There are many ways to go about determining how much of your capital you risk per trade. This is only one of them; this is how I do it. You do not have to do it this way.
If you trade, you need a trading plan, plain and simple. There are two main components to a trading plan. The first is a system. This consists of entry and exit rules - when to enter a trade and when to exit that trade. There are many other things that can be employed when developing a system, which can include determining optimal stops, targets, moves to breakeven, trailing stops, and a dozen other things. I will not cover them. Once you have your system defined, the next step is determining how much to risk on each trade. This is what I will cover.
I've seen too many threads on this forum where people just pick arbitrary risk values. Some people claim their system says to trade 10% per trade. Some say 20% per trade. I believe the beginners sticky thread says risk no more than 3%. So which one do you believe? None. You believe none of them. If you want something done right, you have to do it yourself. Determine your own risk per trade and you will be confidant that it's right, because it will be right. Guessing how many lots to trade is the same as guessing when to enter a trade. Read that sentence again if you need to.
The first thing you need is a sequence of trades. This comes after you develop your "system". How you get that sequence of trades is up to you. If you have a mechanical system, you may want to program it, in which case the program will be able to spit out a sequence of trades for you. I recommend TradeStation; it's features are very powerful. Another way to get a sequence of trades is to do it by hand. You've heard of manually scrolling back on charts to determine if you strategy is profitable, right? Well do the same thing and record each and every trade your system takes. Mainly what were looking for is pips profit and pips lost on a per trade basis.
So, for example, your system has a 50 pip limit (take profit) and a 40 pip stop loss, both of which spread is accounted for. Your sequence of trades could be the following:
50,50,-40,50,-40,50,-40,-40,50,50
This represents 10 trades. How many trades should you consider? Well, this is very controversial; many people use many different numbers. I think it depends on how often trade signals are generated. If you are trading off a 5 minute chart and get several signals a day, maybe a couple hundred trades. If you're trading off a daily chart and get a few signals a month, I would take around 100 trades. I wouldn't go any further back in time than a couple years, and I wouldn't use less than 100 trades. I know it's vague, but that's how trading is.
Once you have your sequence of trades you can determine the maximum drawdown of your system on a fixed lot basis. This is assuming you are trading one standard lot per trade, so a 50 pip trade would be equivalent to $500. The maximum drawdown is the maximum value your account can be "down" at any given time. For example if you start trading with $10,000 and at the worst time your account equity is $8,000, this is a 20% drawdown. If you account is $5,000, this represents a 50% drawdown. Maximum drawdown is also referred as "peak to valley" drawdown. This is the difference between the greatest peak and minimum valley on an equity curve. The following equity curve is from our trade sequence above and a starting account of $10,000:
http://i20.photobucket.com/albums/b2...nchell/ff1.jpg
The maximum drawdown is the difference between the peak ($11,200) and valley ($10,400). This is roughly a 7% drawdown.
Now, this is just so you understand what drawdown is on a fixed lot basis. This does nothing for determining your risk per trade. However, this is where you must choose how much drawdown you are willing to cope with. If you start with $10,000, are you going to have the ability to continue trading you plan if you account is drawndown to $8,000? Or will you give up and go home? Most people, me included, think a 25% drawdown is an acceptable risk. Some are willing to handle more, but most are willing to handle much less. I think this number is very representative of how confidant you are in your trading system. If it has been proven and you've been trading it live for some time, more risk equals more reward. I know one trader that has handled a 40% drawdown, and he's ok if another 40% drawdown comes. You're going to have to determine this on your own, but I would suggest 20% or less.
Now that you have a sequence of trades and know how much maximum drawdown you're willing to accept, you can determine how much of your capital to risk on a per trade basis. From here I would get a piece of software that does that calculations for you. I know, I know, right now you're thinking "so I gotta spend some money on software I'll use once". There is software out there than you can demo for more than you'll ever need it for. What this software does is spit out these optimal values so you dont have to play mathematician. Google "MSA" and you'll find what you need.
Back to our 10 hypothetical trades. How much can we risk per trade? Well, here is what our program tells us:
http://i20.photobucket.com/albums/b2...nchell/ff2.jpg
As you can see if you maximum drawdown tolerance is 20%, then it is optimal to risk 0.12 per trade, which is 12%. Here is our new equity curve with risking 12% per trade:
http://i20.photobucket.com/albums/b2...nchell/ff3.jpg
As you can see from the Y axis our account balance goes from $13,600 to $10,800. This is a 20% drawdown. This is very basic with only 10 trades.
Lets work with something more realistic. Here is a parameter study with a sample of 100 trades:
http://i20.photobucket.com/albums/b2...nchell/ff4.jpg
As you can see, if we take our 20% drawdown, we would be risking 6.93% per trade. Sure beats just guessing doesn't it?
If you trade, you need a trading plan, plain and simple. There are two main components to a trading plan. The first is a system. This consists of entry and exit rules - when to enter a trade and when to exit that trade. There are many other things that can be employed when developing a system, which can include determining optimal stops, targets, moves to breakeven, trailing stops, and a dozen other things. I will not cover them. Once you have your system defined, the next step is determining how much to risk on each trade. This is what I will cover.
I've seen too many threads on this forum where people just pick arbitrary risk values. Some people claim their system says to trade 10% per trade. Some say 20% per trade. I believe the beginners sticky thread says risk no more than 3%. So which one do you believe? None. You believe none of them. If you want something done right, you have to do it yourself. Determine your own risk per trade and you will be confidant that it's right, because it will be right. Guessing how many lots to trade is the same as guessing when to enter a trade. Read that sentence again if you need to.
The first thing you need is a sequence of trades. This comes after you develop your "system". How you get that sequence of trades is up to you. If you have a mechanical system, you may want to program it, in which case the program will be able to spit out a sequence of trades for you. I recommend TradeStation; it's features are very powerful. Another way to get a sequence of trades is to do it by hand. You've heard of manually scrolling back on charts to determine if you strategy is profitable, right? Well do the same thing and record each and every trade your system takes. Mainly what were looking for is pips profit and pips lost on a per trade basis.
So, for example, your system has a 50 pip limit (take profit) and a 40 pip stop loss, both of which spread is accounted for. Your sequence of trades could be the following:
50,50,-40,50,-40,50,-40,-40,50,50
This represents 10 trades. How many trades should you consider? Well, this is very controversial; many people use many different numbers. I think it depends on how often trade signals are generated. If you are trading off a 5 minute chart and get several signals a day, maybe a couple hundred trades. If you're trading off a daily chart and get a few signals a month, I would take around 100 trades. I wouldn't go any further back in time than a couple years, and I wouldn't use less than 100 trades. I know it's vague, but that's how trading is.
Once you have your sequence of trades you can determine the maximum drawdown of your system on a fixed lot basis. This is assuming you are trading one standard lot per trade, so a 50 pip trade would be equivalent to $500. The maximum drawdown is the maximum value your account can be "down" at any given time. For example if you start trading with $10,000 and at the worst time your account equity is $8,000, this is a 20% drawdown. If you account is $5,000, this represents a 50% drawdown. Maximum drawdown is also referred as "peak to valley" drawdown. This is the difference between the greatest peak and minimum valley on an equity curve. The following equity curve is from our trade sequence above and a starting account of $10,000:
http://i20.photobucket.com/albums/b2...nchell/ff1.jpg
The maximum drawdown is the difference between the peak ($11,200) and valley ($10,400). This is roughly a 7% drawdown.
Now, this is just so you understand what drawdown is on a fixed lot basis. This does nothing for determining your risk per trade. However, this is where you must choose how much drawdown you are willing to cope with. If you start with $10,000, are you going to have the ability to continue trading you plan if you account is drawndown to $8,000? Or will you give up and go home? Most people, me included, think a 25% drawdown is an acceptable risk. Some are willing to handle more, but most are willing to handle much less. I think this number is very representative of how confidant you are in your trading system. If it has been proven and you've been trading it live for some time, more risk equals more reward. I know one trader that has handled a 40% drawdown, and he's ok if another 40% drawdown comes. You're going to have to determine this on your own, but I would suggest 20% or less.
Now that you have a sequence of trades and know how much maximum drawdown you're willing to accept, you can determine how much of your capital to risk on a per trade basis. From here I would get a piece of software that does that calculations for you. I know, I know, right now you're thinking "so I gotta spend some money on software I'll use once". There is software out there than you can demo for more than you'll ever need it for. What this software does is spit out these optimal values so you dont have to play mathematician. Google "MSA" and you'll find what you need.
Back to our 10 hypothetical trades. How much can we risk per trade? Well, here is what our program tells us:
http://i20.photobucket.com/albums/b2...nchell/ff2.jpg
As you can see if you maximum drawdown tolerance is 20%, then it is optimal to risk 0.12 per trade, which is 12%. Here is our new equity curve with risking 12% per trade:
http://i20.photobucket.com/albums/b2...nchell/ff3.jpg
As you can see from the Y axis our account balance goes from $13,600 to $10,800. This is a 20% drawdown. This is very basic with only 10 trades.
Lets work with something more realistic. Here is a parameter study with a sample of 100 trades:
http://i20.photobucket.com/albums/b2...nchell/ff4.jpg
As you can see, if we take our 20% drawdown, we would be risking 6.93% per trade. Sure beats just guessing doesn't it?