Uploaded by OneStepRemoved on 30.03.2012

Transcript:

Hi everybody this is Shaun Overton with OneStepRemoved.com

in this video I am going to give you a light introduction

to our position size analysis software the goal is just to show you

over if you know what the expectation of you system is in terms of

percent accuracy and payouts we can give you a model of what kind of returns you

might expect purely due to random chance I am going to go over the parameters in

this software and then in future videos we are going to go over different combinations

that you can consider to analyze your trading system and even random luck so to

start the first Colum here is the initial balance and that ought to be straight forward

it’s the amount of money in your account the next one is the winning percentage

I think that’s also straight forward here we are going to win 50% of the time

and we lose 50% of the time the R multiple is the ratio of the wins to the

loses that’s the amount you win divided by the amount that you lose in this case

I am setting it to 1 and if you are still not comfortable with that you can take it

straight from your backtest if you are familiar with NinjaTrader you will notice that

there is an analytic here that says the “ratio average win / average loss“ in this

case it's 0.68 that is the R multiple it’s exact same thing here it’s we win dollar we

lose a dollar in this next tab it's the amount of risk so for a $100,000

account balance we are risking 1% that means when we win we are going to

make a 1000bucks and if we lose we are going to lose a 1000bucks if we

set it to 2% it means we make 2grand or lose 2grand est. the next

two tabs are a little more tricky the “trades per run” and the “number of runs”

so the trades per run is analogous to a coin toss game so with our 50-50 odds and

1 to 1 payout I am going to flip a coin 200 time whenever I come out

heads I am going to record that I up the balance by a grand and whenever I come

out tails I deduct a grand after I have done that 200 time I will record the

final balance so maybe it comes out to $95,000 I record that and I save it now I am

going to go on to the second trial I flip a coin 200 times I record the change in

the account balance over and over again once I get to the final result then I am

going to write it down on my list say we round up at $102,000

now I am going to do that 2,750 times that if you

notice that this times this is going to wind up to the number of random numbers in

our file and what we want is we want to make this number as large as possible

in order to give us the expectations over certain number of trails to put that in trading

terms if you do 200 trades and you have a sad expectation that might be

the equivalent of your trades per year if you trade less that once per day if you

day trade and you do 5 trades per day then this is the expectations that you would

expect after 30 trades per week times 10 weeks is about 300 so about

6 ½ weeks is what you expect you return to be after this much trading its going to

fluctuate due to random chance and that’s the whole goal of using this software

the last thing I want to explain is how the random numbers in the software work

we are reading them from a file and the file was generated from atmospheric noise

so what that means is you just take the noise you come up with a bunch of bits and

you turn it into a random number between 1 and 10,000 we read the first

number for our trade and we say okay we have a 50% winning percentage

if the number is between 5,000 and 0 then we are going to accept that as

a winner and if the number is above 5,000 we will through it out as a loser

if we change our winning percentage to 65 then we would say if the number

is between 1 and 6,500 it’s a winner and if it’s more than it’s

a loser and so on and just by having that random chance we can very accurately

model the accuracy and because we control the account balance we can come up

with very accurate expectations on how random luck can either work for us

or against us and I am going to push calculate just to put the number up on the graph

and you will see here that we did our analysis on over ½ a million random

numbers in this case 550,000 with numbers between

1 and 10,000 and it give us this nice little statistical analysis that verifies the

results that we would expect so the average winning run is exactly 2 and the

average losing run is almost exactly 2 as well that's perfectly balanced which is what

we expect here you have a random run of 17 consecutive winners purely due

to random chance with 19 consecutive losers this number is never going to

precisely equal 50% it should just be very close and here we are accurate within

hundreds of a decimal place its very good and it's enough for the analysis

so I will leave the video here and in the next video we are going to actually talk about these out comes

and how a 50-50 coin toss game can make or break your account if you are risking too much

My name is Shaun Overton, Thank You for listening.

in this video I am going to give you a light introduction

to our position size analysis software the goal is just to show you

over if you know what the expectation of you system is in terms of

percent accuracy and payouts we can give you a model of what kind of returns you

might expect purely due to random chance I am going to go over the parameters in

this software and then in future videos we are going to go over different combinations

that you can consider to analyze your trading system and even random luck so to

start the first Colum here is the initial balance and that ought to be straight forward

it’s the amount of money in your account the next one is the winning percentage

I think that’s also straight forward here we are going to win 50% of the time

and we lose 50% of the time the R multiple is the ratio of the wins to the

loses that’s the amount you win divided by the amount that you lose in this case

I am setting it to 1 and if you are still not comfortable with that you can take it

straight from your backtest if you are familiar with NinjaTrader you will notice that

there is an analytic here that says the “ratio average win / average loss“ in this

case it's 0.68 that is the R multiple it’s exact same thing here it’s we win dollar we

lose a dollar in this next tab it's the amount of risk so for a $100,000

account balance we are risking 1% that means when we win we are going to

make a 1000bucks and if we lose we are going to lose a 1000bucks if we

set it to 2% it means we make 2grand or lose 2grand est. the next

two tabs are a little more tricky the “trades per run” and the “number of runs”

so the trades per run is analogous to a coin toss game so with our 50-50 odds and

1 to 1 payout I am going to flip a coin 200 time whenever I come out

heads I am going to record that I up the balance by a grand and whenever I come

out tails I deduct a grand after I have done that 200 time I will record the

final balance so maybe it comes out to $95,000 I record that and I save it now I am

going to go on to the second trial I flip a coin 200 times I record the change in

the account balance over and over again once I get to the final result then I am

going to write it down on my list say we round up at $102,000

now I am going to do that 2,750 times that if you

notice that this times this is going to wind up to the number of random numbers in

our file and what we want is we want to make this number as large as possible

in order to give us the expectations over certain number of trails to put that in trading

terms if you do 200 trades and you have a sad expectation that might be

the equivalent of your trades per year if you trade less that once per day if you

day trade and you do 5 trades per day then this is the expectations that you would

expect after 30 trades per week times 10 weeks is about 300 so about

6 ½ weeks is what you expect you return to be after this much trading its going to

fluctuate due to random chance and that’s the whole goal of using this software

the last thing I want to explain is how the random numbers in the software work

we are reading them from a file and the file was generated from atmospheric noise

so what that means is you just take the noise you come up with a bunch of bits and

you turn it into a random number between 1 and 10,000 we read the first

number for our trade and we say okay we have a 50% winning percentage

if the number is between 5,000 and 0 then we are going to accept that as

a winner and if the number is above 5,000 we will through it out as a loser

if we change our winning percentage to 65 then we would say if the number

is between 1 and 6,500 it’s a winner and if it’s more than it’s

a loser and so on and just by having that random chance we can very accurately

model the accuracy and because we control the account balance we can come up

with very accurate expectations on how random luck can either work for us

or against us and I am going to push calculate just to put the number up on the graph

and you will see here that we did our analysis on over ½ a million random

numbers in this case 550,000 with numbers between

1 and 10,000 and it give us this nice little statistical analysis that verifies the

results that we would expect so the average winning run is exactly 2 and the

average losing run is almost exactly 2 as well that's perfectly balanced which is what

we expect here you have a random run of 17 consecutive winners purely due

to random chance with 19 consecutive losers this number is never going to

precisely equal 50% it should just be very close and here we are accurate within

hundreds of a decimal place its very good and it's enough for the analysis

so I will leave the video here and in the next video we are going to actually talk about these out comes

and how a 50-50 coin toss game can make or break your account if you are risking too much

My name is Shaun Overton, Thank You for listening.