Why only 66% approximately?

[DashRiprock]

there is so much crap on these forums I find it offensive to joined-up thinking.

 

[BeginnerJoe]

Crap is good. If you load up enough of it, you can sell it for real cash.

 

[trader-john]

I keep hearing about traders acheieving only around 60-70% success rate with their trading. Why is this the case ? What are the factors that one should be looking out for to increase these odds?

Can they be improved upon?

It would be great to hear from traders who have overcome these kind of problems.

Billy, if you have 60% win rate - its more than enough. If you place 10 trades, 6 will close with TP and 4 with SL. If you have equal SL and TP, you have 2 profitable trades!!!

 

[JRP2891]

Yes, it can. My system runs between 93% to 97% accuracy to a specified target and under a defined MAE. Occasionally, the systems sees what I call the Out of Phase Performance Bracket, where it takes a rapid nose dive down to near 89% to 93%, before popping its nose back up, hammering the throttle forward and soaring into the high 90's again. Of course, it cycles back down to near the 93% and then levels off for a while. It has been following this cycle/patter for the past three (3) years now and I've decided to freeze the system and not over-tweak or over-optimize it to the point of breaking it.

I can live with the performance life cycle at this point and have actually developed a negative performance break-out theory that I am now testing, where I will use the failed trade patterns to artificially boost the overall trade performance. I'll do this by allowing the system to run, but turning "Off" the actual trade signal, when the system gets beyond its historical breaking point 97%. I'll then wait for the system to fail into its Out of Phase Performance Bracket (the high 80's) and then allow the trade signal come come alive again and take what should be a high probability trade sequence back up the ladder to the high 90's.

It should have an afterburner affect on the overall system accuracy. At least, that's my theory until the results dictate otherwise. Yes, it can be done, but not with inside the box thinking and blinders on, which is what you will find on most trading boards these days, unfortunately.

Example: You have so-called "experienced traders" who still don't know how to trade a simple Moving Average Crossing methodology, to its maximum degree of profitability. An MA Crossing methodology is not the most efficient way to trade, but its extremely simple, easy for the Newbie to get their head around and profitable when done properly - though not as profitable as a true Predictive Model for trading that actually works.

So, I would say, be careful when you ask for "expert" advice online and make sure that the people giving you advice, know what they are talking about AND can back it up with proof. I have on thread for Newbies on an MA Cross method that will be clearly demonstrated to work - to the degree that such a method can be profitable. It is a very good method for teaching some very important lessons to the Newbie, such as Magnitude and the importance of knowing that there are certain patterns in the market that love to repeat themselves on a very frequent basis. Those lessons can be carried over into more advanced forms of trading system development down the road and such lessons, once learned, can pay significant benefits in your research.

Learning about pattern recognition and the concept of magnitude early on in one's trading career, is priceless.

Sounding suspiciously like more vendor bullsh1t.

 

[Billy Gates]

Yes, it can. My system runs between 93% to 97% accuracy to a specified target and under a defined MAE. Occasionally, the systems sees what I call the Out of Phase Performance Bracket, where it takes a rapid nose dive down to near 89% to 93%, before popping its nose back up, hammering the throttle forward and soaring into the high 90's again. Of course, it cycles back down to near the 93% and then levels off for a while. It has been following this cycle/patter for the past three (3) years now and I've decided to freeze the system and not over-tweak or over-optimize it to the point of breaking it.

I can live with the performance life cycle at this point and have actually developed a negative performance break-out theory that I am now testing, where I will use the failed trade patterns to artificially boost the overall trade performance. I'll do this by allowing the system to run, but turning "Off" the actual trade signal, when the system gets beyond its historical breaking point 97%. I'll then wait for the system to fail into its Out of Phase Performance Bracket (the high 80's) and then allow the trade signal come come alive again and take what should be a high probability trade sequence back up the ladder to the high 90's.

It should have an afterburner affect on the overall system accuracy. At least, that's my theory until the results dictate otherwise. Yes, it can be done, but not with inside the box thinking and blinders on, which is what you will find on most trading boards these days, unfortunately.

Example: You have so-called "experienced traders" who still don't know how to trade a simple Moving Average Crossing methodology, to its maximum degree of profitability. An MA Crossing methodology is not the most efficient way to trade, but its extremely simple, easy for the Newbie to get their head around and profitable when done properly - though not as profitable as a true Predictive Model for trading that actually works.

So, I would say, be careful when you ask for "expert" advice online and make sure that the people giving you advice, know what they are talking about AND can back it up with proof. I have on thread for Newbies on an MA Cross method that will be clearly demonstrated to work - to the degree that such a method can be profitable. It is a very good method for teaching some very important lessons to the Newbie, such as Magnitude and the importance of knowing that there are certain patterns in the market that love to repeat themselves on a very frequent basis. Those lessons can be carried over into more advanced forms of trading system development down the road and such lessons, once learned, can pay significant benefits in your research.

Learning about pattern recognition and the concept of magnitude early on in one's trading career, is priceless.

Great post. Are you trading on Auto? I ask this because I know another trader who has an automated system and he also has a high rate of success and another thing that you mentioned also..once his analysis falls below a certan level, he also cuts of but demo trades in the and anlalyzes his trades to see if he can improve them.

It seems automation makes the whole thing far more possible.

 

[TheBramble]

If you have equal SL and TP...

Cut your profits. Let your losers run.

 

[BeginnerJoe]

Sounding suspiciously like more vendor bullsh1t.

I am slightly more open to possibilities. I'd be more interested to see proof of assets. At 90%+ win rate, I'd expect a lot of collaterals. So, let's see the beef, and the cow farms.

 

[BeginnerJoe]

It seems automation makes the whole thing far more possible.

I am a computer expert. I can tell you that if you can't trade, even the biggest computer in the world won't do jack for you. The good stuff still has to come from within yourself.

 

[TheBramble]

I am slightly more open to possibilities.

We all were. The first time. Before TV was banned.

 

[Billy Gates]

I am a computer expert. I can tell you that if you can't trade, even the biggest computer in the world won't do jack for you. The good stuff still has to come from within yourself.

Yes I agree.

 

[nunrgguy]

Slowly looks up, blurry eyed, weary and tired.

What?
What's that?

Oh nothing really, just the return of old MP bu1l**** baffles brains.

Head slumps.

I'll go back to watching tumbleweed.

 

[TheBramble]

there is so much crap on these forums I find it offensive to joined-up thinking.

That's precisely why I come here. I thought you did too?

Free association for freely associating ideas totally bereft of any obvious application to trading. In essence, the perfect way to relax.

Consider it an antidote to the wide ranging intellectual exertion, sustained and intense focus and the prolonged application of heightened levels of intelligent awareness (aware intelligence?) we maintain whilst actively trading.

 

[TraderNumber7]

Great post. Are you trading on Auto? I ask this because I know another trader who has an automated system and he also has a high rate of success and another thing that you mentioned also..once his analysis falls below a certan level, he also cuts of but demo trades in the and anlalyzes his trades to see if he can improve them.

It seems automation makes the whole thing far more possible.

The trade signal is automated, but execution is still manual. Right now, I've migrated 65% of my capital onto an institutional platform (I use a Prime Broker). 30% is in Gold and Platinum, with the remaining 5% still sitting on two retail platforms.

I'm working on a full conversion from the prototype into a C++ environment, which will open up the door to automated trade execution as well. On the 5% still in retail, I do have full automation from signal to execution to close. That is an FX Dialogue solution integrated with my prototype Excel based trading engine (system) through MT4, and it runs pretty smoothly as long as MT4 stays connected.

Quite frankly, I would say that automation makes things a lot easier, but the signals being automated have to be good, of course. It also forces a level of consistency and discipline that cannot be replicated manually.

 

[TheBramble]

Qed

 

[TraderNumber7]

MousE

 

[BeginnerJoe]

I'm working on a full conversion from the prototype into a C++ environment, which will open up the door to automated trade execution as well.

What platform/language are you migrating from ? Why only C++ will open up auto execution ? I last developed a FIX execution engine in C#. I could easily have done it in C++. But I randomly chose C#. I could also have chosen some other languages that I can't remember what.

 

[TheBramble]

I last developed a FIX execution engine in C#. I could easily have done it in C++. But I randomly chose C#. I could also have chosen some other languages that I can't remember what.

This epitomises the current attraction of this site for me. You can’t deliberately come up with stuff like this. It’s a gift. A true gift.

 

[BeginnerJoe]

This epitomises the current attraction of this site for me. You can’t deliberately come up with stuff like this. It’s a gift. A true gift.

I am only doing it to hook out the collaterals that I find most interesting in any situation. I prefer a bee line straight to the business end.

 

[megamuel]

I keep hearing about traders acheieving only around 60-70% success rate with their trading. Why is this the case ? What are the factors that one should be looking out for to increase these odds?

Can they be improved upon?

It would be great to hear from traders who have overcome these kind of problems.

Billy,

I would not see 60% win rate as a problem, as you describe it. Even if you traded with a 1:1 risk:reward (stop loss and take profit equal distance from entry), you'd still make a good profit over time. Check out the attached image. It is 10 random equity curves based on a 60% win rate at 1:1 R:R. Are you saying you'd not be happy with an equity curve like that?

I don't know the maths but I think you have to have risk 1, reward under 0.7 for a 60% win rate not to be profitable. E.g. TP at 70 pips and SL at 100 pips.....

 

[tomorton]

Billy,

I would not see 60% win rate as a problem, as you describe it. Even if you traded with a 1:1 risk:reward (stop loss and take profit equal distance from entry), you'd still make a good profit over time. Check out the attached image. It is 10 random equity curves based on a 60% win rate at 1:1 R:R. Are you saying you'd not be happy with an equity curve like that?

I don't know the maths but I think you have to have risk 1, reward under 0.7 for a 60% win rate not to be profitable. E.g. TP at 70 pips and SL at 100 pips.....

Oh yes! This does it for me.