That really looks like something worthwhile to pursue! I basically do the same manually by entering at big moves. Have you tested the same with the ES? My opinion is that the ES reveals trend patterns that are otherwise hidden in the SPY, because it does not trade around the clock. Perhaps your algo will be able to catch moves even more precisely.
ES Trend Following
- Thread starter Bombardier
- Start date
- Status
That really looks like something worthwhile to pursue! I basically do the same manually by entering at big moves. Have you tested the same with the ES? My opinion is that the ES reveals trend patterns that are otherwise hidden in the SPY, because it does not trade around the clock. Perhaps your algo will be able to catch moves even more precisely.
Quote from Wide Tailz:
Yes, perhaps. One of the problems with Dow theory is its inability to know if you are seeing a real trend change or an in-trend correction. Your question exposes this problem.
I don't know the answer.
I eventually gave up on TA and just coded a very primitive trend following algo. It requires huge volatility to take a position. It went short the day after you got stopped out. It caught part of the down move and the subsequent up move.
How many years back have you backtested this algo?
>>10
Many of the entries in this system require an overnight move, so using on 23.5 hour ES data would give a different result. It was developed on SPY. I've thought about that problem and would probably just redesign it for ES, if that were my instrument of choice.
The reason I post is because to duplicate this, you would need to guess 6-7 parameters. If the market changes and this stops working, I can re-optimize and keep going. I actually have a 1 year curve fit version which returns about 40% more than this, but it was flat every year until 2010, when it took off.
The idea behind this is that there is always signal in the market and I can tune my radio dial (the system parameters) to pick it up.
I also believe all systems are curve fits and eventually stop working.
The Sperandeo Trend Method annotated on the recent price chart.
The rules:
1. Trend line under lowest lows from start to present, without going through any
2. Lows that touch trend line are stop points (similar to the Bombardier Method)
3. Once stopped out, reverse signal by first lower high
The same problem with Dow Theory shows up here: the occasional whipsaw will occur but the R values are huge.
I can't figure out how to code this in Tradestation but if I had a couple weeks off work I'd daytrade it and see how doable it is.
The rules:
1. Trend line under lowest lows from start to present, without going through any
2. Lows that touch trend line are stop points (similar to the Bombardier Method)
3. Once stopped out, reverse signal by first lower high
The same problem with Dow Theory shows up here: the occasional whipsaw will occur but the R values are huge.
I can't figure out how to code this in Tradestation but if I had a couple weeks off work I'd daytrade it and see how doable it is.
I am still happily enjoying my long position and have adjusted the stop to 1242, a higher low created on the way up. I tweaked a lot with my charting setup in December to best visualize the trends that I trade. The primary charts are now the 4H ES and 1H SPY each with the well-known rainbow moving averages (EMA). The reason for replacing the 1H with 4H ES is because the trends here appear much more tradeable, while the 1H still shows too much whipsaw. Going forward, this will be my workspace and I'll see how it helps me in my real trading.
I don't think that any action based on an expectation is a good idea at all. Assumingly, you worry that we will crash anytime. It's something I have put aside after a while. I just had enough of worrying all the time. Even if someone screams "fire", I would probably just sit here until I actually feel my hair was catching fire. That's how passive I have become in my trading, and so far this approach has served me very well. Don't trade expectations. That's probably a lesson that needs to be rehearsed and re-learned among all traders. Would it be better to close at 1280? The question rather is, what do you do if we did not crash and we continue to 1300? You'd probably be very angry. Some day you check back and see the market at 1350. Even angrier. But if the market does drop, my stop will take care of the rest and I'll still be happily enjoying a profit. I'm in a very desirable position where I cannot lose anymore. Why would I ever want to give up on such advantage?
Quote from Bombardier:
I don't think that any action based on an expectation is a good idea at all. Assumingly, you worry that we will crash anytime. It's something I have put aside after a while. I just had enough of worrying all the time. Even if someone screams "fire", I would probably just sit here until I actually feel my hair was catching fire. That's how passive I have become in my trading, and so far this approach has served me very well. Don't trade expectations. That's probably a lesson that needs to be rehearsed and re-learned among all traders. Would it be better to close at 1280? The question rather is, what do you do if we did not crash and we continue to 1300? You'd probably be very angry. Some day you check back and see the market at 1350. Even angrier. But if the market does drop, my stop will take care of the rest and I'll still be happily enjoying a profit. I'm in a very desirable position where I cannot lose anymore. Why would I ever want to give up on such advantage?
The kind of unbounded TF you are talking about will work well in single names (like single commodities), but when you are talking about indexes (composed of 500 names), then back-fill is to be very much expected. I have a feeling (untested) that "over-time" taking profits after a 60-80 point 'quick' move in your favor will outperform compared to your approach. What I am saying is if you let market come back 40 point from here till 1242 then only you will stop out. Much better to book profits now (maybe wait till we reach 1290+) and then book profits. Unless and until you expect market to run to 1350 without any significant pull-back.
But I think the more relevant discussion/research point is how to add to your position as it moves in your favor and if you start trading multiple contracts, discussion in above paragraph will be moot - as scaling out will become the natural strategy.
I hope you have read "Way of the Turtles" by Curtis Faith. Highly recommended - the classic text of TF.
Best Wishes!
You point out a very valid issue, gmst. Position scaling is something I will have to deal with in the near future once my accumulated profits justify an upgrade. When I traded CFDs and ETFs back then (which allow very flexible scaling in small amounts), I haven't had much success with adding to a position. The market usually turned against me the moment I increased, so I was always better off if I hadn't done anything. I'm sure there are traders who do it successfully, and I would like to hear their thoughts. As far as I am concerned, I find scaling out a more relevant tactic to explore.
- Status