Hopes and Dreams shattered... - page 2

New comment
 
Dear All,

Thanks for all encouragements and insights...let me give an example of using human expertise...consider that your hourly EA (that has positive expectancy and quite robust) gives you a trade entry signal 10 minutes before non-farm payroll on EURUSD pair...would you allow the EA to execute the trade or would you like to wait till NFP data is out and market is settled a little? On such occasions, I would like to intervene and delay the trade...in my mind, it is not second guessing the EA, but just money management.

TA can definitely exploit the variability of the market, however, the uncertainty of the market needs human intervention. Ensuring that you provide your EA better opportunities helps with better money management. However, I do agree that except for such uncertainties, second guessing the EA can only deterioriate the EA performance in the long run.

Second is regarding the feasibility of EAs over long run...markets keep on changing in terms of their liquidity, volatility and maturity..classic example being GBPCHF. ..a great pair in 1980/90 with lots of volatility and good trends...but, look at the pair now..it trades pretty maturely...so, the EA that was tuned to exploit GBPCHF characteristics in 1980/90 may no longer be profitable today and what is profitable today may not be profitable in 2010/15, therein lies the excitement of these markets!

Of course, this is just my opinion..based on what little I've seen of ForEx in the past 1.5 yrs and I may be wrong... :)
 
Only one thing I can add here, that it would be better to have all trades closed until friday's economic releases. The rest of the day is probable to have very high volatility, plus that 2 weekend days generate nothing but rollover charges. .. However, not only economic releases are dangerous... I remember I read about an ECB intervention that shift EURUSD upside 500 pips in 15 minutes... and there's no indicator able to protect you then... Maratha is also right about how pairs importance shift in time...But does this affect systems with short timeframes ? Automation's holy grail would be a system able to generate profit using a lot of trades on short timeframes, but on a statistical basis, not needing to know where the market is going... And here comes the problem of false positives, generating potentially more losses that earnings on good trades...
 
Benna:

After a search of many, many months I came to the conclusion that there is no Expert Advisor that can actually consistently make money over the long term. Can anybody prove me wrong?


Benna,

I am new to MQL4 but very experienced with automated trading (primarily in the futures market). I am in charge of porting some of our futures trading strategies to the forex market and I can assure you that automated trading does make money.

Having said that, I will pose a question. Why would anyone sell a profitable trading strategy let alone publish it. I am in the final stages of testing our first application (ported from the futures) that consistently makes 1%+ per day.

It is basically a modified scalping application and the reason it "only" makes 1% per day is because of the inherent limits in the forex market that make scalping so difficult. Without those we would be making at least twice that amount. We don't sell it and I would be fired (probably executed) if I published even a hint of what we actually do.

I have read threads about brokers trading against you and stealing your ideas. Let them trade against me because that will cause the market to move which will simply make money in a different strategy we run. And, why would they steal my ideas when they have teams of traders and programmers to come up with their own ideas (have you heard the phrase "not invented here") not to mention the fact that they don't need my ideas because they already get the spread.

I heard it said best by Mark Fisher; most traders fail because they didn't attend BYA University (bust your a**). Work hard enough, long enough and with the right people and you will be successful at anything.

I am trying to offer realistic encouragement and hope I do not sound like an arrogant jerk.

Scott
 

All,

I've wrote an MT4 strategy that duplicates one that I wrote on Tradestation. However, the one on MT4 has a negative sloping equity line and the Tradestation one is positive. I can't for the life of me figure out what it is doing. Has anyone else came from Tradestation to MT4 experienced this problem?

Some questions I have regarding it:

  1. How does MT4 calculate slippage. I noticed if I click on the symbol properties it shows a spread of 2 for EURUSD but does it calculate these into the results? Can the defaults be changed?
  2. I've manually looked at a trade results that was placed through MT4 and it was off by $100 in the profit/loss from what I calculated. Why is this?
  3. I've discovered the AccountEquity() command in MT4. But I believe this command pulls your current account equity regardless of if you are back testing or live. Is there another command that will actually keep track of your account equity during a back test simulation run?
 
  1. How does MT4 calculate slippage. I noticed if I click on the symbol properties it shows a spread of 2 for EURUSD but does it calculate these into the results? Can the defaults be changed?

    You sell at Bid and Buy at Ask. Spread is included in backtest.
  2. I've manually looked at a trade results that was placed through MT4 and it was off by $100 in the profit/loss from what I calculated. Why is this?

    Don't know. Examine the differences.
  3. I've discovered the AccountEquity() command in MT4. But I believe this command pulls your current account equity regardless of if you are back testing or live. Is there another command that will actually keep track of your account equity during a back test simulation run?

    It pulls test equity under the tester. Initial deposit can be set in Expert Parameters.
 
ScottB:

Benna,

I am new to MQL4 but very experienced with automated trading (primarily in the futures market). I am in charge of porting some of our futures trading strategies to the forex market and I can assure you that automated trading does make money.

Having said that, I will pose a question. Why would anyone sell a profitable trading strategy let alone publish it. I am in the final stages of testing our first application (ported from the futures) that consistently makes 1%+ per day.

It is basically a modified scalping application and the reason it "only" makes 1% per day is because of the inherent limits in the forex market that make scalping so difficult. Without those we would be making at least twice that amount. We don't sell it and I would be fired (probably executed) if I published even a hint of what we actually do.

I have read threads about brokers trading against you and stealing your ideas. Let them trade against me because that will cause the market to move which will simply make money in a different strategy we run. And, why would they steal my ideas when they have teams of traders and programmers to come up with their own ideas (have you heard the phrase "not invented here") not to mention the fact that they don't need my ideas because they already get the spread.

I heard it said best by Mark Fisher; most traders fail because they didn't attend BYA University (bust your a**). Work hard enough, long enough and with the right people and you will be successful at anything.

I am trying to offer realistic encouragement and hope I do not sound like an arrogant jerk.

Scott

All traders,


Why would someone sell a profitable EA or trading system? The eternal question, the answer is simple.


Because the EA you receive is limited to 1 currency and 1 time frame. Maybe theirs is unlimited? Plus they issue licenses limiting the amount of users they will allow.


What's so unbelievable?


Could it be you bought an EA who's advertisers promoted it by using a period of time the EA did well, like 2 months, and you bought it. You back test it over 3 years and it's a loser.


I'm a believer in profitable commercial trading systems.

 
Benna:

After a search of many, many months I came to the conclusion that there is no Expert Advisor that can actually consistently make money over the long term...

Free EAs tend to be oversimplified. They usually are very short programs totaling few hundreds lines or less. They tend to be statically designed for a highly dynamical environment. And, most of them lack sophisticated reasoning, planning and learning algorithms...


Working Expert Advisors do exist, certainly not for free; and chances are they will become better and better as Artificial Intelligence technology leaps forward.


Unless the majority of traders change their attitudes toward paying for working systems, good programmers will keep refusing to release their best work to public domain. After all, we all have to pay for rent, clothing, food, heath insurance, and those REALLY expensive programming books. So, the next time a good programmer asks you for a reasonable price for his programming jewel: open your mind ...and don't be stingy!

 

and chances are they will become better and better

I think a comparison of the results of the 3 years of The Contest show that idea in action.

https://championship.mql5.com/2012/en/

https://championship.mql5.com/2012/en/

https://championship.mql5.com/2012/en/

12
New comment
Reason: