Hi guys! I want to share my observations in my time trying to figure out the markets,
maybe it can help a few of you traders on what to think about when creating your own trading systems.
we hear every every time and all the time that the win to reward should be 2 to 1 or at least 1 to 1,
this is ideal in a perfect world but will it work? I can't tell you how many times this will not work, because there is so many factors playing in,
and you have to balance them all to get the result you try to achieve.
the 3 things are
1. win ratio
2. risk to reward
3 chance of success (odds)
not to say that these 3 are the only ones, there are plenty more
such as money management etc, but these are good to start with.
if we begin with the win ratio, let's say that you find something that work 60% of the times,
you feel that 60% isn't enough in your case because it is only 1 to 1 risk to reward,
you look at the equity curve in the back-tester and it's choppy not really going straight up as you had been hoping for.
you are surely winning but it's slow, so easy fix right? just change the risk to reward to 2 to 1 and that solves the problem,
no this actually make it even worse, because the odds that you will reach 2 to 1
rather than 1 to 1 is smaller, so now your win rate will suffer and suddenly your strategy is loosing.
so these 3 things must be very balanced, and they may differ between time frames and pairs,
and be aware that if you change a tiny detail, the whole system, every other function of your EA or manual strategy
will be out of it, so after any change you must fine tune all functions, filters, etc
so each strategy must be tailored to the instrument it is supposed to be trading,
this brings up the next problem you must avoid at any cost, and that is curve fitting,
this means that in your quest to find your holy grail, you are testing it on a part of the market,
lets say Eur usd, 3 months back, so everything look great,
and you start trading the system only to find out a week later you lost half your account,
this is because the market is always changing, and how it was 3 months ago will never happen again,
so any system must be tested over a whole range of different markets to make it more reliable,
this also will have the advantage to remove any curve fit over a period of time.
but hold on, we are not done yet, the market have different modes,
uptrend, downtrend sideways, and your EA or system must know how to trade through all the different market conditions,
or what is likely to happen is that when the market changes mode, you will give all of it back and start losing.
actually there is more than 3 things the market can do, the uptrends and downtrends are obvious but
there are different uptrends and downtrends and also different sideways conditions that all have different characteristics.
your EA must be able to trade all these market conditions, or to trade just one of them, and stay out of the rest.
do not expect to be profitable trading an trending idea through all these market conditions,
most of the systems around on the net and in the forums are trend trading ideas,
and not one of them is based on sideways market conditions, even though the market is going sideways 80% of the times,
isn't sideways what we should be trading? why trade something that only work 20% of the times?
do you know how many times you will be wrong and stopped out trading this for one year,
you will be stopped out 80% hmm didn't they say most traders are losing 80% ?
I hope this will at least help someone, good luck!
maybe it can help a few of you traders on what to think about when creating your own trading systems.
we hear every every time and all the time that the win to reward should be 2 to 1 or at least 1 to 1,
this is ideal in a perfect world but will it work? I can't tell you how many times this will not work, because there is so many factors playing in,
and you have to balance them all to get the result you try to achieve.
the 3 things are
1. win ratio
2. risk to reward
3 chance of success (odds)
not to say that these 3 are the only ones, there are plenty more
such as money management etc, but these are good to start with.
if we begin with the win ratio, let's say that you find something that work 60% of the times,
you feel that 60% isn't enough in your case because it is only 1 to 1 risk to reward,
you look at the equity curve in the back-tester and it's choppy not really going straight up as you had been hoping for.
you are surely winning but it's slow, so easy fix right? just change the risk to reward to 2 to 1 and that solves the problem,
no this actually make it even worse, because the odds that you will reach 2 to 1
rather than 1 to 1 is smaller, so now your win rate will suffer and suddenly your strategy is loosing.
so these 3 things must be very balanced, and they may differ between time frames and pairs,
and be aware that if you change a tiny detail, the whole system, every other function of your EA or manual strategy
will be out of it, so after any change you must fine tune all functions, filters, etc
so each strategy must be tailored to the instrument it is supposed to be trading,
this brings up the next problem you must avoid at any cost, and that is curve fitting,
this means that in your quest to find your holy grail, you are testing it on a part of the market,
lets say Eur usd, 3 months back, so everything look great,
and you start trading the system only to find out a week later you lost half your account,
this is because the market is always changing, and how it was 3 months ago will never happen again,
so any system must be tested over a whole range of different markets to make it more reliable,
this also will have the advantage to remove any curve fit over a period of time.
but hold on, we are not done yet, the market have different modes,
uptrend, downtrend sideways, and your EA or system must know how to trade through all the different market conditions,
or what is likely to happen is that when the market changes mode, you will give all of it back and start losing.
actually there is more than 3 things the market can do, the uptrends and downtrends are obvious but
there are different uptrends and downtrends and also different sideways conditions that all have different characteristics.
your EA must be able to trade all these market conditions, or to trade just one of them, and stay out of the rest.
do not expect to be profitable trading an trending idea through all these market conditions,
most of the systems around on the net and in the forums are trend trading ideas,
and not one of them is based on sideways market conditions, even though the market is going sideways 80% of the times,
isn't sideways what we should be trading? why trade something that only work 20% of the times?
do you know how many times you will be wrong and stopped out trading this for one year,
you will be stopped out 80% hmm didn't they say most traders are losing 80% ?
I hope this will at least help someone, good luck!
Bulls are stupid Animals!especially when Im short!