Joined Nov 11, 2007
1,458 Blog Posts

# ATR Position-Sizing and Stops Can Be Superior: Part 2

Part 1 established the following:

1. A stock’s movement is its volatility.
2. Average True Range (ATR) is a measure of volatility.
3. ATR can be used to position-size to account for volatility.

And we examined the chart below and performed some calculations…

Using the ATR(10) indicator in the bottom pane of the chart, we divided the ATR(10) by the close and multiplied that result by 100 to get a percentage.

(.803/\$33.91=0.023)*100=2.3%

It is important to for anyone wanting to explore ATR position-sizing to know how to figure out  what 1ATR equals in percentage terms. If 1ATR=2.3%, then 3ATR=6.9%. I’ll jump a couple of steps ahead and say that if a 3ATR stop is applied to MHP, it will be -6.9% beneath its close of  \$33.91

We now have most of the information we need to position-size and set a stop using ATR.

Here are the 4 necessary pieces of information to position-size using ATR:

1. ATR(10) calculation, readily available at Stockcharts. (I prefer a 10 period setting but the default setting is 14).
2. The ATR multiple, or what number you will use to multiply by the ATR (typically between 1.5 to 5).
3. The most recent closing price.
4. How much \$\$\$ to be risked. (We will use 1% of equity as our amount to be risked).

### Here is the ATR Position-Size Calculation:

(R/(ATR*M))*C=position-size

Where R=risked amount.
ATR = Average True Range with a period setting.
M= Multiple applied to ATR.
C= Most recent closing price.

Assuming account equity of 100K and 1% risk, our 3ATR position-size calculation for The McGraw-Hill Companies, Inc. [[MHP]] would look like this:

(\$1,000/(.803*3)) =415 shares

415*\$33.91 = \$14072.65 position-size

The easy part is that the stop is already calculated- it is simply the ATR multiplied by the multiple (we used 3). So the stop is .803*3 = \$2.409

### Lets look at another example:

Above is [[TPI]] , which was the PDS pick on Wednesday evening. Yes, it is another gorgeous pullback, the kind that the PDS finds daily.

Let’s run through the calculations, using the same parameters, with a 3 multiple for the ATR.

\$1000/(.322*3)=1035 shares (We would complete this calculation the night before, and put in the order for that number of shares. Remember that \$1,000 is 1% of our equity and the amount we are risking.)

1035*\$4.27 = \$4419.45 position-size (On Thursday morning, TPI opened at \$4.27).

Stop = .966 beneath the entry of \$4.27

### Let’s Make Sense of It All

Carefully note the following:

Both MHP and TPI used an ATR multiple of 3 to position-size, yet this yielded two completely different sized positions.

For example, the MHP position is over 3x larger than the TPI position, yet the TPI stop is 3x wider than the MHP stop.

MHP position = \$14,072 with a -6.9% stop.
TPI position = 4419.45 with a -22.6% stop.

What does it all mean?

The simplest way to understand it is that TPI is more volatile than MHP. Thus, we take a smaller position in TPI and use a wider stop.

Since MHP is less volatile, we might expect it to move 2% in our favor, while we might look for the more volatile TPI to move 6%. (In fact, MHP was closed on 12/24 for a 1.45% gain.)

MHP = 14,072*2% = \$281.44 profit.

TPI = 4419.45*6% = \$265.17 profit.

Even though the TPI position is smaller, we know that it can earn similar profits to the larger-sized position in MHP.

It is always fun to think about profits, but it is equally important to consider losses. If we took an equal-sized position in TPI as we did in MHP, how would it feel to have TPI move 7-14% against us? It would hurt, yet we should expect that TPI could move that quickly as its average daily range has been over 7%.

Theoretically, building positions that account for volatility help ensure that each position will add the same amount of “heat” to the portfolio. In other words, if you have loaded up with 5 volatile penny stocks, volatility position-sizing will have you buying 5 small positions, leaving a significant amount of cash. Conversely, if you want to fill your portfolio with blue chip stocks (which are typically not relatively volatile), then volatility position-sizing will have you buying a few large positions, and will probably use most of your cash. To see this balance in action, you could buy 7 positions in stocks similar to MHP, using all of your 100K. If you bought 7 positions in stocks similar to TPI, you would have used only 30% of your cash.

Finally, the ATR multiple, be it 1.5, 3, or 5, determines how aggressive your allocations are. A lower ATR multiple will result in larger positions with smaller stops.

If you have any questions or anything I have presented is unclear, please speak up in the comments section.

The PDS site includes a 2% risk, 3ATR model that earned 64.64% annualized, with an average trade profit of 1.65%.

1. Excellent stuff Wood – the only note of caution I would put on ATR-based stops is that in the situation where a stock is very, very un-volatile, you can have a situation where you make the position huge, and then a sudden burst of volatility will crush you like a bug. My trick for dealing with this is to have a maximum position size to make sure I don’t end up in a situation where a single position can take me out.

• 0
• 0
• 0 Deem this to be "Fake News"
• Damian, excellent point, and one that should not be forgotten. I will edit the post to reflect your cautionary note.

• 0
• 0
• 0 Deem this to be "Fake News"
2. thanks, wood. hard core and sensible — really appreciate the insight.

• 0
• 0
• 0 Deem this to be "Fake News"
3. If my earliest math classes were like this post I’d be smarter with the numbers today.

Thanks, Shed-ro.

• 0
• 0
• 0 Deem this to be "Fake News"
• Anton, I was a liberal arts major, but I’ve been reformed!
Math is all about the presentation.

• 0
• 0
• 0 Deem this to be "Fake News"
4. just a remark, the link for Power Dip System doesn’t work.

• 0
• 0
• 0 Deem this to be "Fake News"
• Thanks…not sure why but we’ll fix it.

• 0
• 0
• 0 Deem this to be "Fake News"
5. Wood,

I also like to use ATR in my trading, but when I place a trade, I am usually looking for 2-3 times risk to reward ratio. If I am reading you correctly in these scenarios, you are risking about \$800 and \$1000 based on your stops to gain less than \$300. Do you feel more confident with more risk because your system has 65-70% success rate? Also, are you hedging your overall market vol risk? If so how? Thanks.

• 0
• 0
• 0 Deem this to be "Fake News"
6. That situation was hypothetical in terms of risk/reward, but yes, I would risk 1000 to make 300, day after day. And the CRUCIAL element which you have addressed is percentage of winning trades. As the percentage of winning trades decreases, typically the winners become larger, just less frequent. I would expect that once the win% gets less than 60% and down as low as 50%, most systems will have a risk profile that you would like, somewhere near 2:1 or 3:1 The Power Dip, using ATR position-sizing, has a risk/reward of 1.33:1, meaning the losers are slightly larger than the winners.

I’m going to say a little more about how I build positions…

To be clear, using a percent-risk formula to build a position is just another way to skin a cat.
When using a percent-risk formula, the risked part is already figured out. You are just deciding to risk x% of your overall account on each trade. That risk could be 0.5% or 3%, or anywhere in between.

Let’s say you decide to risk 1% of account equity on every trade. If you trade a 25K account, then you are risking 250.00 You would then need a stock that would earn you 750.00 The problem is that depending on the price of a stock (cheap or expensive, i.e., 2 bucks or 50 bucks) and the volatility, you may have a hard time finding one that will meet your 3:1 ratio, and still allow for a good stop and profit target.

I think the 3:1 ratio is a good rule of thumb, especially for beginning traders, as it requires focus on the two most important aspects of trading: risk, and an exit. These two are often overlooked for focusing on the exit.

Any monkey can enter a stock. The secret sauce is knowing what to do once you have entered.

When you have an edge, you know your exit works, you know your entry works, and it is just about increasing the opportunity. Risk is managed because you know over enough trades, you will win more than you will lose (theoretically).

HTH.

• 0
• 0
• 0 Deem this to be "Fake News"
7. I sent your ATR post to Zee, Zstock7.com . Zee has similar methods/formulas of figuring trades. Zee emailed me back and said this:

“Woodshedder is a 5 star trader ( that’s my highest ranking) , I’ve followed him quite awhile.” – Zee

… Just thought I would pass this along to you .

• 0
• 0
• 0 Deem this to be "Fake News"
• Thanks Nergo! Z is a good guy as well.

• 0
• 0
• 0 Deem this to be "Fake News"
8. Wood,

A couple of questions re Power Dip. What is the avg. slippage you are experiencing with the system since you started trading it with real money? How much is it in terms of % per trade? Are you using market orders? Who is your broker?

• 0
• 0
• 0 Deem this to be "Fake News"
• Tilo, by slippage, do you mean the difference between the consolidated opening price vs. the price that I actually receive on my order? Or, do you mean the difference the price might have been had my order not existed? I can measure the former but it is very very hard to measure the latter.

My friend and I were tracking the difference between the consolidated opening price (this is the price usually listed by yahoo, google, etc. as the “open”) vs. the official exchange opening price (which is the price you would receive had you used a market order on the open). This difference is important as when backtesting, you are using the consolidated open price and not the official open price, but when trading, you do not always receive the consolidated open price but you are guaranteed by the exchange to get the official opening price.

When we first started tracking this stuff, we had logged over 300 trades from a couple of different systems and were seeing positive slippage.

We stopped tracking. Then, on my own, I started tracking this again, in November. Currently I am seeing negative slippage of -0.04%, but I do not have enough trades logged for the results to really be robust. Also, we have found (and this is verified by others) that there is very very little slippage when using market-on-close orders. The bulk of the slippage then comes from one side of the trade, which is the open.

I expect to see less slippage than what is currently being reflected as more trades are logged.

I use market orders. I use Tradestation and Interactive Brokers. Tradestation tends to have slightly better fills, especially when trading an odd-lot.

• 0
• 0
• 0 Deem this to be "Fake News"
• Thanks, Wood.

With IB, do you use market order GTC or market-on-open order? Do you actually need to be there at 9:30 to transmit the order?

Best –

• 0
• 0
• 0 Deem this to be "Fake News"
• Tilo, with IB, the order is called an OPG, short for opening. It is good for NYSE, AMEX, and NASD issues. You can place the order anytime you want, 9:30 the night before, 6:30a.m. the morning before it opens, etc. So no, once the order is placed it is held with your broker or on the exchange until the market opens.

• 0
• 0
• 0 Deem this to be "Fake News"
• (i) To make sure I understand you, you are using OPG orders, not market orders set on GTC, correct?

(ii) How do you back test your results if you are using OPG, where the prices often vary greatly from the consolidated open price?

Sorry for all the q’s, couldn’t find answers elsewhere.

• 0
• 0
• 0 Deem this to be "Fake News"
9. Tilo, OPG orders are the same as setting GTC market orders before the open. Tradestation does not offer an OPG order, neither does Scottrade. You still get the opening price. A market order placed before the open is assumed to be an opening order.

Data vendors record the consolidated open as the opening price. Exchanges report the official open as their opening price. There is no way to backtest with the official exchange opens. What you must do is track the trades to make sure that your official price does not differ significantly from the consolidated open. Right now, after recording a month or so of trades, that difference as been -0.04%, meaning that for long positions you have been paying slightly more than the consolidated open. As I stated earlier, I have also seen times over hundreds of trades where you had positive slippage, meaning you generally got prices better than the consolidated open.

I will be doing a post on this soon, and have built a project around figuring it all out.

• 0
• 0
• 0 Deem this to be "Fake News"
10. Great explanations. Well worth the investment in the subscription to the PDS!

• 0
• 0
• 0 Deem this to be "Fake News"
11. Great article.

What are your thoughts on using just an ATR(5) with a multiplier of 1 for a swing trade?

If you are entering a swing trade intraday, what values would be best to use? The ATR from the prior day since that is the only “final” ATR value available? What about price? The prior day’s close, or your entry price you entered intraday.

What about subtracting the ATR to the low instead of the close?