares each month. The results he discusses are assuming monthly re-balancing and holding all stocks that qualify in the screen.
Buffet averaged 25-30% a year compounded to amass his fortune. Of course, he dodged a lot of taxes with his buy and hold strategy. I cite this as a baseline to judge the AAII screens.
In order to get a total (pre-tax) return of 1000% over 10 years, you need to average 26% a year. A 1500% return over 10 years is 39% compounded per year. A 2000% 10-year return is 50% per year compounded.
So, when I look at the AAII screens, I am looking for anything with a total return of 1000% or more since 1998 - or returns Buffet would be happy with. I know, of course, that I'll likely be paying 28% cap gains tax on my profits, so I need as much cushion as possible.
The other things I am looking for in these screen results:
1. Consistency in returns. If a screen has a 1200% total return, but has 6 winning years and 4 losing years, that's not as attractive to me as a 1200% screen with 9 winning years and 1 losing year. When a screen has a big return but several losing years, it implies that it is not suitable for use in all market conditions.
2. No horrible years. I consider a horrible year a double digit loser.
3. Emphasis on recent performance. If a screen has a great total return, but 1998-1999 played a big role in it and recently it hasn't done as well, I will downgrade the screen.
4. 15 or less qualifying stocks on average. I don't want to invest in more than 20 positions at a time, and I am really interested in 6-12 positions at a time. Therefore, a screen that gives me 45 qualifying stocks on average is of marginal use. Further, if I think I can screen the screen (like with Zacks rankings), then the more stocks I am screening out with Zacks, the more the AAII total return % becomes meaningless since I am doing something completely different than AAII did. Finally, if I invest the same in each stock, if I use a screen that produces 45 stocks and 1 that produces 15, I will end up weighting the 45 screen more. Which is probably bad since Zweig and CANSLIM have the best results, and they produce fewer stocks.
With that in mind, let's examine the candidates.
Currently, Graham Enterprising Investor is at 923%. It produces 4 qualifying stocks per month. It lost money in 1998 and 1999. While the losses were small (single digit), the market had great years in 1998 and 1999. Thus, this screen vastly underperformed the market in those years. It has vastly outperformed the market since. In fact, from 2000-2007, it's total return is 1109% (which is over 50% per year compounded for 8 years!!). This tells me that the screen works well in periods when fundamentals matter. Fundamentals didn't matter much in 1998-99 - it was mania. Finally, the screen is up this year! I will use this screen for current market conditions.
Foolish Small Cap 8 Revised - 623.1%. Through 2007, this screen was very impressive at nearly 1000% 10-year total return. Plus, it had 9 winning years in 10. It's only loss was small (3.9% in 2004). Otherwise, it made at least 12% every year. Very consistent. Plus, it produces only 7 passing stocks a month on average. However, this year, it is down 35%. Until I can figure out what's going on, I am not planning on using this screen. I am going to keep an eye on it, however.
O'Shaughnessy Growth - 652%. Another very consistent screen. 10 straight winning years of at least 10% gains. Until this year (down 13%). But it produces 50 stocks per month. Not suitable for my purposes, but I will watch it.
O'Shaughnessy Small Cap Growth & Value 985% - Another impressive screen. 9 excellent years and flat in 2002. But it's losing this year (down 13%). Plus it produces 25 stocks a month. Not suitable for my purposes, but I will watch it.
O'Shaughnessy Tiny Titans - 2262%. Awesome total return. But poor recent performance. $100 invested in Jan 2005 would be $114 today. Clearly, the weakening dollar has helped the larger companies recently (who have international operations) at the expense of small domestic companies. This may explain the drop off in performance. Plus, it produces 25 stocks a month - too many for my purposes.
Value on the Move - PEG with Est Growth - 833%. Another 1000% winner with 10 straight winning years until 2008 (down 15%). After a poor 1998-99, it made 970% from 2000-2007 or 49% compounded, which is incredible. But it produces an average of 47 stocks a month. Not suitable for my current purposes.
Zweig - 2107%. By far the best, most consistent screen on AAII from 1998-2007. Never returned less than 17% a year. 10 straight winning years in all sorts of market conditions. Until this year. I am willing to forgive it 2008 (down 12%). I am also willing to forgive the "slowdown" from 2005-07 when it "only" earned 18-27% a year. However, I am keeping my eye on Zweig to make sure it doesn't go south on me. Produces 15 stocks a month on average, in my sweet spot. I will use Zweig.
CANSLIM - 1471%. Excellent total return. Very consistent. 9 winning years of at least 20% a year vs. 1 modest losing year (3.8% in 2004). All on 9 stocks a month. Down only 3% this year, which beats the market. My second favorite screen behind Zweig. I will use CANSLIM
Est Rev Up 5% - 1302%. Excellent total return, but got a huge boost in 1998-99. This screen focuses on positive earning surprises, and everyone was buying/selling on news (instead of fundamentals) back in 98-99. Still, had excellent years in 2003-07. But it produces 43 stocks a month. Plus, when I cross-screen Zweig and CANSLIM through Zacks, I get stocks with good earnings surprises. As you might imagine, more of the Est Rev up 5% stocks get 1 ranks in Zacks than Zweig stocks do. So, if I used this screen, I'd be heavily diluting Zweig. I will watch it, but I don't plan to use it.
So, there's my analysis of the screens. I will focus on Zweig, CANSLIM and Graham Enterprising, which should produce a combined ~30 stocks a month. Edited for Zacks 1-2 ranks, that will bring it down to a manageable 6-12 stocks a month.
I plan to fully invest my portfolio each week, and invest in each stock equally. So, at 1% risk, 12 stocks means I need to put a 12% stop loss on, and 6 stocks means I need to put a 6% stop loss on. I saw your 2007 stats. For Zacks 1 + Zweig, your average loss as -3.2% on your losing trades. Therefore, I'd think a 6% stop loss is way sufficient. Unless you had a lot of experience with stocks losing more than 6% off the bat and coming back to be winners. Which I doubt.
Theses are the types of stats that I love to gather for myself. I think that thinking through issues like these often make you a much better trader. I know it has helped me put together my own ideas, and has helped me tremendously during periods of drawdowns.