1. Smart money always attempts to control risk before counting the profits. Traders with experience pay first attention to potential loss, secondly to potential gain. “Potential” is the operative word in this scheme, realizing there are no “gimmees” as we construct setups with Reward/Risk ratios and safety-net stop-losses. Let’s use a selection made on 2/28/03 (EYE) as a case study for this discussion. The closing price on the day of the pick was $8.25.

2. The first important question to answer is where to put the initial stop-loss. The stop level should be tough to reach with normal trading variation, so that when it is reached its clear we made a bad choice. Then, the next question… what is normal trading variation? In our parlance, price variation is the channel between price highs and price lows. That is a given, there must always be a difference. Our first measure of price variation then is simply the range, high price minus low price during a trading session. We can create an instant measure on the sort tabs with a personal criteria formula, (H – L), giving us that range for the most recent session. There is greater strength in averages so a five-day average range would look like this:
   ((H – L) + (H1 – L1) + (H2 – L2) + (H3 – L3) + (H4 – L4))/5

3. This could have been written for a 10-day average range, or for any number of days for that matter. This is a personal preference, but as the time window gets too long, we miss the character of the more recent price activity.

4. In 1978, J. Wells Wilder, Jr. in his classic “New Concepts in Technical Trading,” showed that the simple average range calculation misses changes in price from one day to the next, particularly with gaps, offering what he calls the true-range. True range for a days trading activity includes what happens the day before. True Range (TR) = The greater of (Hi - Low), (Hi - yesterday's close), (Low - yesterday's close), and Average True Range (ATR) = a moving average of TR. This is easy to setup on a spreadsheet. However, it is difficult with a PCF since we do not have a Maximum function. A proxy for Wilder’s ATR is the following PCF to find a 5-day ATR. This measures inherent noise in the price pattern (again, this can be modified for a 10-day ATR, etc.):
   (((((H4-L4)+(H3-L3)+(H2-L2)+(H1-L1)+(H-L))/5)+((ABS(C5-H4)+ABS(C4-H3)+ABS(C3-H2)+ABS(C2-H1)+ABS(C1-H))/5)+((ABS(C5-L4)+ABS(C4-L3)+ABS(C3-L2)+ABS(C2-L1)+ABS(C1-L))/5))/3.5]

5. On February 28th the 5-day ATR for EYE was $0.40 using this PCF. We’re saying we expect this much variation under normal conditions, without any change in character. So if we set our stop-loss at $8.25 - $0.40 ($7.85), chances are we would likely get stopped with normal variation. So, a better stop-loss is some multiple of the ATR, we like 2X, or $8.25 – 2X0.40 = $7.45. This is another choice, with multiples of 1.5X to 3.0X not uncommon. With our initial stop-loss at $7.45 we maintained the position to enjoy quite a run.

6. Exit strategies can make sell-decisions for us. One such exit strategy is to sell when a target gain is reached. Had we used a 10% sell-target on EYE, our exit would have occurred on 3/17 when price closed at $9.60. A lot of money has been made using a fixed % profit sell-strategy. However, a quick look at the more recent price pattern shows how painful it is to leave so much money on the table. Another approach suggests we beat out a lot of singles, yet strive for home-runs when possible. That is, work both ends of the trading equation. The few large gains will pay for all our small losses. To this end, a trailing-loss strategy is another option. This raises the stop-loss trigger as price increases, never reducing the stop-level with a decreasing price.

7. EYE is a good illustration for this principle, with the buy on 2/28/03 at $8.25. As each new closing price was reached, keeping a stop at twice the ATR, the price moved ahead handily. Critical point: Never reduce the stop from that derived at the highest price reached. In this long run-up, the highest price reached was $16.91 on 5/12/03. The ATR at that time was $0.57, giving us a stop at $15.77. Notice, the closing price Friday, 5/16/03, was $15.75. That kicked us out, with a price gain of 91% over 77 trading days. There are numerous other exit strategies which may see further gains, but for this technique, that was quite a run.

8. But a word on the ATR. A popular initial stop-loss is 7% or 8% below the buy-price. If we divide the ATR by the price, we get the percent which is helpful for comparison. On one end of the spectrum, a price pattern with a very low ATR would see 8% as too far away for practical purposes. Whereas, one with a high ATR may see 8% too close for normal variation. For example, look at the following for 2 and 3% ATR’s: GTRC, SHOO, SPN, CBI, HOTT, IPO. On the other hand, look at the following for ATR’s between 6 and 15%: REFR, SSYS, FDCC, GLDN, PILL, NICE.

9. Lastly, the PCF shown is not an exact replicate of Wilder’s ATR. Point: There is no “true” range, it being the construct of someone’s idea. But the principle is valid for trading techniques. Whether we use an Average Range, Wilder’s ATR, or the ATR PCF shown, the differences are in the 2% to 5% range which are meaningless in this game of guesses and guestimates. Concern over these differences is a classic case of swallowing the camel and gagging on a gnat. The principle is the camel, a great methodology, with the percent differences in detail being the gnat. Pick an option, then stick with it. Your trading success will benefit. We use the PCF since it is easy to execute, with more time available for heady decisions about which stocks to buy.

Main Picks:
ALSC,AUGT,CIG,EAS,IGLD,IMA,MDRX,STHLY,SWIR,TTMI
Breakouts:
ARTI,FARO,GSIC,IDBE,MRGE,PRCP,SIFY,SIGM,UAG,WCI
QuickPicks:
CNP,CTLM,GFI,LAB,MVL,RDWR,SCUR,SPN,TRZ,USG