Also, TS simulation doesn't allow you to test trade-thru entry (bid-ask spread). You can try to get around it but the result is still not realistic if your profit margin is low. Testing intraday strategy is tough with TS. The thing I trade cannot be automated and requires some "art" because it is not indicator based strategy but the numbers came out much better in real life. Maybe you can test some pair trading (you can put two "something" on the same chart) or something less vanilla.
i've tried my luck at automation. i.e. programming, seeing what you see. however, i find the best results are discretionary. in fact, i would suggest, a profit factor over 10, is quite reasonable. Yoda: "...no different, only different in your mind. you must unlearn what you have learned"..."no; try-not, do, or do-not, there is no try"...
A profit factor of 10? A profit factor that high seems to have come from the same bin as the Canadian dollar came from. KH
Profit factor is one of the best trading metrics out there, however, there is at least one major drawback to the calculation that I haven't seen mentioned in this thread: Time. Peter Brandt is one of the best long term discretionary traders on the planet - his PF might be four? I'd have to check. However, he holds trades for weeks. So, going back to the OP's question, what is a reasonable PF for intraday trading? Clearly that's a different answer than for a trend follower. To illustrate the point - a HFT firm will hold trades nanoseconds - they certainly don't have PFs of 1.20. Maybe not even 1.01? The point is that they execute tens of thousands of trades a day with a tiny edge. So, the correct answer is that a good profit factor lies somewhere along a continuum. Steve Winn made billions with a profit factor of less than 1.05 at his casinos - but look how many bets his casino was taking a day. If he only took one bet a month, he probably wouldn't have Rembrandts hanging in his office.
I think the comments of TSGannGalt and Vikana earlier in the thread (back in 2006!) have it about right. I consider profit factor to be one of the three most important metrics I use to evaluate a backtested strategy, along with win rate of ~50% or better and a reasonable profit/maxDD ratio. I look at a couple of other things like $ per trade and most consecutive losing trades, but they're not nearly as important. As I've gotten better at designing new rules-based systems, I have begun to require them to give profit factors close to 2.0 or better. Real world trading is never as good as the backtests, so I feel that I need the backtests to produce results that are at least this strong. That being said, I have seen several profitable systems with profit factors of 1.50 or less, and have traded some too...
%% NOTHING like a good trend. QQQ, up past ten times from $20 area, $30 area............................................. BUT may want to do something in addition to that. Prices in USD ;more than that with dividends. NOT quite up 10 times rom starting price of $50 area