Donchian Channels automatically produce time-based stops. But I like the concept opposite of tightening stops as a trade goes deeper into profits, I would prefer to widen my stop if I were to make such changes intratrade. Suppose you enter with a 4 week channel stop. After that stop breaks even you could let the ten week channel catch up and then move the stop with it. This gives you a chance to let the trade run longer and deeper into profits. “Fast and furious with the OTE.” That is what Jerry Parker likes to say. That was the thinking of the Turtles. But they and he actually use a different approach - System Diversification. Because I can get wicked good granularity with an Oanda account, I have moved on to that approach also.
-Adrian
Drinking Rolling Rocks tonight. You know, for a cheap beer, they aren’t that bad.
you are right concerning the time-based stops…i just thought of time-based stops as closing the trades after a certain amount of time (1 hour, 1 day, 1 week, etc…)…but of course, a trailing stop is also time-based…
well, one could argue that this way you also give the trade a chance to take your profits away again…
may i ask you to run a backtest for DC 20 (= entries) with a trailing stop of 2x ATR…without manually trailing the stops on DC 10, as i did up to now…if this is profitable we could take great part of the manual trading out of the game (only entries would still be placed manually)…
Still thinking about these trailing stops. Initially there looks to be no advantage stepping the periods down (12 to 9 to 6) to lock in profits. However Adrian’s take of stepping out the periods (10 to 20) does have merits. Particularly after a swing low. See what we can come up with.
I have decided to make a tweak to my implementation.
I have opened separate accounts with my dealers in which to trade different systems. This way I am not trading multiple channels in the same pair on the same account. So I can have an account that only trades the four week channel, another that trades the ten week, and another that trades the twenty week. This will clean up any troubles that can result from the first-in-first-out rule and avoid any confusion about which stop goes with what channel.
The basic idea of an antimartingale strategy is that you risk more as you win and less as you lose. Allowing the stop to widen as the trade progresses in your favor is in line with that logic. While it will in many cases cause you to give up potential profits, it will also in some cases keep you in a trade through a counter-trend move and take you on to higher profits.
All that said, I like simply trading multiple systems to cover all bases. And with the spectacular granularity I can get in Oanda accounts that is simply too attractive. So I am not stepping out to longer channels intratrade anymore.
With respect to giving up profits, I made this meme a while back and I just can’t get enough of it:
The USD/CHF pair has been up to sideways since late April but has not broken any new highs in the four, ten, or twenty week ranges. Although it did touch over the 200 day moving average a bit last week only to remain below it this week. With that, I am still short on all three ranges. Enough time has now passed to get the January 15 week out of the twenty week range so all three ranges are now more similar to normal conditions.
Since GBP/JPY went back north of the 200 day moving average in mid-April it has shown further upward movement to the point of my stop on the four week channel has passed the point of break even. This trade is currently a major component of my positive OTE.
Gday Mike. The results are in. How we interpret them is up to how each of us trade.
First the bot used to back test. Rules very simple. If no order exist, open limit stops orders at the upper and lower 20 period channel. OCO. Stops set at 2 x ATR period 20. Trailing stop active immediately and trails at 2 x ATR period 20 at last candle close. Risk at 1% to match other tests.
Test period May 2005 to May 2015. Same 10 year period. Lets go with the EURUSD first.
Na fairly easy even for me. Remember, the reason we test is to gather statistics on the past and use this data to build our mechanical systems. Having the knowledge that our system has whether the storm of time gives us the confidence to stick to and trade by the rules However a well put together bot can have the ability to trade when we can’t.
I just switched to ctrader about 2 months ago, I really like it the only thing I miss is the b clock plus, I think thats what its called where it gives me the the margin requirement for lot and mini lots and its right on the screen.
Yes. Of interest for is the fact on the hourly charts a trailing stop based on ATR seems to be the best option. Yet on the 4 hour I use fib levels and a fixed R:R then out on the daily the turtles have it. Lots of work to do over the next year. Or two, maybe three, quite possibly four…
