Why are bad habits so hard to break, especially in trading? To start the New Year fresh, why not take a look at your bad trading habits and make adjustments. In particular, look at your stops; are they too close or too far away? Maybe you should use a different type of stop. Applying new research in stop placement may mean the difference between a rising profit curve and bill collectors calling your neighbors for help prying the last pennies from your bank account. And this may be a good thing as the point of using stops in the first place is to protect profits and/or minimize losses.

A stop loss (stop) order is an order to sell below the current market price. The stop order becomes a market order for immediate execution when the stock drops to or below a specified price. Thus, if you place a stop order at ten dollars a share and price opens trading at nine, you’ll likely be stopped out near nine. And there is no guarantee that your order will be executed at the specified stop price.

There are many types of stops, but to keep the sophisticated traders interested, let’s talk about these four in particular: beta-adjusted trailing, trailing, Fibonacci correction and volatility.

The Beta-Adjusted Trailing Stop
The idea behind this stop is that highly volatile stocks should have stops placed farther away to avoid being stopped out on normal price movement. I call beta-adjusted trailing stops BATS for short. The idea for BATS wasn’t mine, but I tested it and refined it. Table 1 is an example of a beta-adjusted trailing stop. I used a combination of statistical methods backed by experiments to make sure the stops worked as intended.

The left column shows the current daily high price for the stock. The top row shows the stock’s beta. Beta is a measure of a stock’s price volatility against the market’s, and is available on a variety of websites; Yahoo’s finance website, Value Line, and so on.

How do you use the table? Say you buy ABC Gum for 18.50 and it has a beta of 0.75. If the stock has a high price of 19 on the day you bought, the intersection of the 19 price and 0.75 beta suggests a stop 11 percent below the current high. That would place the stop at 16.91. If the stock made a new high of 22, then you would recalculate and find the stop should be placed ten percent below the high, or at 19.80. If a stock has negative beta (meaning price usually drops when the market rises, or the reverse), then ignore the minus sign.

Table 1 shows two trends. First, as beta increases, meaning the stock is more volatile, the stop needs to be farther away. For a $5 stock with a 0.4 beta, the stop would be ten percent below the high. If beta is 2.0, then the stop would be 18 percent below the high, nearly double the earlier number. Second, as price rises, the stop becomes closer. Scan down each column and you see that the stop decreases for increasing price. The reason for this is that low priced stocks are more volatile. A $1 move for a $5 stock means a 20 percent change. A $1 move in a $50 stock means a two percent change. Finally, remember that the large percentage values help compensate for using the intraday high price, not the close. You may want to adjust the stop percentages as needed for your situation, especially if you use the intraday low or closing prices.


click image for larger view

The Trailing Stop
Trialing stops are the most popular stop type. A trailing stop is one that chases price higher. You move the stop up when price reaches a new high. Figure 1 shows an example of how this is done.


click image for larger view

Look at the inset where it shows three minor lows, pointed to by red lines. I define a minor low as a price valley separated from other valleys by at least a week. Don’t get hung up on the definition. Minor lows (MLs) should appear as price turning points, and they act as price support zones, so they make for good locations beneath which you place the stop.

As shown in Figure 1, our trader, let’s call him Jacob, bought the breakout from the descending triangle and received a fill at 43.50. Where should he place the initial stop? A good location is below the bottom of the triangle, because triangles are congestion regions that often support price. The bottom is at 41.60 for a potential loss of 4.4 percent. Since he doesn’t like round numbers (like 50, 55 or 60) where others often place their stops, he placed his at 41.53. Price throws back to the pattern (B to C) before resuming the uptrend and making a new high at A. Since A is above B, Jacob looks for a closer minor low (C). Since the low at C doesn’t differ much from his current stop price, he decides to leave the stop location alone. Price moves up and makes a new high at D (it rises above the prior high at A). Jacob raises the stop to just below the minor low at E, 42.34. He moves the stop again after price rises to F, placing the stop just below G. Price rounds over and in late May it hits his stop, closing out the position at 44.63.

Do you see any flaws with this method? Jacob bought at 43.50 and sold at 44.63. He made 2.6 percent, but the stock peaked at 50.80 and if he had sold then, he could have made 14.4 percent. In fact, location G was the only stop location where his position was profitable, despite price being above the buy for a good portion of the time.

As you search for a stop location, a minor low need not be the only pattern you place the stop beneath. Any support zone will work. A study I conducted showed that placing a stop below a minor low will work 65 percent of the time, so they are a good first choice. Price congestion regions, prior minor highs, trendlines, moving averages, round numbers, long price bars and so on, all have advocates saying price will find support there. Many make for good stop locations. Try some of them and see if your prior trades would have worked better. And what happens if there is no minor low or other congestion region to hide underneath? A Fibonacci-based stop may be the answer.

The Fibonacci Correction Stop
Figure 2 shows a stair step advance that our trader, Dave, was lucky to buy into near the bottom, at $7. Where should he place his initial stop? That’s a difficult question to answer because, on this chart, there is no congestion region nearby. Looking back on the monthly chart, the stock had not been this low since 1993 and perhaps even earlier (but my data stops there). A beta-adjusted stop would work well in this situation as would a volatility stop. Let’s ignore volatility stops for the moment and see what price does.


click image for larger view

In about a week, price shoots up, reaching a high at B of 10.61. There’s a small congestion zone at D (7.93) where he can place a stop, but that still means a large give back (25 percent) if price were to decline. Imagine that point A is a new, all-time high, so there is no nearby support region to hide underneath. What does Dave do for a stop then? A rising price trend often pauses before the uptrend resumes, as Figure 2 shows at C. These small downturns are called corrections or retraces of the uptrend. Some traders believe that price will stop on or near a Fibonacci correction of 31.8 percent, 50 percent or 68.2 percent of the prior up move.

To calculate the Fibonacci retrace, measure the rise from the swing low (price turning point) at A to the swing high at B. In this case, it’s 3.76. Then multiply the difference by your Fibonacci retrace of choice (like 68.2 percent) and subtract it from the swing high. Place a stop below the result. In this example, place the stop at 8.05. Since the low at C is 9.13, the 50 percent or 68.2 percent Fibonacci corrections would work well. But are they the best choice?

I decided to find out using 766 stocks from 1994 to 2006, and found 1,956 samples of price in an uptrend where it moves from a swing low to a swing high and then retraces. Few stocks covered the entire period, and in many I only looked at the July 2005 to August 2006 period. I found that the median retrace was 59 percent (meaning half retraced less than 59 percent of the prior up move and half corrected more). The most frequent correction was 61 percent followed by 56, 50 and 55 percent. Two thirds of the samples retraced less than 67 percent, so I consider that the best stop location. In other words, measure the move from the swing low to high, take 67 percent of it, and subtract the result from the swing high. The result is the stop price and it should protect you 67 percent of the time from being stopped out. Unfortunately, giving back such a large percentage of the prior swing can mean disaster to your wallet or pocketbook. Consider using a volatility stop instead.

The Volatility Stop
A volatility stop is similar to a beta adjusted stop only you measure the average high-low price variation to determine the stop location. To find price volatility, take the difference between the intraday high and low for each day of the prior month, average the readings and multiply by two. That gives you the average volatility. Subtract it from the current low to get a stop price. When price makes a new high, recalculate the volatility stop price. If you are lucky and price goes vertical like the move from C to F in Figure 2, you may need to adjust the stop daily.

For example, Figure 2 has a volatility reading of 76 cents at B. The low price at B is 9.93, so you would place a stop no closer than 9.17 to avoid being stopped out on normal price fluctuation.

In extensive testing comparing the high-low method of calculating volatility, average true range and standard deviation, I found that taking the difference between the high and low prices gave the best performance. I also found that a multiplier of two worked best with a 22 bar look back (that is, use one month’s worth of data).

Returning to Figure 2, if you used the volatility stop of 9.17, you would have been stopped out at C. That is unfortunate because price continued rising to F. However, the point of using stops in the first place is to protect profits and/or minimize losses.

For the start of 2007 why not reassess your “bad habits” and make adjustments. What new type of stop would you like to try? Know that every time price hits your stop, you are trading like the smart money. And the smart money wins.