In the last two posts of this series on trading tactics, I discussed overall money management; the allocation of funds among various accounts, financial instruments, and strategies. That was the big picture. Now it’s time to zoom in and concentrate on just a single account and its individual positions. Specifically, we’ll be looking at ways to manage stop loss orders and position sizes in order to control risk within a forex account.
I’ll be using my own methods as an example. However, that doesn’t mean that my way is the only way to control risk. It’s just the method that makes the most sense to me, and lets me sleep at night. You may want to tweak, modify, overhaul, or completely ignore this method in favor of your own risk control strategy. That’s ok. The examples I give here will just serve as a starting point. They also serve as a guide to those members of CapTraderPro that want more insight into how I’m sizing my initial positions each week.
Obviously, before applying risk control measures to a position, we need to decide what position(s) to take in the first place. That’s the subject of having a directional edge, which I discuss in the “Trading Profitably” section of the site. So check that out if you haven’t already.
The classic approach
Anyway, once you’ve decided on a position, you need to decide two things related to risk control. The first is where to place your stop loss order, and the second is your position size.
The order of these is important! A common mistake made by many new traders is to pick a position size first. Then, based on how much they’re willing to lose on the trade, they decide where to place their stop. This is backwards! The market doesn’t care how many pips you can lose on a trade. Your stop needs to be placed at a technical level that makes sense based on the market’s behavior, not based on what’s in your account.
So let’s start by looking at what I’ll call the “classic” approach to position sizing (not the method I use, by the way). This consists of a rule regarding what percentage of your account value you’re willing to lose on a single trade, picking a stop point based on technical analysis, and using those two factors to determine your position size. Here’s an example of how that works.
Common figures thrown around by traders for the acceptable loss on a single trade typically range from about 1% to 5% of total account equity. I’d guess that the most common one I’ve seen is 2% of the account’s value. So let’s say you have an account value of $1000. That means that you’d be willing to risk losing $20 on a single trade.
Now suppose that you want to go long EUR/USD at the current price of 1.2000. You look for a convenient level below that for your stop, and find an area of heavy consolidation between 1.1850 and 1.1900. You decide to place your stop below that at 1.1800. So based on your likely entry price of 1.2000 and the stop loss price of 1.1800, you’re facing the possibility of losing 200 pips on the trade. So you need to size your position so that 200 pips equals about $20.
Most forex trading platforms will have a built-in calculator that allows you to see the risk of a potential trade based on your entry and stop loss points. So in practice, you can just adjust your position size until your risk amount is correct, and then place the order. But just for completeness, let’s do the math here.
If you buy a single euro for $1.20 and sell it at a loss for $1.18, your risk is 2 cents per euro. So if you want to risk $20, you’d need $20/$.02 = 1000 euros. That’s your position size for the trade.
Let me stop here for a moment to touch on something I haven’t mentioned yet; your account leverage. This is simply the factor by which you multiply your account value to find your maximum allowable trading size. But this isn’t as important a topic as many traders think it is. If you manage your position risk using this classical approach, you’ll rarely (if ever) approach the limits of your account leverage. It’s just a non-issue.
But now let’s turn to a real issue with the classical approach. The problem with sizing positions this way is that, if you have a good directional edge, it’s too timid. It leaves far too much money on the table, and is thus an inefficient use of your leveraged resources. So what’s the solution to this?
The Kelly Solution
A logical way of trying to optimize position sizes is to base them on the profitability of your trading strategy. Basically, the better your trading strategy is, the higher your position size should be. The Kelly Criterion provides a way to quantify this. Here’s a basic explanation from Investopedia.
Another way of remembering the Kelly formula is that it’s simply “edge/odds.” Your trading edge, or expectancy is:
[(win %) x (avg win) – (loss %) x (avg loss)] / (avg loss)
This is a way of saying that for every trade, you risk your average loss amount, and you either lose it or you win some percentage of it.
Your “odds” is just your average gain divided by your average loss. This “edge/odds” formula is equivalent (after a little algebra) to the formula given in the Investopedia article.
For example, as of this writing, after nearly 50 trades using my current edge, my win rate is about 66%, so my loss rate is about 34%. My average win is 2.66% of my account value, while my average loss is 0.98% of my account value. So my edge is:
[(66%) x (2.66%) – (34%) x (0.98%)] / (0.98%) = 1.45.
In other words, for every dollar I risk, I expect to get it back with an additional 45 cents to boot. My “odds” value is just (2.66)/(0.98) = 2.71. So my Kelly criterion value is 1.45/2.71 = 53%.
As an exercise, you may want to confirm that the formula in the Investopedia article gives the same answer.
Ok, great! So I’ll just go ahead and risk 53% of my account on every…wait, what??!! Are you nuts? I’m not doing that!
This brings us to the next problem. The Kelly criterion recommends unreasonably high position sizes, especially in the case of a pretty good historical trading edge like mine. So next, I’ll describe my own solution.
First of all, I turn the concept of diversification discussed in that Investopedia article on its head. The article uses the Kelly value to determine the number of positions in the account. The example they use is that if your Kelly criterion value is 5% then you should have 1/(5%) or about 20 positions in the account. So with my Kelly value of 53%, should I only have 2 positions in my account?
The answer is no, because I have a known value for the maximum number of positions in my account. It’s four. That’s because I only trade the eight major currencies, currencies trade in pairs, and I try not to have the same currency in more than one position. So for example, right now I have postitions in GBP/USD, EUR/JPY, NZD/CHF, and AUD/CAD. Eight currencies, four positions, no currency in more than one position.
What I do instead is use the Kelly value for the total risk in my account, not for single position sizing as described in the article. Since I can have up to four positions in the account, I’m only willing to risk 1/4 of the Kelly value on any given position. So instead of risking 53%, I’m only willing to risk about 13% on a position.
That still sounds like a lot, but there’s one more risk reduction step. Remember that position size is inversely proportional to the distance to the stop loss point. So the wider the stop, the smaller the position. Thus I use extremely wide stops. I look at a weekly bar chart of the pair over a period of 3-4 years (that’s about the maximum zoom of my trading platform). I then simply use the high or low extreme over that period as the stop. If the pair is at one of the extremes, I just use the price range of the period as the stop distance. When I say extremely wide stops, I mean it.
Does this still sound risky? Well it would be if I managed the trade in the classic way; that is, waiting for it to either hit a profit target or the stop. But that’s not how I trade. The stop is there only to manage the risk of an extreme “black swan” type event. The lifetime of my trades is not based on the stop (and I don’t even use profit targets at all). I open and close trades based on my directional analysis of the pair as I describe in the “Trading Profitably” section.
When I’m in a position, and my weekly analysis no longer supports the direction of the position, I do two things. First, I tighten up the stop by moving it from the far reaches of the galaxy to a much more conventional nearby technical level. Then, if price happens to still be going in my direction, I trail the stop agrgressively behind the price, using moving averages and/or local highs or lows.
So for example, suppose I wanted to enter a long EUR/USD position at 1.2000 as in the “classic approach” example above. If my account value were $1000 as in that example, and my Kelly criterion value were 53%, I would first divide that by 4 to get about 13% or $130 total risk for the position. Looking at a long-term weekly chart, I see that the low for EUR/USD was about 1.0300, so to make the math easy I put my stop at 1.000.
I’m (very improbably) “risking” 2000 pips on the trade, which should translate to $130. If I buy one euro at $1.20 and sell at a loss for $1.00, then I’m risking 20 cents per euro. Thus, my position size should be $130/$0.20) = 650 euros.
Notice that this is actually smaller than the position size in the “classical approach” example. But notice also that my probability of being stopped out is extremely small, since the stop is so far away. This is a smaller trade, but it has a lot of room to breath. Since my directional edge is pretty good, the trade has a high likelihood of either going my way or at least not going too far in the wrong direction before my analysis changes and I tighten up the stop.
Often what happens is that my analysis changes at some point, I tighten the stop, and begin trailing it behind the price. Then the pair continues to move in my original direction (because my analysis generally leads price by a matter of weeks), and I continue to trail the stop well into profitable territory before the trade finally gets stopped out.
But all of this talk about trailing stops and so forth is really more suited for a discussion of trade management. And that’s what I’ll cover in the next post. So until then, be sure to control your risk and…keep pipping up!
NOTE: As always, comments are closed due to WordPress spambots. However, to comment on this or any other post, just go to my Forex Factory journal.