Dreamliners’ MPAS EAs, Indicators, etc | Page 7 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Dreamliners’ MPAS (Mechanical Price Action Strategy) System 486 replies

MT4 doesn't save trendlines, indicators, etc 1 reply

Quantum leaps in tweaking MT4 (indicators, EAs, etc.) 0 replies

Merc's special Fibonacci trading,in combo w/QQE Gann etc.etc. 7 replies

indicators, etc... 12 replies

Attachments: Dreamliners’ MPAS EAs, Indicators, etc
Exit Attachments

Dreamliners’ MPAS EAs, Indicators, etc

  • #121
  • Jan 2, 2011 4:40pm Jan 2, 2011 4:40pm
I see value in backtesting, live, by hand. I see no value in an EA backtesting whatsoever. That's just me, and it stems from having such a negative view of EAs.

Quoting Musashi
Do you see no point in backtesting?
  • #122
  • Jan 2, 2011 4:42pm Jan 2, 2011 4:42pm
Oh, sorry, didn't see any sarcasm, guess that's hard to see in text, unless there is a or some other way to tell. I do agree, though, even though he was being sarcastic. I don't think you'll find a profitable EA in here, or anywhere in Forex Factory, TSD, or any other forum. Just my opinion.

Quoting tazmaan
What's funny is crodzilla's post was sarcastic, DL's was not
  • #123
  • Jan 2, 2011 4:50pm Jan 2, 2011 4:50pm
As azjeff and smjones have iterated here, which is the absolute truth that most don't want to admit is, EA's will never work over the long term. EA's are written to take advantage of a short term aberration/pattern in a market which will eventually become part of the market, as it is taken advantage of, and eventually disappears.

Every EA marketer knows this, and takes advantage of every naive EA buyer.

The sooner people realize this, the sooner a person realizes that there is only one Expert Advisor... that's you!
  • #124
  • Jan 2, 2011 4:53pm Jan 2, 2011 4:53pm
Quoting Dreamliner
I see value in backtesting, live, by hand. I see no value in an EA backtesting whatsoever. That's just me, and it stems from having such a negative view of EAs.
I understand your negative views on EA's however I do believe they can be useful especially for speeding up the manual backtesting approach.

Using the Visual Test option in Strategy Tester is very useful for comparing manual testing and mechanical testing results.

A statistically large enough sample is what is required to determine if a strategy is likely to succeed in the future.

I would say at a minimum +100 trades is required to have some idea.
  • #125
  • Jan 2, 2011 4:58pm Jan 2, 2011 4:58pm
Quoting crodzilla
Yes, there are no profitable EA's here. Everyone should abandon any further research of this method using EA's. Its a complete waste of time...
Assuming your summary statement is valid, and in light of your long experience
with both Fx trading and MQL Programing of many EA's, would you please dispell or attest to the idea profitable EA's do or do not exist? And that methods with an "edge" statistically and mathematically do exist both in manual form and automated form for the readers/traders of this thread.

I personally am party to profitable Fx trading methods and the propretary methods which early on were tested manually are profitable EA's which are Price Action based (non-indicator driven) systems having been profitable over the last four years. This does not mean they will be forever profitable, however the profits made over the life of these trading tools have far exceeded their probable future loss of trading capital.

In fact, in reading this thread and your input I see your early trials were represented to be a high win:loss ratio with this method using your version of your EA? And, on this small statistical sample it initially looked to you this method had promise? Now, it does not??? Interesting.

Perhaps in the spirit of our earlier comments, you could be encouraged to outline not a method but the characteristics you know from your own personal experience are the framework of profitable manual trading systems which ultimately have proven to be profitable automated trading systems?

I like your style and candid approach to Fx Trading and would appreciate, as I'm sure other here agree, your elements outlined that must be present to culminate in a profitable trading method, both manually and automated.

Cheers,
Atlas1
  • #126
  • Jan 2, 2011 5:05pm Jan 2, 2011 5:05pm
Quoting Dreamliner
methinks somebody was offended at the challenge of his backtesting methods. Hard to harbor grudges in life, my friend, it will get you in the end. My encouragement to you is to drop this method entirely from your trading EA arsenal; it is clear that it is not profitable for you (and maybe for nobody).
Offended at the challenge of my backtesting method...no, not really. I know the method of using accurate tick data is the best you can do in MT4.

I put my backtest results out there to give everyone a sense of the long-term profitability of the system(s). I honestly thought it would be helpful. I had quite a few people PM about the results and asking questions about the results. Thus, it seemed like a worthwhile exercise.

But I do get bothered when people say things cannot be done or that there is only one way to do things. But no grudges here...as I said before, debate can be very productive. zamanib is a complete dolt, but I certainly don't hold any grudges against him or you. I wish both of you lots of luck with your trading.

I appreciate your encouragement. I do not use your method and never created a trading EA of your method. And my EA arsenal is just fine already, thanks.

Quote
EAs are simply not profitable

This may very well be true for you.
  • #127
  • Jan 2, 2011 5:14pm Jan 2, 2011 5:14pm
Quoting crodzilla
As azjeff and smjones have iterated here, which is the absolute truth that most don't want to admit is, EA's will never work over the long term. EA's are written to take advantage of a short term aberration/pattern in a market which will eventually become part of the market, as it is taken advantage of, and eventually disappears.

Every EA marketer knows this, and takes advantage of every naive EA buyer.

The sooner people realize this, the sooner a person realizes that there is only one Expert Advisor... that's you!
I think I just stumbled upon a really very informative and excellent discussion on systems. There is a lot of truth floating around and a few near misses but essentially whatever can be done with a mechanical system can also be done with an automated system (EA). This means that if a mechanical system requires periodic adaptation to be profitable, then the automated strategy also has to adapt over time as well. Elementary enough concept, but the idea that a single static set of rules will have a positive edge over a long period of time is really logic defying.

I know that's the dream the system sellers sell - buy this system, get rich, you don't even have to know how it works!

The truth is really well summarized in Crodzilla's post above that I would restate as follows: a single non-adaptive EA won't work over the longer term unless the market stops changing. Since the markets are always changing, if you hope to run an automated strategy (EA) over the longer term, then you must design your strategy to automatically adapt to market changes, or you must adapt it manually as a manual trader would do - by observation and testing. Designing a self-adaptive system is a lifetime's work. I happen to be working on this concept right now. If anyone is working on the same concept and would like to compare notes, shoot me a PM.
  • #128
  • Jan 2, 2011 5:21pm Jan 2, 2011 5:21pm
For the most part, any strategy that has a clear set of rules, and the rules can be quantified, can be automated (be it an EA in MT4 or some other automation technique). And the automated trader's results should match those of the manual trader. Usually it is just simple mathematics and code.

It is when you start adding discretion to the strategy that difficulties arise in automating it. Trying to quantify discretion can drive even the best programmer crazy.
  • #129
  • Jan 2, 2011 5:24pm Jan 2, 2011 5:24pm
Quoting Atlas1
I personally am party to profitable Fx trading methods and the propretary methods which early on were tested manually are profitable EA's which are Price Action based (non-indicator driven) systems having been profitable over the last four years. This does not mean they will be forever profitable, however the profits made over the life of these trading tools have far exceeded their probable future loss of trading capital.

Cheers,
Atlas1
How many times in the 4 year period were parameters, symbols, timeframes, MM rules etc., changed to fit current volatility / market conditions?
  • #130
  • Jan 2, 2011 5:38pm Jan 2, 2011 5:38pm
I have traded/evaluated hundreds of automated trading systems. I have developed several private systems for other markets. I have developed dozens of private EA's for forex.

Let me clarify... an automated trading system that works over time, will never see a public audience, ever.

ATLAS1:
I could never answer you in just the several keystrokes that I'm willing to put forward here. Profitable methods exist, both manually and (semi-) automated. Most people are not willing to put forth the effort required both mentally or strategically to put together a profitable method. I have been involved in the investing "game" for over 20 years, I'm still learning. I still want to learn more. As technology has advanced, trading methods need to advance. Strategic/forward thinking needs to advance. If you don't have the thirst for knowledge and are just looking for an easy "fix," you are wasting your time. This forum is not a waste of time in my opinion. There are nuggets of interesting methods/ideas on FF. There are several very smart people that hang here.

I'm done...
  • #131
  • Jan 2, 2011 6:15pm Jan 2, 2011 6:15pm
Well, the short EURUSD that I didn't close on Friday in my live account ended up closing for 60 pips on the gap. This is quite a cool experiment though...
  • #132
  • Jan 2, 2011 6:51pm Jan 2, 2011 6:51pm
Quoting Chris_B
Hi!!

It's not a bug...RangeBars Charts are printed using only BID data, when you
Hope this helps...
Hello Chris. Thanks for the clarification.

My take: taking the bid/ ask into consideration would by right for at market orders but not when we are placing stop orders several pips in advance. However, if you can provide an option to disregard spreads in all calculations (including trailing stops/ stop loss etc) that will be gr8.

Another thing: The ability to specify trailing stop and break even is great - it increases profit in backtests. In live tests, the Breakeven and stop loss are conflicting with each other it seems. In a live trade this morning, I noticed the trailing stop being activated repeatedly, but getting reset again and again to the breakeven value. You might want to check part of code to avid this type of conflict.

Look forwar to testing your new version.

Thanks
  • #133
  • Jan 2, 2011 7:08pm Jan 2, 2011 7:08pm
Quoting msood
Hello Chris. Thanks for the clarification.

My take: taking the bid/ ask into consideration would by right for at market orders but not when we are placing stop orders several pips in advance. However, if you can provide an option to disregard spreads in all calculations (including trailing stops/ stop loss etc) that will be gr8.

Another thing: The ability to specify trailing stop and break even is great - it increases profit in backtests. In live tests, the Breakeven and stop loss are conflicting with each other it seems. In a live trade this...
Hi!!

Yes...I'm fixing the StopToBreakeven/Trailing conflict...you're right I didn't try it...

About the spread...I've added a switch so you can turn it off..even if I think you have to consider it also if you use stop orders, because you want to put an order at a predetermined price (Open of the current bar), but that price is the Bid price of the data feed.
A practical example:

let's say the open is at 1.60000, you want to open your Buy trade at that price right? not before...
If you don't consider spread it will trigger at 1.59980 (Bid) for example (because the Ask price will be 1.60000) ...and it's not the price you want...
the price may reach 1.59980, make a High and then reverse...without really making your 1.60000...
Anyway with the new version you can choose if use it or not...
Market is like a puzzle, YOU have to fit the pieces!!
  • #134
  • Jan 2, 2011 9:20pm Jan 2, 2011 9:20pm
All ea lovers-- this system in not profitable -- over long time so - why waste you time -- all who thinks is is waste system -leave move on-- for me i am sticking with it-- i trade it manually -- when i am infont of the computer-- dreamliner was very innovative in this strategy - and i give him full credit for it--
PROFITABLE TRADERS ARE NOT LUCKY BUT ARE WELL PREPARED
  • #135
  • Jan 2, 2011 10:07pm Jan 2, 2011 10:07pm
'for the ea lovers' ?
this is the ea version of the thread, you could always not post in this thread as well you know
  • #136
  • Jan 2, 2011 10:24pm Jan 2, 2011 10:24pm
Quoting FXEZ
How many times in the 4 year period were parameters, symbols, timeframes, MM rules etc., changed to fit current volatility / market conditions?
None. Some years had higher gains than others with the same set files, but all were profitable and with no "curve fitting".

Atlas1
  • #137
  • Jan 2, 2011 11:53pm Jan 2, 2011 11:53pm
[quote=Chris_B;4280719]Hi!!

Yes...I'm fixing the StopToBreakeven/Trailing conflict...you're right I didn't try it...

Thank you. Yes, I see your logic, but its good to have that option. When you r ready, I would like to test it on live account - I use micro lots for testing and any potential loss is offset by more time tested strategies - so its worthwhile to get the real market feedback.
  • #138
  • Jan 3, 2011 12:47am Jan 3, 2011 12:47am
I look forward to your final vs Chris. The set files for the different strategy's type is like ice cream on the cake.
  • #139
  • Jan 3, 2011 2:30am Jan 3, 2011 2:30am
Almost all trades so far from the opening of market today are losses.
I am using basic rules 30 pips bar with SL=10 and TP=20. Everybody have similar results? Just wondering.
AshN
  • #140
  • Jan 3, 2011 2:54am Jan 3, 2011 2:54am
Quoting netFX
Almost all trades so far from the opening of market today are losses.
I am using basic rules 30 pips bar with SL=10 and TP=20. Everybody have similar results? Just wondering.
This strategy not yeilding same results as back-testing which I did over the weekend. It is not anywhere close to it. In back-testing I got amazing % wins but on live/demo so far all losses.
Back-testing I am referring to the one which is done by Ea on offline chart generated by RangeBar package as demonstrated on U-tube. This back-testing is not realiable.
AshN
0 traders viewing now