I'm not sure I'd call it a widows and orphans fund. It's based on Jim O'Shaughnessy's What Works On Wall Street, which tested the Compustat database back to 1954 to find the best-performing attributes (low price to sales and rising relative strength were two that I recall), then holds the most promising 50 for the year ahead. It returned about 18% annually over all rolling 10-year periods, but I do recall some bad years, with the worst year an 18% loss, as far as I can remember. I can't find the book offhand; buried in a box in the basement somewhere, but I do recall it having some volatile periods.
I think that and Ken Lee's Trouncing the Dow are the two best FA-based approaches I've seen. Lee's has done maybe 30% a year since 1973, but the problem with that one is it holds as few as 1-3 stocks at a time; that's a little too much concentration for my nerves. Still, it's had a total return of about 73% since 2000, versus about 45% for HFCGX, so I keep looking for ways to time it. Weekly stochastic crossovers have been the most promising so far, but even they wind up lagging a lot of years.
On the third quarter, I think since 1986, August has been down on average maybe 0.12% on the S&P, September down 2.6%, and October up 2%+, so September really is the month to sit out (all those stats are from memory). Did not know about the fourth quarter when the third is up; will have to watch that.
Paul |
