Most novices set their stop losses on the basis of feel. They'll pick some arbitrary percentage like 5% below entry as that "sounds about right", or they'll put the stop at some round number which they think is safe. The issue is that every stock moves in a different way. A 5% stop could be far too tight on a volatile small cap which might swing 8% on a normal day and far too wide on a very steady large cap. The ATR (Average True Range) fixes this by measuring how much the stock actually moves so your stop is based on the stock's actual volatility rather than your own guess.
ATR is the average true range which is the average size of the daily range over a period of days (usually 14). It's a true range as it takes gaps into account rather than just high minus low. If a stock has an ATR of $2 it means that on average the stock will move about $2 from high to low in a day. It tells you nothing at all about direction, just how much movement is normal for that stock at the moment.
This is absolutely useful for stops. If your stock is normally going 2 dollars a day and you've just set your stop 50 cents below your entry you’re basically asking to be whacked out by the normal amount of noise. That kind of stop will get hit all the time even if your trade is absolutely sound. The usual approach is to place your stop a multiple of the ATR away from your entry or from a key level (often somewhere in the 1.5 to 3x ATR range). The tighter the multiple the more suited it is to a quick trade and the wider the multiple the more room you give the trade to play.
Here’s a simple example. Suppose you want to buy a stock at $50 and its ATR is $1.50. Using a multiple of 2 your stop would be 3 dollars below entry at $47. You now know your risk per share is $3. If you’ve decided you’re willing to risk $100 on that trade you simply divide $100 by $3 and you end up with around 33 shares. That’s how ATR directly relates to position sizing. The more volatile the stock the higher its ATR which will give you a wider stop which will give you a smaller position for the same dollar risk. The less volatile the stock the lower its ATR and the more shares you can hold for the same risk. This automatically keeps your risk consistent from trade to trade which is precisely what you want.
It’s best to use the ATR in conjunction with price structure rather than just going with it. The ideal stop will be below a logical level (say a recent swing low or a moving average) and at the same time be far enough away that normal volatility won’t hit it. If the nearest logical level is less than one ATR from your entry then your stop is probably too tight and you may want to either wait for a better entry or simply walk away. If the logical level is five ATRs away then the trade is probably too risky to be worth it. Using the ATR in this fashion is a good way of checking whether the trade actually makes sense.
ATR also works very well for trailing stops. Rather than moving your stop around randomly you can trail it by a fixed number of ATRs below the highest close since you entered (this is often referred to as a chandelier style exit). As the stock goes up the stop goes up with it but it never moves down. This allows your winners to run whilst protecting your profits and it does it automatically. If volatility increases the stop gives the stock more room. If volatility decreases the stop comes in tighter.
You can also use ATR to set realistic profit targets. If a stock normally moves $2 a day it's probably unrealistic to expect it to gain $20 in two days. The number of ATRs it takes to reach your target is a good indicator of how long the trade is likely to take and whether the target is a credible one. Targets that are a few ATRs above entry are far more reasonable than those that require an absolutely enormous move.
Another good habit is to look at how the ATR is changing over time. A rising ATR indicates that volatility is increasing which is typically the case around the time of news, earnings or a trend break. A falling ATR indicates that the stock is calming down which can occur during a consolidation prior to a breakout. You can compare the current ATR to its recent history to determine whether the stock is in an unusually active or unusually quiet phase and that context is absolutely vital when you're considering a trade.
There are a few caveats to bear in mind. ATR is based on historical data and so is not going to be able to anticipate a sudden spike due to a piece of surprise news or an earnings gap. It doesn’t in any way replace a proper trade thesis. A stop at 2 ATRs is only going to be useful if the setup is in fact valid. Like any tool it’s more useful as a guide than it is as a hard and fast rule.
If you’re not currently using ATR then give it a go for a few weeks. Before every trade write down the ATR, your planned stop distance in ATRs and your position size. Then go back and look at how often your stops were hit by noise as opposed to how often they were hit because the trade was actually wrong. Most people find that switching from arbitrary percentage stops to volatility based stops significantly reduces the number of unnecessary stop outs and makes their risk far more consistent.