There is always a lot of debate on these boards as to how one should go about evaluating and comparing certain systems.
Just starting out can be hazardously confusing and painfully account reducing.
Although I am no expert I do feel as though I have learnt a lot in these areas, both through the many pages here and other sources and thought I'd consolidate everything I've found useful into this thread.
Hopefully some of you will also find it useful, if not please feel free to shout at me for any mistakes I may have made.
1) Understanding your system
When developing your system it is essential to be able to provide suitable answers to the following questions, whether your system is discressional, mechanical or a combination of both.
A) Have I tested the system for at least 100 trades?
B) What is my average Take Profit (in pips)
C) What is my average Stop Loss (in pips)
D) What is the ensuing reward to risk ratio?
Once you can answer the above points you are ready to set out in evaluating your system.
Ok, I've tested a system. I've named it the 'Wonderpip'.
A) I tested it for 100 trades.
B) My average take profit was 100 pips
C) My average stop loss was 50 pips
D) I won 60 times, I lost 40 times. Therefore my reward to risk ratio is as
follows.
(60 * take profit): (40 * stop loss)
= (60 * 100) : (40 * 50)
= 6000 : 2000
= 3 : 1
I win 3 pips for every 1 I lose.
Golly, Wonderpip is wonderful!
2) Expectancy
Expectancy is a particularly simple calculation for determining how much you expect to earn from any given trade. It is very important because it gives you a specific confidence every time you make that next trade.
It is very easily gleaned from the above information.
Put simply the equation for expectancy is as follows
Expectancy =
(Probability of Win * Average Win) - (Probability of Loss * Average Loss)
In terms of Wonderpip this would be
(0.6 * 100) - (0.4 * 50)
= 40
we would expect to gain 40 pips every time we entered a trade.
If your system is unfortunate enough to reveal a negative expectancy it means its time to start looking for a new system.
3) Determining a risk model
So far we know that Wonderpip has a positive reward:risk ratio. We also have confidence in the fact that each time we enter the market we do so with a specific expectancy of earning 40 pips. For emphasis; this is very important because it gives you a vast amount of faith in entering your very next trade.
However, how do all these happy statistics resolve themselves in dollar terms?
Fortunately it is extremely easy to translate expectancy from a per trade figure to a per dollar one.
$expectancy
A) $ risk.
We know that on average we stand to lose 50 pips every time we enter a trade.
If we allow 50 pips of risk to be equal to $1 of risk
$1R = 50 pips
it is irrelevant what the 50 pips stand for as long as R is always equal to $1.
B) $ profit
we also know that we stand to gain an average of 100 pips every time we enter a trade.
1* profit = 100
therefore
1 * profit = 2 * risk
expressed in terms of risk this gives us
$profit (P) = $2
C) We know our win rate = 60%.
We know our loss rate = 40%
We know our sample of trades to be 100
to calculate $expectancy we do the following calculations
i) We multiply P by the number of winners = $2 * 60 = 120
ii) We multiply R by the number of losers = $1 * 40 = 40
iii) Subtract the losers from the winners = 120 - 40 = 80
iv) divide the result by the trade sample size = 80/100 = 0.8
$expectancy = 0.8
What this means is that over a large number of trades you expect to gain 80 cents for every dollar you risk.
In my opinion $expectancy is the most potent yardstick by which to measure a systems performance, since that is in effect what we are all in forex for.
Once we have found our $expectancy we can then use it to determine our rate of return over time.
This is expressed as an R multiple. (remember that for our purposes R = $1)
take our expectancy figure of 0.8 and multiply it by the number of trades in our sample (100)
0.8 * 100 = 80R
For every 100 trades we make we expect to gain 80 dollars for every 100 risked.
Trade Frequency
The final factor to consider in analysing a systems performance is frequency of trades.
You might well have an expectancy of increasing your account balance by $80 every time you complete 100 trades, however if it takes you till Armageddon to achieve that number then you're system isn't worth the time spent building it.
A system which achieves 80R in a week is 52 times more profitable than one which achieves 80R in a year.
Whether you are more comfortable with ratio's, pips per trade or return on dollars we have now determined in each of the three ways that Wonderpip is profitable.
It is also expressed in a way that has no bearing on account size, lot size or anything else.
HOORAY!
Just starting out can be hazardously confusing and painfully account reducing.
Although I am no expert I do feel as though I have learnt a lot in these areas, both through the many pages here and other sources and thought I'd consolidate everything I've found useful into this thread.
Hopefully some of you will also find it useful, if not please feel free to shout at me for any mistakes I may have made.
1) Understanding your system
When developing your system it is essential to be able to provide suitable answers to the following questions, whether your system is discressional, mechanical or a combination of both.
A) Have I tested the system for at least 100 trades?
B) What is my average Take Profit (in pips)
C) What is my average Stop Loss (in pips)
D) What is the ensuing reward to risk ratio?
Once you can answer the above points you are ready to set out in evaluating your system.
Ok, I've tested a system. I've named it the 'Wonderpip'.
A) I tested it for 100 trades.
B) My average take profit was 100 pips
C) My average stop loss was 50 pips
D) I won 60 times, I lost 40 times. Therefore my reward to risk ratio is as
follows.
(60 * take profit): (40 * stop loss)
= (60 * 100) : (40 * 50)
= 6000 : 2000
= 3 : 1
I win 3 pips for every 1 I lose.
Golly, Wonderpip is wonderful!
2) Expectancy
Expectancy is a particularly simple calculation for determining how much you expect to earn from any given trade. It is very important because it gives you a specific confidence every time you make that next trade.
It is very easily gleaned from the above information.
Put simply the equation for expectancy is as follows
Expectancy =
(Probability of Win * Average Win) - (Probability of Loss * Average Loss)
In terms of Wonderpip this would be
(0.6 * 100) - (0.4 * 50)
= 40
we would expect to gain 40 pips every time we entered a trade.
If your system is unfortunate enough to reveal a negative expectancy it means its time to start looking for a new system.
3) Determining a risk model
So far we know that Wonderpip has a positive reward:risk ratio. We also have confidence in the fact that each time we enter the market we do so with a specific expectancy of earning 40 pips. For emphasis; this is very important because it gives you a vast amount of faith in entering your very next trade.
However, how do all these happy statistics resolve themselves in dollar terms?
Fortunately it is extremely easy to translate expectancy from a per trade figure to a per dollar one.
$expectancy
A) $ risk.
We know that on average we stand to lose 50 pips every time we enter a trade.
If we allow 50 pips of risk to be equal to $1 of risk
$1R = 50 pips
it is irrelevant what the 50 pips stand for as long as R is always equal to $1.
B) $ profit
we also know that we stand to gain an average of 100 pips every time we enter a trade.
1* profit = 100
therefore
1 * profit = 2 * risk
expressed in terms of risk this gives us
$profit (P) = $2
C) We know our win rate = 60%.
We know our loss rate = 40%
We know our sample of trades to be 100
to calculate $expectancy we do the following calculations
i) We multiply P by the number of winners = $2 * 60 = 120
ii) We multiply R by the number of losers = $1 * 40 = 40
iii) Subtract the losers from the winners = 120 - 40 = 80
iv) divide the result by the trade sample size = 80/100 = 0.8
$expectancy = 0.8
What this means is that over a large number of trades you expect to gain 80 cents for every dollar you risk.
In my opinion $expectancy is the most potent yardstick by which to measure a systems performance, since that is in effect what we are all in forex for.
Once we have found our $expectancy we can then use it to determine our rate of return over time.
This is expressed as an R multiple. (remember that for our purposes R = $1)
take our expectancy figure of 0.8 and multiply it by the number of trades in our sample (100)
0.8 * 100 = 80R
For every 100 trades we make we expect to gain 80 dollars for every 100 risked.
Trade Frequency
The final factor to consider in analysing a systems performance is frequency of trades.
You might well have an expectancy of increasing your account balance by $80 every time you complete 100 trades, however if it takes you till Armageddon to achieve that number then you're system isn't worth the time spent building it.
A system which achieves 80R in a week is 52 times more profitable than one which achieves 80R in a year.
Whether you are more comfortable with ratio's, pips per trade or return on dollars we have now determined in each of the three ways that Wonderpip is profitable.
It is also expressed in a way that has no bearing on account size, lot size or anything else.
HOORAY!
quote of the day:
.
It's like deja-vu all over again