What I've been learning lately
My quant system continues to work for me, I'm beating the market for the year (real life trades) though I've been lagging a little the last few weeks. There wil always be periods of lagging with any system, sometimes these periods can persist and that's when most traders (and fund managers) blow it, they give up or change a system that would be effective over the long-term if they stuck to it but it's really really hard to stick to it if you lag for a year or two. Nevertheless, 60% is the best anyone can hope for and a 60% success rate means that with truly random distribution there will be long periods that lag. A lifetime 60% free throw shooter will have LONG stretches in his career where he will miss more than half of his shots.
But that's not my point. Timing has always been my biggest challenge, especially when to sell. My success rate at selling is probably a bit above 50%, slightly better than throwing darts. I rarely sell too late, but I often sell too soon.
I was thinking about the many successful systems that tell you to buy a stock that meets their criteria, hold it for exactly one year and sell. Totally arbitrary perhaps, but it seems to work out pretty well. But I think one year is too long today. A lot of my sells DO tank long before a year is up.
My average holding period has been about 13 weeks, my analysis of my sells convinced me that this is a bit too short. Too many stocks have been flat for the 2-4 months I've held them and then popped up nicely right after I let them go. (Extreme example: I held HII for months, it did NOTHING, I sold it on 8/12 and it's up over 8% in less than a week!)
So I'm going to go with a longer time frame and I'm going to take some of the subjectivity out of selling.
My new time frame (real life, because of CAPS scoring quirks I won't use it here) will be 20 weeks. I may tweek later, but I'm pretty sure that 13 is too short and 30 is too long. I'll hold more stocks (40 vs 26) in smaller dollar amounts and I'm going to be much more systematic and disciplined about selling the two lowest scoring stocks (no more and no less) each week and replacing them with two others. That will make a 20 week average holding time though stocks though I will hold some much longer or shorter. (I also hold a number of etf's long-term).
We'll see how it goes. The system remains the same, I'm just working on timing issues.
And my system has never worked well for ADRs (I've never been able to consistently outperform VT for foreign stocks). I've never held very many (I have several international and em funds) but I'll probably just give up on them almost completely and stick with funds for the foreign portion of my portfolio. I like FNDE, FNDF and VINEX (expense ratios matter!).
Meanwhile, though I screen for a multitude of factors, P:FCF remains my most effectively predictive single screen. While momentum is also a major factor, with currently high valuations I have been generally avoiding stocks with high EV:EBITDA ratios.
Recent buys include: AYR, HIHO, BX, INTC, LUV