How to tell if a system works?

[Kat888]

Applying your hypothesis/ses in real time as price unfolds.

hmmm whats the point if you can just trade your hypothesis and see if it works ( ur take profit will be hit without stop losses reached before)?

 

[PieterSteidelmayer]

hmmm whats the point if you can just trade your hypothesis and see if it works ( ur take profit will be hit without stop losses reached before)?

I'll turn out the light on my way out.

 

[timsk]

hmmm whats the point if you can just trade your hypothesis and see if it works ( ur take profit will be hit without stop losses reached before)?

Hi Kat,
I think what PieterSteidelmayer was trying to express in his post is that what you suggest is akin to putting the cart before the horse. Without forward testing your hypothesis on live data - you don't know for sure that it works. That said, many traders plough ahead anyway along the lines you suggest. If you do this, be sure to trade the smallest position size your broker allows - and be fully prepared to lose your money.
Tim.

 

[Kat888]

I'll turn out the light on my way out.

I think you should use some place out for that sort of remarks. It doesn't do you any favour

 

[Kat888]

Hi Kat,
I think what PieterSteidelmayer was trying to express in his post is that what you suggest is akin to putting the cart before the horse. Without forward testing your hypothesis on live data - you don't know for sure that it works. That said, many traders plough ahead anyway along the lines you suggest. If you do this, be sure to trade the smallest position size your broker allows - and be fully prepared to lose your money.
Tim.

Tim,
it's just a waste of time or simply a sign you are not confident in what you are doing. You developed the system, back tested it, next step is actual trading. It's even more impractical when you are trying to analyse if your system works wasting your time forward testing instead of actually trading it.

 

[timsk]

Tim,
it's just a waste of time or simply a sign you are not confident in what you are doing. You developed the system, back tested it, next step is actual trading. It's even more impractical when you are trying to analyse if your system works wasting your time forward testing instead of actually trading it.

Hi Kat,
There are many confident traders who have blown their accounts. I know, I'm one of them! As I said in my last post, you can miss out the forward testing phase if you want, but it's akin to a motor manufacturer going into full production from a 3D computer model without building and testing some prototypes first. It's risky and, more often than not, it will fail. Having said all that, I completely understand your impatience and wanting to get on with it. Been there, done it, bought the T shirt.

Tim.

 

[neil]

'Kinell

Tim,
it's just a waste of time or simply a sign you are not confident in what you are doing. You developed the system, back tested it, next step is actual trading. It's even more impractical when you are trying to analyse if your system works wasting your time forward testing instead of actually trading it.

One is constantly surprised by the standard of "trader" as reflected in their abuse of logic and willingness to rush along the road to financial ruin . Still, someone is happy to be on the other side of their trades

I'll turn the lights off on the way out again

 

[ZEN archer]

Why is it so dark in here?

 

[tar]

Ofcourse if you keep tweaking the back testing results of any system then "it will work" ! Sadly you cant do that when you trade live as results cant be changed ...

 

[Kat888]

Hi Kat,
There are many confident traders who have blown their accounts. I know, I'm one of them! As I said in my last post, you can miss out the forward testing phase if you want, but it's akin to a motor manufacturer going into full production from a 3D computer model without building and testing some prototypes first. It's risky and, more often than not, it will fail. Having said all that, I completely understand your impatience and wanting to get on with it. Been there, done it, bought the T shirt.

Tim.

Markets change ok? And your system might work well during that forward test period ( OH BTW CAN ANYBODY TELL ME HOW LONG THE FORWARD TEST PERIOD IS?????)
and will fail as soon as you start real trading.

 

[timsk]

Hi Kat,

Markets change ok? And your system might work well during that forward test period and will fail as soon as you start real trading.

This is very true. However, the point that everyone is in agreement about (well, everyone but you, lol!) is that your probability of success when trading for real is much greater if you've forward tested your methodology first, using a demo account in a live market, than it is if you skip this phase altogether and jump from back testing to live trading.

( OH BTW CAN ANYBODY TELL ME HOW LONG THE FORWARD TEST PERIOD IS?????).

It depends on the trading style (day, swing or position trader), market and instruments traded and type of methodology employed. However as a very rough rule of thumb, I'd say in the region of 100+ trades for a day trader, 40-50 trades for a swing trader and 15-20 trades for a position trader. Care needs to be taken regarding the prevailing market condition during the test period. For example, if you position trade and your methodology is a trend trading strategy based on moving average cross overs, this will work well in a trending market and fall down in a sideways choppy market.
Tim.