BooYa!! I'm hyped up about Forex again :)) - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Thanks again Phillip for taking the time outside your busy schedule to provide clearly defined information on this site.
Yeah, go for mathematical justification for your systems. If that's giving you a hard time settle for logical justifications. If thats still not working then attack the popular systems. Still no luck and you're finding you need some motivations for your hard work. Then start curve fitting and playing around with meta-trader those will put a smile on your face and keep you learning. Looks like you're already doing that with the specific charts. Demo test your curve fitted systems, if they're failing then return to step 1.
Alright then. I still can't match you here, but I've gotten some working EAs going. Always something to learn. Thanks :)
I use MAE and MFE as my critical metrics for assessing the quality of market forecasting my strategy is performing to. [...]
That is a great info & graphic. Thanks Phil...
I don't have much to contribute on this yet but keep it going Ubzen, it's great thread.
Just thought I would add these graphs to show money can be made with an EA and give hope to you all. The back test is for my scalping EA over 3 weeks and the picture is what it did today on cable. I'm not sure if it will continue because cable has been very trending but it made 1% increase in account in a few hours today.
scalping is kinda the "trivial" way to make profits...it is rather easy to make profits in live testing with a scalper, the challenge is finding a broker who can be convinced that it is also in their best interest to keep letting you scalp because most scalping trades end up taking money from the broker for a variety of reasons unless they are truly an ECN (in which case you are going to be trading integer lot amounts and so on, none of this fractional lot stuff)
Are you showing us live account or just forward testing on a demo account? Regardless, stick with your current EA for another month or two and you'll see exactly what I am alluding to..
Nice job Ruptor and thanks for sharing your live experience. Hope you can keep us updated on the progress of your hopefully successfully future trading. IMO the biggest obstacle with Forex trading is developing the confidence to fund real money behind these EA's we develop. I once seen a GBP.USD scalper which made Heavy profits (10,000+Orders) in 5-years and then Failed Badly (10,000+Orders) within the last 5-years. The programmer was giving it away as charity to help others make money too. After the first 5000 wins, I was a believer, I mean it was statically valid #of trades. How could something so good go so bad. Up till now, I still don't know if the results good or bad is because of the Demo Data, the System Being Countered or the Market Changed. But one thing was certain, this was why the programmer was giving away the system. Hence where do I get the confidence to trade any system if 5-years performance is not good enough?
The only Scalping I believe in at this junction are the Arbitrage kind <i.e> you already know where price is headed. All the brokers I'm attracted to do not allow me to take 4-8 pips profit. It's usually above 10 pips so I naturally assumed thats the industry standards. Seem even if Non-Arbitrage Scalping works, the brokers are too scared of single-digits-pip-profits and therefore would start to re-route your orders to the Real markets which may cause delays in minutes and re-quoits and break most of those types of systems. In either case, the broker is gonna ban you. Looks like you're using $500 macro lots and hoping to slide under the radar lol. That might work in Casino Table games where you can rat-hole the chips but here, they know exactly how much you're taking. If they're not banning you yet - then maybe it's because you haven't reached the cut-off point yet. Or too low stakes to make enough money in the long-run for Them to care about.
What I like about the BooYa system I posted on the first page is that, the First back-test was without any Optimization. I watched the market the week before, programmed it, and boom 3 months of similar performance. Any system really can break with market change, so I guess we all have to keep an eye on the "when to quit" factor. All IMHO.
Hence where do I get the confidence to trade any system if 5-years performance is not good enough?
It really comes down to asking the right question and knowing what question to ask.
Performance over a historical period of time, regardless of length, can not be expected to provide any indication of performance over a future period of time unless that future period of time just so happens to coincidentally mimic the market actions of the backtested period of time.
This is where the terms over-optimization and curve-fitting come into use. But the problem starts even sooner. The problem is the wrong kinds of questions are asked regarding the performance during backtesting. The questions to ask are "why did this strategy perform like it did over the backtested time period?" and "how can I keep it performing that way in the future?".
Asking "what were the profits?" or "what is the drawdown?" are value-less questions, they add no value to your ability to determine if it can be expected to yield profits in the future (unless one happens to naive enough as to believe that past performance is always indicative of future results).
So looking at that 5yr good scalping system the question to ask of it is why was it good for five years, what was its decision tree setup to assess before making market entries and exits and why was that producing profits for five years? Usually such analyses will yield the crucial information in terms of the assumptions and Achilles heel of the trade algorithm and from there you can determine (without any forward testing whatsoever) whether the system is bound for greatness or bound for ruin (given time).
I've said before I don't backtest looking for profits or looking for drawdown, those are answers to the wrong questions. I look for accuracy in market prediction. How "good" was my algorithm in predicting a market up or market down in the past? Why is the algorithm doing a good job making such predictions?
If I can't answer these questions then I certainly have no justification for placing any confidence in the ability to extract profits from markets on future price activity.
.... All IMHO.
You may be nearer to an overall strategy than you believe
> watched the market the week before
Well a month or so of demo trading is better but never mind
> First back-test was without any Optimization
A very good pointer to something with promise
> Any system really can break with market change
Just knowing that is extremely useful - but could you act on that - kill an EA you spent months on?
If you can - and do a post mortem, find what killed it and how, the next one will be better
> keep an eye on the "when to quit" factor
Every EA has one - its just spotting it - e.g. when drawdown, or max consecutive losses, etc is out of line with expectations
You're looking for the factor/s that are counter to your system
FWIW
-BB-