Ed Seykota recently made an interesting post on his site:
"A [MAR ratio] of 1.0 is pretty good, better by far than the one you get by buy-and-hold on S&P. People who claim much larger ones such as 2.0 and up don't usually make that claim for long."
This comes from a guy who has been trend following, professionally, since the 1970s. His opinion would also seem to be supported by the results reported at the iasg.com site.
Do the real-time results of TB users bear this out? Or is Ed's assessment more accurate? For what it is worth, my personal results are far closer to Ed's, though currently I do not trade a fully diversified portfolio.
What is Pretty Good to You?
There was a post a few years ago by c.f. where someone asked him "what is an extraordinary trader?" or something like that, and he replied. I searched for the post just now and can't find it, perhaps it was in the section of the Forum devoted to the now-obsolete product Veritrader. The Veritrader areas of the site are gone.
He listed several criteria and what I found extremely interesting was the order of the listing. The first item that he listed was "greater than 70% winning months" which is a criterion that very few gurus talk about. I've forgotten the others, I suppose I found them unremarkable, but this criterion is still surprising and striking, to me. Very few of my mechanical systems are able to sustain 70% winning months for backtests of 20 years or longer. In my experience testing mechanical systems, the ones that do exceed 70% winning months, really are "extraordinary" as he said.
No surprise that he and Seykota focus on different criteria!
He listed several criteria and what I found extremely interesting was the order of the listing. The first item that he listed was "greater than 70% winning months" which is a criterion that very few gurus talk about. I've forgotten the others, I suppose I found them unremarkable, but this criterion is still surprising and striking, to me. Very few of my mechanical systems are able to sustain 70% winning months for backtests of 20 years or longer. In my experience testing mechanical systems, the ones that do exceed 70% winning months, really are "extraordinary" as he said.
No surprise that he and Seykota focus on different criteria!