Hi Carnino, did you try a variation like. After first loss, EA would sleep for 1 or even 4 hours, if after that the breakout still not happened, send a new pending stop. I guess if that level reach strong resistance, it can breakout with less retracement later.
I’m just guessing, maybe I’m wrong but we can try this.
Regards
No, I did not try that, but from a general look at the results, I do not think it is worth pursing.
I guess that it was a rash idea on my part that just complicates things. Most of the time, simpler is usually better.
Carnino, when you were originally doing your back tests did you notice any larger drawdowns similar to what we are experiencing now? I am just thinking that this might be a normal thing with the system. I would assume that all systems have some periods with large drawdowns. I guess the real question is if that is all that this is, or is this a sign that it does not really work in these specific conditions, whatever those might be.
If you really think about it, it would only take a couple of decent runners on our side to get out of this.
Yes, given that [I]Vijay’s[/I] system is based on a fixed stop size, it is bound to undergo stressful periods under certain conditions, specially when the market is highly volatile and erratic. If you look at the back-tests that I published, you will see that some months did badly, but not to far away from break-even.
A solution with which I have been having some success, is by the use of a variable stop size. Initially, I started testing with an ATR based stop size, which is an improvement but is still not ideal and it was affected negatively this month as well.
So, I decided to take it a step further, by using a variable stop size based on MFE and MAE analysis, but this method is still in its infancy.
Both the ATR and MFE/MAE based implementations in my EA, are still crude and have not yet properly matured. I have not been working on them the past few weeks due to other personal obligations, but within the next few weeks I will be focusing more on this task and will soon start my own thread for those methods.
Hopefully in the coming weeks, I will be able to give you more concrete details backed with published back-testing and will also start a public forward test on [I]MyFXBook[/I] as well.
I will analyse it and let you know. However, it is 00h15 here and am going to bed, so will let you know my findings tomorrow.
Happens to the best of us, lol. Thought it was good idea as I’ve often noticed that the same thing happens.
Another idea: A filter for when news breaks, I’ve seen some EA’s that have them but I don’t how they are set up; meaning keyed off a news site for major announcements or via the spread. A filter to put the EA to sleep for 15-30 minutes if the spread is larger then 10 pips might be worth looking into. But obviously only re-activated if the price returns to with in the previously pending buy/sell orders, otherwise cancel them.
Vijay, I had an idea that might improve the profitability of your original system. Once pos 1 is hit move SL’s for pos 2 & 3 to BE (or perhaps +1 pip); when pos 2 hit then move SL for pos 3 to + 5 pips. Not sure if it’s worth factoring in the spread too. Thoughts on this anyone?
I will run some back-tests of [I]Vijay’s[/I] original strategy with your break-even (and Trailing) proposal and will publish it here later today or tomorrow.
Thanks Carnino! Was just thinking it might help and also be a viable alternative until either you or Vijay gets the bugs worked out on the EA with TS.
What do you mean by “viable alternative”?
Both Vijay’s original strategy and my Single Order with Trailing strategy are both viable! Irrespective of using an EA or not, both can be traded manually on MT4 (even the trailing can be done automatically with MT4 even without an EA).
Sorry, perhaps poor choice of words again. I simply meant using it as an auto trading EA until you get yours up and running. I am of course aware of using it manually as I am doing so now, but sometimes I am away from my machine or don’t have an internet connection, so it would then be a good alternative to use.
Even if you are away from your machine, leaving MT4 open will continue to run the trailing for you. It is a hybrid trailing with step for break-even and continuous from then on.
As for not having Internet connection, that would be a BIG problem. Even an EA would require a continuous Internet connection almost 100% of the time. So, if you have Internet problems, you will not be able to run an EA.
That’s true, but doing it manually you still have to enter the pending orders every day. If you are away or traveling then it may not be possible to do it before the market moves past your entry.
As for not having Internet connection, that would be a BIG problem. Even an EA would require a continuous Internet connection almost 100% of the time. So, if you have Internet problems, you will not be able to run an EA.
Also true, unless you run it on a VPS, which I am considering due to the generally crappy and unreliable internet service in the Philippines - be it WiFi, 3G or 4G.
Here are the comparison results of the Break-Even/Trailing proposal, by [I]piphauler[/I] for [I]Vijay’s[/I] original PARMAR-3P strategy:
[ul]
[li]Symbol & Testing Period: GBP/USD (from 2012-12-31 to 2013-12-27)
[/li][li]ZimeZone: UTC+2 (NY Close)
[/li][li]Trading Days: Monday-Friday
[/li][li]Spread: 1.5 pip
[/li][li]Breakout Offset: (spread) 1.5 pip only on High (0.0 on Low)
[/li][li]Percentage Allocation: Original 60%; 20%; 20%
[/li][li]Stop Size: 5.0 pips
[/li][li]Risk: 1% (Initial balance $10000)
[/li][/ul]
PARMAR-3P (Original)
[ul]
[li]Profit: $12006.10 (120.06% Gain)
[/li][li]Profit Factor: 1.74
[/li][li]Relative Drawdown: 7.02%
[/li][li]Win Rate: 50.15%
[/li][/ul]
PARMAR-3P (MT4 Style Hybrid Trailing - Break-Even Step/Continuous Trailing)
[ul]
[li]Profit: $10470.48 (104.70% Gain)
[/li][li]Profit Factor: 1.84
[/li][li]Relative Drawdown: 5.89%
[/li][li]Win Rate: 61.95%
[/li][/ul]
PARMAR-3P (Discreet Trailing - Break-Even and Stepped Trailing)
[ul]
[li]Profit: $10748.10 (107.48% Gain)
[/li][li]Profit Factor: 1.84
[/li][li]Relative Drawdown: 6.19%
[/li][li]Win Rate: 61.95%
[/li][/ul]
PARMAR-3P (Discreet Trailing & Spread Compensating Trail Gap)
[ul]
[li]Profit: $11200.85 (112.00% Gain)
[/li][li]Profit Factor: 1.87
[/li][li]Relative Drawdown: 5.96%
[/li][li]Win Rate: 61.95%
[/li][/ul]
I will leave it up to you to draw your own conclusions!
Regards,
Carnino
Thanks Carnino
Thanks for your advise Vijay,
I will try admiral Markets.
Cheers
Suraj
Mr. C, first and foremost thank you very much for running the test!
I’m a bit surprised at the outcome. I would have thought by moving the SL’s to BE once the first 5pip target was hit would have increased profitability simply by limiting the losses when the other two targets were not hit. In addition by adding a trailing stop, as you have already demonstrated in prior testing, should have helped too.
What I don’t understand though is the gain was down, but both the profit factor and win rate were up. Any thoughts on how can that be the case?
You have to remember that [I]Vijay’s[/I] original strategy sets Take Profit points, thus limiting potential profits. Trailing Stops too, usually allows less room for trades to evolve and thus also limit profits in their own way. Put the two together and profits will most definitely drop.
However, that is not the complete picture. Since the Win Rate and the Profit Factor increases, it means that the strategy is less risky, and you can therefore increase the risk in order to increase profits.
Most “good” traders, look for “safe” strategies, even if they do not have very high gains, because that means that they can risk more of the capital “safely”, for better rewards. The 1% Risk rule is only a guideline that depends on many factors.
It has been a woeful month so far, but I am optimistic about the merit of the strategy over the longer term.