Risk Management: Position Sizing Using ATR (Average True Range)

One of the biggest dangers of trading is blowing up your account. Traders only blow up for one reason: lack of risk management. A trader who wisely uses risk management techniques will trade for a lifetime, through the ups and downs of the markets and their trading system.

One of the keys to a good trading system which employs proper risk management is making sure you size each trade position correctly.

Buying too much can mean either too much trading (if you use close stop losses), or large losses on each trade. Buying too little means not taking proper advantage of each opportunity.

There are several easy methods you can use to calculate the proper size each position you enter.

In this post I will specifically talk about position sizing using the Average True Range ("ATR") of a security.

To get your position size, it is not necessary to understand exactly how the ATR indicator is calculated. But it is important to know what ATR does.

Average True Range is a calculation which, in effect, measures the dollar amount the security in question moves during the trading day, which is then averaged over a set number of trading days.

Average True Range is essentially a volatility measurement which changes from day to day.

Using ATR to size positions means the risk is based on recent volatility. A more volatile security will have a smaller position size while a less volatile security will have a larger position size. That is good risk management!

Steps to Calculate Your Position Size

These are the steps to identify all the components you need to complete a proper calculation of maximum position size using Average True Range.

  1. (R) Determine the maximum amount of equity you are willing to lose for each trade. This should be based on your total account equity at the time you enter the trade. (New traders should risk less than 1 percent per trade.)
  2. (P) Identify the current price of the security. I do most of my trade entries near the end of the trading day as volume tends to be higher. If you do your calculations after hours, use the closing price of the security.
  3. (V) Calculate the previous 20-day ATR. You do not need to calculate this manually. Just use a trading platform, your brokerage account trading window, or even Yahoo Finance. I choose 20 days because it is a sufficient time period (4 weeks) to get a good feel for recent market activity.
  4. (X) Determine your ATR multiple. Your ATR multiple will form the first stop loss on your position. A more patient, less active trader could use an ATR multiple which is 5 times the current 20-day ATR. A more active trader may choose an ATR multiple which is 3 times the current 20-day ATR.

The calculation for ATR-based maximum position size is as follows:

R/(V*X) = U (Total Number of Units)

U*P = Max Position

or, in a single calculation:

[R/(V*X)]*P = Max Position

Example

This example will be a position size using the Average True Range method with GLD (SPDR Gold Trust).

  1. (R) This investor has a $100,000 investing account. His risk per trade is 1 percent. (The maximum amount of money he wants to lose if this trade goes against him is 1 percent of $100,000.) Therefore R = $1,000.
  2. (P) The closing price of this security is $113.44. Therefore P = $113.44.
  3. (V) The current 20-day Average True Range for GLD is $0.88. Therefore V = 0.88.
  4. (X) This investor is not highly active (not a day trader type), so his will use a multiple of 5 times the 20-day ATR. Therefore X = 5.

R/(V*X) = U

Calculation for U: 1000/(0.88*5) = 227.27 units

U = 227

Calculation for Max Position: 227*$113.44 = $25,750.88

The maximum position size of GLD for this trader purchasing today will be $25,750.88, or 227 units of GLD.

Based on the calculation, the trader will enter a purchase order to buy 227 units at a limit price of $113.44 per unit. On this trade, he will allocate 25.75 percent of his account equity to GLD.

The trader might choose to round this purchase down to 200 units if he wants to buy in round lots, thereby adjusting his risk down by a further 12 percent.

The initial stop loss price will be $109.04 (calculated as P-(V*X)). If the price of GLD falls to less than $109.04 near the close of the trading day, he will sell the position.

Alternatively, the trader can set a Stop Limit Order and have the trade automatically executed if the price falls to this level.

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.

Risk Management: The Key to Trading

Focus on What is Important

All traders need a trading system; a methodical, repeatable system of buying a selling a range of market securities with a long-term positive expectancy.

The major components of a trading system include:

  • Target Markets: You must monitor a portfolio of index ETFs, stocks, bonds, commodities, LEAPS, and/or currencies that you deem to be investable and manageable for your trading account.
  • Signals: You must continually look at your target markets for buy signals and sell signals based on your own thoroughly backtested method.
  • Risk management: You must carefully analyze the size of each market position relative to your portfolio equity.

Markets and Signals

Many traders, especially new ones, focus on the least important components of their trading system: the target markets and signals.

Identification of target markets can be an endless task. There are literally thousands of market securities out there and you can't follow them all.

I believe the best choice for most traders is to choose less than twenty market securities which have a broad range of correlations to each other. Some equities, some bonds, some commodities, and some currencies.

Traders are absolutely captivated by charts, trend lines, moving averages, oscillators, bands, indicators, and the other bells and whistles of trading software.

This focus on the signals is fun, but completely misguided. They can clutter your process and become too market specific. I am a proponent of easy-to-identify signals with wide applicability.

Most of the work you put into target markets and signals will only give a moderate advantage over buying a market with a bit of common sense.

Timing Periods

Within markets and signals, choosing appropriate timing periods is the most important part of this process. It is also largely a personal decision.

Systems based on short timing periods, such as short moving average systems and short period breakouts will demand a lot of concentration, a lot of trades, and typically a lot lower risk allowance per trade as more trades are likely to fail.

Longer period trading systems, such as those based on weekly or even monthly signals, long-term moving averages, and so on will be slower and easier to manage with fewer trades and more risk allowance per trade.

Understanding Losses

It is not uncommon for a successful trading system to lose money on more than 50 percent of trades. In fact, some famous traders lose money on more than 70 percent of their trades.

However, by keeping the losses per trade very small, they can still make enormous profits. A trader is likely to see large profits on maybe 10% of the trades, but those can easily make up for the losses.

Traders must be willing to take their losses quickly when they recognize a trade is moving against them. Retention of account equity is vital!

Let's make no mistakes. Trading is a very difficult game. There are a lot of people who lose a lot of money trading. Some people can't sell and crystalize their losses so they see too many traders go down to zero. Others double down on bad bets, which is even worse!

These losing traders eventually get kicked out of the game and then spend their time spreading their loss experiences on the internet. These stories of catastrophic loss have promoted a significant fear culture surrounding trading.

Fear of failure is why trading is not popular with many investors, or in the personal finance blogosphere. It is deemed to be risky, egotistical, frivolous, time consuming, and foolish.

It is important to remember that all of the people who have lost their money through trading have one thing in common—poor risk management.

Benefits of Trading

Some investors want to trade because they recognize the opportunities there. Trading has created some extremely wealthy individuals, many who have followed the same trading system for decades.

There are successful traders who let trading be the centre of their life. A good example of this might be a currency trader who follows a system of 15 minute bars. Currency markets are open 24/5, so this trader will be highly caffeinated (a good Tom Basso term)!

There are many successful traders who take things much slower. They use daily signals, or even weekly signals, and spend no more than a couple hours a week executing their trading strategy. You always get exactly what you want.

Trading also extends beyond the reasonable limitations of buy-and-hold investing. As a trader you can access commodity markets and currency markets, both of which are not very conducive to long-term buy-and-hold strategies.

These two markets, currencies in particular, are also great because you can trade both sides of the position (long and short) without any inherent penalty for shorting.

Commodities have a slight long-term upwards bias based on inflation. Currencies are all priced relative to each other in a pair format. There is no inherent upwards bias in the currency pair market.

Betting on both sides of currencies and commodities can offer great portfolio protection and a unique return stream that is not correlated to equity markets.

Stocks on the other hand have a strong upwards bias. Short-selling (betting against) equities is extremely difficult. Even many of the best traders in the world only short equities to hedge long equity positions.

Trading is a humbling experience that is far from inherently risky. The logical analysis and dedication required for successful trading is far from foolish.

In trading, just like in life, you will get exactly the experience and results you want.

Proper Risk Management

Traders who are successful have an extremely stringent risk management process which they follow meticulously.

Regardless of the market being trading, whether its gold or Japanese yen or Apple shares or marijuana stocks, the risk management process should be exactly the same. Only the numbers themselves change.

Risk management focuses on two key areas that work closely together to determine your risk per trade:

  1. How much to buy, and
  2. When to take a loss if the trade moves against you.

Even great trading systems will have the positions move against you 50 percent of the time, or more. You have to know when to take the loss!

Understanding (R), Your Risk Per Trade

The first step of risk management is calculating how much of your account equity you are willing to risk for each trade. This means the amount of money you are willing to lose on the trade if it goes against you.

Static Percent of Equity

Setting your risk per trade is generally done best when it is based on a percentage of your account equity.

Use of current account equity is very important! This means new positions will increase when your account is going up and decrease when your account is going through a drawdown.

A beginning trader should risk no more than 1 percent of account equity on each trade. A more experienced trader may risk as much as 3 percent of account equity per trade.

Only the highest conviction systems used by extremely experienced traders should ever risk more than 3 percent of account equity on each trade.

Kelly Criterion

A more sophisticated way of calculating risk is using the Kelly Criterion. This method is mathematically optimized for maximum gains and "perfect" position sizing.

Pure Kelly Criterion trading is highly aggressive with large drawdowns. Even very experienced traders who use Kelly will throttle it down to "half Kelly" or something similar.

This advanced system requires extensive backtesting of your markets and signals. You must understand your system's win percentage and the average size of the wins compared to the losses for each market.

The problem is that markets are not mathematically certain going forward, even with the most thorough backtests over long time periods. The best data you can obtain is still just a reasonable estimate.

Personally I think the Kelly Criterion is somewhat flawed for market trading because markets are not purely mathematical—there is a human element. Markets change over time. Past performance of a system does not predict future results or performance for the exact same trading system going forward.

While Kelly Criterion is worth understanding, I believe it is better applied in pure mathematical situations where the win percentage and size of wins in a situation are fixed, not estimated.

A good example of position sizing using Kelly Criterion where your inputs are mathematically certain might be a bet at a card game like poker or blackjack.

How Much to Buy

The size of each position at entry will be based on your risk of equity per trade and how your trading system is set up to exit a position that moves against you.

As I stated above, a new trader should risk less than 1 percent of their equity per trade while an experienced trader might risk as much as 3 percent of their equity per trade.

Personally I use 2 percent of equity per trade as my risk number for the calculations. Then I round down the position size somewhat when placing the order.

There are a number of ways to establish a position exit point on new trades. Some ways of calculating exit points to determine initial position size include:

  • Average True Range
  • Percent Risk
  • Box Lows
  • LEAPS Systems

I will go over each system method with the corresponding calculations in a series of future posts.

When to Take the Losses

Systematically selling a position that is going against you is critical to your risk management! You must be willing to take losses!

There is little point to position sizing based on a predetermined exit point if you don't sell when the market moves down to that exit point because of a psychological impediment.

Often times, your trading system will have you taking several small losses in a row before one of two of your positions start making a good profit. Sometimes those repeat small losses will happen several times on the exact same position!

This is where the importance of backtesting and understanding your trading system comes into play. There is little new under the sun.

When you get discouraged because you have been stopped out of a position several times in a row, look back into your backtests to find a similar situation. Many times you will find shortly after a market whipsaws several times, that same market had a fantastic and highly profitable move.

Remember, trading is a long-term mathematical game. It appears more doable when thought of in this sense.

If you have a system that uses 2 percent risk per trade with a 50 percent success rate on each trade, you only need to make an average profit equaling a little more than 2 percent of equity on the 50 percent of successful trades to break even overall after trading costs.

On this same system, if your average profit per successful trade is 3 percent, you will have a net positive return of nearly 1 percent of equity per trade.

If you trade once per year, your account will grow at around 1 percent per year. If you make ten trades per year, you account will grow at 10 percent per year on average.

The key to a good trading system is making sure you have enough equity to make the next trade. Over time, if your system has a positive expectancy, you will be successful. This is why risk management and taking small losses is so important.

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.