Most sucessful? - page 2
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
You might want to consider the spreadsheet that I have attached for you and for the benefit of other members of this forum before you start writing code or looking for the best EA. This is something that you rarely see in beginning books on Forex and its a shame because its probably one of the main reasons the majority of traders lose all their money, regardless if they trade manually or use an EA.
.
.
.
Very nice worksheet. Based on your graphs it looks like the lower end of "win probabillity" is somewhere between 35-40% to have a profitable system with a 1:1 SL to PF. So should I focus less on the number of winning trades in my strategies? Also, do you take spread into consideration? Using my 1:1 example above I should be able to "gamble" 50/50 and still make a profit.
First of all I found a bug in my first spread sheet. The corrected and simpler spread sheet is attached below. It only shows the equity curve for the fixed risk case. When I have time I will correct the other graph.
In answer to your first question whether you make money in the long term depends on both your winning probability and your payoff. With a 50 per cent winning rate and a reward/risk ratio of 1, your graph will not be consistent and this is a bad sign. You want consistency. You want your graph to slope from the lower left to the upper right. You don't want the graph bouncing all over the place.
If you put in 40% or less win probability with a 1:1 ratio, look at the graph. It slopes down to the right. You will lose money.
In my example I assumed a 2 pips spread for the EUR/USD. However, in the real world the spread is not fixed and your EA must always monitor the spread before it gets into a trade, especially during times of volatility in the market-- non-farm payroll,etc. The changing spread level is the nemesis of all scalpers and can create losses very fast. (Not to mention re-quotes and non-fills of orders. Scalping has many pitfalls.)
If you opted to create an EA or trade with a long term trend strategy, you might get better results. Lets say the average payoff for a long term trend trade is 4 to 1, but you you only have a 30% win rate. Will your expectation value be good? Absolutely! Plug in the values and see what the equity curve shows for 100 trades.
First of all I found a bug in my first spread sheet. The corrected and simpler spread sheet is attached below. It only shows the equity curve for the fixed risk case. When I have time I will correct the other graph.
In answer to your first question whether you make money in the long term depends on both your winning probability and your payoff. With a 50 per cent winning rate and a reward/risk ratio of 1, your graph will not be consistent and this is a bad sign. You want consistency. You want your graph to slope from the lower left to the upper right. You don't want the graph bouncing all over the place.
If you put in 40% or less win probability with a 1:1 ratio, look at the graph. It slopes down to the right. You will lose money.
In my example I assumed a 2 pips spread for the EUR/USD. However, in the real world the spread is not fixed and your EA must always monitor the spread before it gets into a trade, especially during times of volatility in the market-- non-farm payroll,etc. The changing spread level is the nemesis of all scalpers and can create losses very fast. (Not to mention re-quotes and non-fills of orders. Scalping has many pitfalls.)
If you opted to create an EA or trade with a long term trend strategy, you might get better results. Lets say the average payoff for a long term trend trade is 4 to 1, but you you only have a 30% win rate. Will your expectation value be good? Absolutely! Plug in the values and see what the equity curve shows for 100 trades.
Ah, ok. I made a mistake too. I fiddled with the non green values even though you said not to! So using the new sheet I can see I need to have about a 2.5:1 payoff for a 40% win probabillity to comfortably make a profit. My current EA that I am working on falls somewhere in the 40-45% win rate but I think my profit expectations are way off. I think using your sheet and tuning my profit target will help. Thanks for the good work.
Ah, ok. I made a mistake too. I fiddled with the non green values even though you said not to! So using the new sheet I can see I need to have about a 2.5:1 payoff for a 40% win probabillity to comfortably make a profit. My current EA that I am working on falls somewhere in the 40-45% win rate but I think my profit expectations are way off. I think using your sheet and tuning my profit target will help. Thanks for the good work.
Your welcome. Good luck on your EA.
First of all I found a bug in my first spread sheet. The corrected and simpler spread sheet is attached below. It only shows the equity curve for the fixed risk case. When I have time I will correct the other graph.
In answer to your first question whether you make money in the long term depends on both your winning probability and your payoff. With a 50 per cent winning rate and a reward/risk ratio of 1, your graph will not be consistent and this is a bad sign. You want consistency. You want your graph to slope from the lower left to the upper right. You don't want the graph bouncing all over the place.
If you put in 40% or less win probability with a 1:1 ratio, look at the graph. It slopes down to the right. You will lose money.
In my example I assumed a 2 pips spread for the EUR/USD. However, in the real world the spread is not fixed and your EA must always monitor the spread before it gets into a trade, especially during times of volatility in the market-- non-farm payroll,etc. The changing spread level is the nemesis of all scalpers and can create losses very fast. (Not to mention re-quotes and non-fills of orders. Scalping has many pitfalls.)
If you opted to create an EA or trade with a long term trend strategy, you might get better results. Lets say the average payoff for a long term trend trade is 4 to 1, but you you only have a 30% win rate. Will your expectation value be good? Absolutely! Plug in the values and see what the equity curve shows for 100 trades.
Thanks for the explanation----this is good stuff, if I may say so. I will try to digest this as I have time.
Also may I run an idea by you for a EA?
Know anything about using this site?
http://www.freelancer.com/jobs/
Thanks for the explanation----this is good stuff, if I may say so. I will try to digest this as I have time.
Also may I run an idea by you for a EA?
Know anything about using this site?
http://www.freelancer.com/jobs/
Thanks. You can run an idea by me, but I am not a professional programmer. I am self-taught on MQL4. I have written many EA's for years that work fine for my purposes. I could never sell them as a commerical product because they have a lot of inefficiencies and vulnerabilities in them. A real programmer reading my code would cringe if they saw the potential problems I could have with my EA's.
On this site you will get help on specific problems you have with EA's or coding, but no one will write a whole ea for you for free. This is to be expected. Remember, the programmers probably on this forum went through a rigorous 4 year computer science program, and many more than likely incurred student debt to do it. You can see why they would not just give away their services.
They are helpful on specific points, but they expect you to learn MQL4 on your own. They will help you if they sense you are taking the initiative to do the grunt work of coding. There are good tutorials on this forum that will help you learn MQL4 programming. Click on the "Book" at the top of this page as a starter.
What's your idea?
Hello,
I'd like to share my opinion with you and it is the following: trading and markets are complex. You can write a great EA that makes you tons of profits, but it all depends on how good is your trading strategy and your hability to code it. We are all looking for our holy grial in different directions, and we would all point you to different places. So I'll point you into mine, which comes from my personal experience: the simplest way to make a killing EA is to find a simple hedge against the market (52%-48% is enough) and use an aggresive money management strategy to exploit it. The goal of my ea's are to behave like a casino, being my broker or whichever is taking the opposite trade the irreponsable gambler.
Great traders will tell you that a measly 30% winning trades selection must be enough to make a fortune, and for some of them, it is.
Exploit your hedge and make the broker play against you like a madman.
That being said: trading is fun for me. But I make the big bucks by value investing.
Cheers.
This has become a very interesting thread. Has opened my eyes to different angles. After all that I read, I am not sure that there is a really good ea out there. If someone had a good one, why would they share it---not. We all know that there is no Holy Grail of eas - with that said there are a lot here that look interesting and I have spent a lot of time on one of them that ended up having a deficit with the money management. I am thinking that the successful scalping eas are hard to find. If someone has a service that claims such success----they are probably making more from the sale of it than they are from trading it.
Guess I am getting tired ----must admit it----spent a lot of time on this. I still have this idea of a ea that uses stoch, macd and moving average-----used to use it on stocks when the trading was good. I am sure that someone has already made one and it is probably floating around here in this forum. It would not be a scalping ea but a little longer term-- on the four hr chart.
The excell chart was excellent.
thanks
After all that I read, I am not sure that there is a really good ea out there.
There are good ea's, but how they are used can be bad.
Before I go into that, let me share some more things with you that might give you some other insights. I have attached another spread sheet that you might find enlightening. When you open it, you will see a chart. Zoom in to 50% and you will see it better. The green line is a 100 sma, the red line is a 20 sma, the grey line is an 8 sma, and the black line is a line graph of the price. Use the bottom bar to scroll the chart left and right. Hit cntrl c to generate a new chart.
Now, someone who knows about reading charts will tell you there are obvious patterns on the chart-- double and triple tops and bottoms, head and shoulders, support and resistance, consolidation periods, and strong trending periods. They might even venture to give you the fundamentals of why these patterns occur.
But surprise, this chart is generated by a random sequence of numbers. Zoom in to 100% and look at the cells in column A. You will see random functions.
The chart is generated by the following rules: a "1" means an uptick, a "0" means no change, and a "-1" means a downtick. These 1's, 0's, and -1's, were than continously summed, each cell representing an increment of time. The resulting sum was the cumulative price for that point in time. The results were than graphed as seen in the chart.
Interesting, isn't it? The chart resembles a "real" chart, but the whole thing was generated by a random sequence of numbers!
There has been an on going debate among academics about the nature of the real market. Is it just pure randomness or is it driven by other deterministic factors?
Well, evidence-- historical data-- shows it is not just a random process. There are too many things that occur in the market that do not agree with standard finance models, such as ones that use the normal bell curve to describe the market. If you want more info on this, read MandelBrot's, The (MIS)Behavior of Markets. (Mandelbrot is the discoverer of fractals.) Reading this book is where I got my inspiration to make this excel chart.
Now, whether or not the real market is random, pseudo-random, or all of it is driven by fundamentals, doesn't really matter for a trader or one who writes ea's. What matters in my opinion is that trends occur in the real charts and even in random charts. The trend is your friend and it is the core reason why you can make any money trading, no matter what system or method you use. Think about, you go long or short, but you only make money if the price continues in the direction you initated the trade. If the market trends, you make money no matter if it trends for 1 minute, 5 minutes, 1 hour, or a day. It's the only way you can make money.
Now, here is my philosophy of using EA's. First, I think they are mis-used when they are over-used. I am very leery of any EA system that has the EA going 24 hours a day, 7 days a week. This is stretching the "artificial intelligence" of the EA to its extreme, and beyond. Read how many commercial EA's out there that make spectacular claims, that bomb out in the long term. There are many garden variety of them.
We need to remember an EA is a computer program. It uses Boolean logic to "think". It is a machine, that uses the impedance states of transistors to store logical ones and zeroes. It could just as well, but not as fast, use mechanical switches to store ones and zeroes. Its a stretch for an EA to "read" the market like a human can. Some computer scientists might create ultra-fast systems that mimic this, but who can afford to have a Cray multi-processing computer in their house or apartment running an EA? The advantage that an EA has over a human is two fold: an EA is fast and it can store lots of information. And yes, it doesn't need sleep or food.
When I trade using an EA, I like to use the analogy of the falconer. I am a forex falconer. The ea is the falcon or golden eagle or whatever raptor you want.
How does a falconer operate? He takes his bird out for the hunt and looks at the conditions-- the weather, the sky, and then releases the bird. He does not release the bird in a storm or leave the bird for days and nights to hunt and return days later to see how the bird is doing. No, the falconer releases the bird to capture the prey only as long as the falconer needs to, to accomplish his goal of capturing a prey, and then the falconer goes home with the bird.
When I trade I look at the market conditions, not my ea, for the dips and rallies in trends at or near fibonacci levels. Then I use an EA to get into the trade to buy on a rally or sell on a dip after a moving average crossover occurs. The crossover is the trigger point in most cases for my bird to be released, my EA to take the trade, if this crossover occurs near a fibonaaci retracement. Then, I let the trade fly has long as the market conditions allow it. I use the EA because I might be asleep when the crossover occurs. If the trade reaches a point when the trend ends, my bird is tired and ready to come home. This can happen when one ma crosses over another ma or I might let it fly to a fibonacci extension. When the trend is clearly ending, I get out and stay out for the next trade when the market conditions are ready for another trade.
Whipsaw is a market storm, keep the bird home when this happens. The ma's will be a tangled mess. Cherry pick when you will release your falcon EA. A diving falcon has been clocked at 180 mph, the fastest mammal on earth. But the falcon won't do this in a storm.
Any success I have had in trading has occured when I use this one-time trading, falconer-bird partnership method, when I don't over-trade. Don't try to over-use an ea or make one that runs all the time. There might be some out there that claim to work, but has there really been a third-party objective test of their claims? Time is the final sifter that allows the true expectation value to fall out.
This has become a very interesting thread. Has opened my eyes to different angles. After all that I read, I am not sure that there is a really good ea out there. If someone had a good one, why would they share it---not. We all know that there is no Holy Grail of eas - with that said there are a lot here that look interesting and I have spent a lot of time on one of them that ended up having a deficit with the money management. I am thinking that the successful scalping eas are hard to find. If someone has a service that claims such success----they are probably making more from the sale of it than they are from trading it.
Guess I am getting tired ----must admit it----spent a lot of time on this. I still have this idea of a ea that uses stoch, macd and moving average-----used to use it on stocks when the trading was good. I am sure that someone has already made one and it is probably floating around here in this forum. It would not be a scalping ea but a little longer term-- on the four hr chart.
The excell chart was excellent.
thanks
Exactly. Nobody is going to publish a wining EA. Even if the EA is commercial and claiming it will make millions then why sell it? Why does the person that wrote the EA not just make the millions and have a happy life? Forget about all the EA's you see on the net other than to look at coding techniques and to get ideas.
On your second part the only success I have had with an EA I have developed so far is using exactly the method that you describe above - macd to find entry, stoch and ma's as filters. This is over about six months of trying to find a strategy that works without having commited one cent in the real market. And I have coded LOTS of EA's. I am currently refining the above strategy which should take me another couple of months. One important lesson that I have learned is not to mess with lower timeframes - I started at 30mins and have now moved up to 4h timeframe and get the same equity gain over time but with a fraction of the number of trades and consequently far lower risk.