Whenever you enter a trade, three factors must be crystal clear in your mind-where to get in, where to take profits, and where to bail out in case of an emergency. Daydreaming about profits will not make you rich. You must decide in advance where you'll take your winnings off the table or cut and run if the market turns against you.

Beginners keep looking for promising trades and assume that find ing them will make them money. They search for entries, while pro fessionals spend a lot of time and energy planning their exits. They always ask themselves where to take profits or cut losses. Survivors know the essential truth-you don't get paid for entering trades, you get paid for exiting them.

Why think about an exit before you enter a trade? Isn't it better to get in, monitor the trade, and exit in response to price action? There are two main arguments for deciding on an exit before you enter a trade.

First, knowing your targets and stops allows you to weigh rewards and risks. If you have a clear signal to buy, with a price target of $2 above and a stop $4 below, is that trade worth taking? Do you want to risk $4 to gain $2? Price targets and stops prompt you to focus only on trades whose potential rewards far outweigh the risks. The ability to walk away from a potential trade is as important as the ability to decline a drink when you want.

Second, setting profit targets and stops before entering trades helps sidestep the pernicious "ownership effect." We get attached to things we own and lose objectivity. That old ratty jacket hanging in your closet should have been given to the Salvation Army years ago. That trade you entered last week is starting to come apart at the seams. Why don't you chuck it? You own both the jacket and the trade; they feel comfortable, familiar. That's why you need to decide on exits before you enter, before a trade becomes yours.

A friend who is a top-notch technical analyst temporarily took a job as a broker after falling on hard times when his hedge fund closed. Naturally, I moved one of my accounts to him. Whenever I called to place an order, he would not let me hang up unless I gave him a stop. Sometimes I'd plead for a little more time and he'd take my order, on the condition that I call back within five minutes to give him a stop, or else he would call me. Six months later he found work as a technical analyst, and I have never seen another broker like him since. He really stressed the need to know your safety parameters before you took on market risk.

Before entering a trade, you should set two specific price levels, a target and a stop, one above and the other below current prices. This is all you need for a short-term trade, where you shoot at a clearly visible target. You may find a broker who accepts OCO (one cancels other) orders. Then, if your profit target is hit, the stop is automatically cancelled, and vice versa. If your broker does not accept OCO orders, give him your stop and keep an eye on your profit target.

What if you're planning a fairly long-term trade that may last for several days or even weeks? Profit targets shift with the passage of time, and protective stops need to be tightened as the trade moves in your favor. You must write down your rules for exiting trades and follow them from that piece of paper without arguing, haggling, or hoping for a better price. For example, you may decide to exit if the upper channel line gets hit or if the market makes lower lows for two days in a row. Whatever rules you use, write them down and execute immediately once the market hits a profit target or a stop-loss level in accordance with your rules.

Very experienced traders know how to recognize unusually forceful trends, during which they shift exit strategies, take partial profits, and carry the balance of their positions with a modified exit strategy for runaway trends. As you become more experienced, you can become a little more relaxed with your plans, but a beginning or an intermediate trader must be very strict. Entries are easy because any clown can buy a lottery ticket, but exits separate winners from losers.

Channel Targets

Traders all over the world squint at their charts, trying to recognize patterns, and let their imaginations run wild. Statistical studies, however, consistently confirm only one pattern-the tendency of prices to fluctuate above and below value. Markets may be chaotic most of the time, but their overbought and oversold conditions create islands of order that provide some of the best trading opportunities. Markets swing between elation and despair, and we can make money from those moods.

Channels are technical tools that help us take advantage of market swings. We draw them parallel to the moving average in the inter mediate timeframe, usually daily. A well-drawn channel contains ap proximately 95% of recent prices. Its upper line represents the manic and the lower line the depressive moods of the market.

If we buy value near a rising moving average, we can sell mania in the vicinity of the upper channel line. If we go short near a falling moving average, we can cover at the depressed level near the lower channel line. Channels provide attractive targets for profit taking.

People say that a neurotic is a person who builds castles in the clouds, a psychotic lives in them, and a psychiatrist is the fellow who collects the rent. Channels help us collect rent from what drives most investors crazy-the relentless swings of the markets. The idea is to buy normalcy and sell mania or go short normalcy and cover depression.

Straight channels or envelopes work better for profit taking than standard deviation channels, or Bollinger bands. Those bands grow wide when volatility rises and narrow when it declines. They help options traders, who depend heavily on volatility, but those of us who trade stocks or futures are better off with straight channels.

Channels are for traders, not investors. If you want to invest in a $10 stock and ride it to $50, channels are not for you. Exits from investments or very long-term trades are based on the fundamentals or such long-term technical signals as the reversals of a 26-week moving aver-age. Envelopes or channels work best when you trade relatively shortterm swings between undervalued and overvalued levels.

If you buy near a rising EMA, place a sell order where you expect the upper channel line to be tomorrow. If the upper channel line has been rising a half a point a day for the past few days and closed at 88 today, then you can place a sell order at 88.50 for tomorrow. Adjust this number every day as the channel moves higher or lower.

Whenever I teach a group how to use channels for profit taking, someone raises a hand and points to an area where prices overshot their channel. Taking profits at that channel line would have caused us to miss a large part of a rally. What can I say? This system is good but not perfect. No method, except for hindsight, nails down all tops and bottoms. Robert Prechter, who used to be a famous market analyst, put it well when he said, "Traders take a good system and destroy it by trying to make it into a perfect system."

If a trend is very strong, you may want to ride swings a little farther. Sell half of your position when prices hit the upper channel line, but use your judgment to dance out of the second half. You may monitor intraday prices and sell on the first day when they do not make a new high. Use your judgment and skills, but do yourself a favor-give up the idea of nailing tops. Greed is a very expensive emotion.

If a rally is weak, prices may begin to sink without reaching their upper channel line. There is no law that says the market must become manic before returning to value. Force Index can help measure the strength of a rally. When the two-day Force Index rises to a new high, it confirms the power of bulls and encourages you to hold until prices hit the upper channel line. If the two-day Force Index traces a bearish divergence, it shows that the rally is weak and you better grab profits fast.

An A trader is someone who takes 30% or more out of a channel. That is a little more than half the distance from the moving average to the channel line. Even if you buy slightly above the moving average and sell below its upper channel line, you can be an A trader and profit handsomely. Channels help catch normal tops and bottoms, and you can become very rich by steadily raking in normal profits. Channels help set up realistic profit targets.

Protective Stops

Amateurs swing between fantasy and reality, making most decisions in the realm of fantasy. They dream of profits and avoid unpleasant thoughts about possible losses. Since stops force us to focus on losses, most traders resist using them.

A friend told me she needed no stops because she was an investor. "At what price did you buy that stock?" I asked. She had gotten in at 80 and now it was at 85. "Would you still hold it if it fell back to 80?" She said she would. "What about 75?" She said she would probably buy more. "What about 70?" She winced. "What about 55? Would you still want to own it?" No, no, she vigorously shook her head. "Well, then you need a stop somewhere above 55!"

I recently had dinner with a lawyer who obtained inside information that a certain penny-stock company was about to announce a strategic partnership with a telecom giant. Questions of legality and morality aside, he put most of his money into that stock at an average price of

16.5 cents a share. Once the announcement came out, his stock ran up to $8, but by the time he told me his secret over a tray of sushi, it was down to $1.50. He had no stop. I asked him whether he would con-tinue to hold if his stock slid to 8 cents, half of what he paid. He was shocked and promised to put in a stop at $1. Did he do it? Probably not. It is easier to dream and close one's eyes to reality.

You must put in a stop immediately after entering a trade and start moving it in the direction of that trade as soon as it starts moving in your favor. Stops are a one-way street. When long, you may raise, but never lower them. When short, you may lower but never raise them. Only losers say, "I'll give this trade a little more room." You already gave it all the room it needed when you placed your stop! If a stock starts moving against you, leave your stop alone! You were more rational at the time you placed it than you are today, with prices hovering and threatening to hit it.

Investors must reevaluate their stops once every few weeks, but traders have a harder job. We must recalculate our stops every day and move them often.

A Deadly Delusion Many traders think they can stay out of trouble without using stops, thanks to their superior trading market analysis. Trading is a high-wire act. You may walk that wire a hundred times without a safety net, but the very first fall can cripple you. You cannot afford to take that chance. No amount of brains will help you if you abandon stops. Several years ago I got a call from a world-famous developer of trading software. He invited me on a camping trip, and mentioned in pass ing that he had developed a fantastic system for futures or stock trading or investing. It was based on computerized pattern recognition, backtested on 20 years of data with breathtaking results. He had no money to trade that system, having lost his capital on an earlier venture, but showed his discovery to a group of money managers. They were so impressed they began to set up a hedge fund for him, and in the meantime gave him what they called a small account-$100,000.

I flew across the country and spent the first evening admiring my friend's system. "Are you trading anything now?" The system had given him six signals: in soybeans, Swiss francs, pork bellies, and three other markets. He had entered trades in all six. "How much did you allocate to each?" He had divided his account into six parts, one for each market. No reserves. Fully margined. "Where did you put your stops?" He told me in so many words that real men didn't use stops.

He had mathematical proof that stops lowered profitability. Safety lay in trading unrelated markets. If one or two went against him, others would move in his favor. "What about a catastrophic event, with all mar kets moving against you?" He assured me that was impossible because he traded unrelated markets, and there was no correlation between Swiss francs and pork bellies. Furthermore, his system had not a single wipeout in 20 years of backtesting.

I suggested we forget about camping and stay closer to the screen, since his entire capital was at risk. My friend insisted he had total confidence in his system, and so we drove to the Sierra Nevada, some of the most breathtaking scenery in America. We had a grand time, and on the last day my son, who was about 8 at the time, hauled in a plastic pail full of gold nuggets. It was fool's gold, of course, but to this day I keep one of them on my desk as a paperweight, engraved "All that glitters is gold."

By the time we returned to civilization the impossible had happened. All six markets went against my friend, nearly wiping out his investors portfolio equity. The next morning we watched in horror as the markets opened one after another, continuing to go against him in huge gobs. I talked him into closing two out of six positions, but then it was time to drive to the airport.

A few days later I called to thank my friend for the trip. His account was wiped out, and he bitterly complained that his money men weren't gentlemen; they weren't returning his phone calls. I resisted the temptation to say that if he lost $100,000 for me, I would not return his phone calls either.

A trader who doesn't use stops will eventually take the mother of all losses. Spectacular disasters hit brilliant individuals who believe that their general sharpness and great systems override the need for stops. A careless trader can get away with no stops for a while, but if he trades long enough the market will kill him.

No amount of brains will save a trader who doesn't use stops. A Nobel Prize for research in the financial markets will not help him. Look at Long-Term Capital Management, a hedge fund owned and operated by a clutch of certified geniuses, including a former Salomon director, a former Fed governor, and two Nobel Prize winners. Those people were too bright to use stops. They came to the brink of failure in 1998 and didn't bust out only because the US Federal Reserve stepped in to arrange a bailout to avoid disrupting world markets.

No amount of intelligence, knowledge, or computer power will save you from a disaster if you trade without stops. Stops are essential for your survival and success.

What about mental stops-deciding on your exit point and then watching the market? If it violates that level, you exit your position, try ing to get the best price. This is a common strategy among pros who have a lot of experience and iron discipline. A beginner, on the other hand, watches the market the way a rabbit watches a snake-frozen in fear, unable to move. He must place actual stops.

Stops do not provide total protection because prices are not continuous and can gap across the stop level. You can buy a stock at 40 and place a protective stop at 37, but a bad earnings report or an announcement can force that stock to open at 34 tomorrow, filling your order at a much worse level than expected. This is not an argument against stops. An umbrella with holes is better than no umbrella at all. Also, money management rules provide an extra level of protection.

Stops in Two Dimensions Placing stops is one of the hardest challenges in options trading, more so than finding good trades. You want to place them close enough to protect your capital but far enough to avoid being stopped out by meaningless noise. It is a delicate balancing act.

Most trading books repeat the same advice: place a stop below the latest low when long or above the latest high when short. This method is so simple and common that tons of stops become bunched up at the same obvious levels. Professionals are not blind; they look at charts and know where those stops are. They gun for them, trying to trigger stops with false breakouts.

When a stock hangs just above support, the inflow of fresh buy orders dries up, and those with stops below support just suck air and wait. The pros sell short, giving that stock a bit of a push. It stumbles below support, setting off a flurry of sell stop orders. The pros who shorted at a higher level begin to cover, buying on the cheap from amateurs whose stops they hit. As soon as the decline slows down they redouble their buying, and go long. The market rallies, and the pros who bought below support now sell into the rally. Most breakouts from trading ranges are false breakouts-fishing expeditions by the pros gunning for stops at common levels. Once those stops are cleaned out, the market is ready to reverse. Most traders become so disgusted after getting hit by several false breakouts that they give up using stops. That's when a real reversal catches them. They end up losing money, with and without stops, and wash out of the markets.

Placing stops at obvious levels is not a good idea. You are better off placing them a bit closer to protect capital or a bit farther to reduce the risk of getting hit. Try not to do what everybody else does. Be sure to place your stop where you do not expect the market to go. If you expect prices to fall to a certain level, why place a stop there? You're better off closing your trade without waiting.

There are two inputs into placing stops: technical analysis and money management. You can combine them to find the right size for your trade as well as the right place for your stop. The first step is to decide how many dollars you should risk on the trade you're about to take. Later, in the section on money management, you'll learn to limit your risk on any trade to a tiny percentage of your account. If you're not fully confident, risk an even smaller percentage. Once you have the dollar figure for your maximum risk, turn to technical analysis to find where to place your stop. A stop based on technical analysis is almost always tighter, that is, closer to the market, than the money management stop. Your account is now starting to look like a submarine with a double hull-softer on the outside and harder on the inside.

Your money management stops belong in the market. They repre sent your maximum allowable risk level, which you may not violate under any circumstances. If your technical analysis stops are closer to the market, you may hold them in your mind as you monitor prices and are prepared to exit if those levels gets hit.

Here, I want to share with you two advanced methods for placing stops. Try to program them into your software and test them on your market data. Until now, I have never disclosed SafeZone to traders, except to small groups in Traders' Camps, where I like to share my latest research. It is my principle not to withhold information from my books. I write as I trade and maintain my edge not by secrecy but by developing new methods.

The SafeZone Stop

Once in a trade, where should you put your stop? This is one of the hardest questions in technical analysis. After answering it, you'll face an even harder one-when and where to move that stop with the passage of time. Put a stop too close and it'll get whacked by some meaningless intraday swing. Put it too far, and you'll have very skimpy protection.

The Parabolic System, described in Trading for a Living, tried to tackle this problem by moving stops closer to the market each day, accelerating whenever a stock or a commodity reached a new extreme. The trouble with Parabolic was that it kept moving even if the market stayed flat and often got hit by meaningless noise.

The concept of signal and noise states that the trend is the signal and the nontrending motion is the noise. A stock or a future may be in an uptrend or a downtrend, but the noise of its random chop can obscure its signal. Trading at the right edge is hard because the noise level is high. I developed SafeZone to trail prices with stops tight enough to protect capital but remote enough to keep clear of most random fluctuations.

Engineers design filters to suppress noise and allow the signal to come through. If the trend is the signal, then the countertrend motion is the noise. When the trend is up, we can define noise as that part of each day's range that protrudes below the previous day's low. When the trend is down, we can define noise as that part of each day's range that protrudes above the previous day's high. SafeZone measures market noise and places stops at a multiple of noise level away from the market.

We may use the slope of a 22-day EMA to define the trend. You need to choose the length of the lookback period for measuring noise level. It has to be long enough to track recent behavior but short enough to be relevant for current trading. A period of 10 to 20 days works well, or we can make our lookback period 100 days or so if we want to aver age long-term market behavior.

If the trend is up, mark all downside penetrations during the lookback period, add their depths, and divide the sum by the number of penetrations. This gives you the Average Downside Penetration for the selected lookback period. It reflects the average level of noise in the current uptrend. Placing your stop any closer would be self-defeating. We want to place our stops farther away from the market than the average level of noise. Multiply the Average Downside Penetration by a coefficient, starting with two, but experiment with higher numbers. Subtract the result from yesterday's low, and place your stop there. If today's low is lower than yesterday's, do not move your stop lower since we are only allowed to raise stops on long positions, not lower them.

Reverse these rules in downtrends. When a 22-day EMA identifies a downtrend, count all the upside penetrations during the lookback period and find the Average Upside Penetration. Multiply it by a coefficient, starting with two. When you go short, place a stop twice the Average Upside Penetration above the previous day's high. Lower your stop whenever the market makes a lower high, but never raise it.

I anticipate that SafeZone will be programmed into many software packages, allowing traders to control both the lookback period and the multiplication factor. Until then, you will have to do your own pro gramming or else track SafeZone manually (see Table 6.1). Be sure to calculate it separately for uptrends and downtrends.

Here are the rules for calculating SafeZone using an Excel spread sheet. Once you understand how it works, try to program SafeZone into your technical analysis software and superimpose its signals on the chart. Compare the numbers from the spreadsheet and the trading soft ware. They should be identical; otherwise, you have a programming error. Comparing results from two software packages helps overcome pesky programming problems.

Rules for Longs in Uptrends When the trend is up, we calculate SafeZone on the basis of the lows because their pattern determines stop placement.

1. Obtain at least a month of data for your stock or future in highlow-close format, as shown in Table 6.1 (lows are in column C with the first record in row 3).

2. Test whether today's low is lower than yesterday's. Go to cell E4 , enter the formula =IF(C3>C4,C3 - C4,0) and copy it down the length of that column. It measures the depth of the downside penetration below the previous day's range, and if there is none, it shows zero.

3. Choose the lookback period and summarize all downside penetrations during that time. Begin with 10 days and later experiment with other values. Go to cell F13 , enter the formula =SUM(E4:E13) , and copy it down the length of that column. It will summarize the extent of all downside penetrations for the past 10 days. 4. Mark each bar that penetrates below the previous bar. Go to cell G4 , enter the formula =IF(C4<C3,1,0) and copy it down the length of that column. It will mark each downside penetration with 1 and no penetration with 0.

5. Count the number of downside penetrations during the lookback period, in this case 10 days. Go to cell H13 , enter the formula = SUM(G4:G13) , and copy it down the length of that column. It will show how many times in the past 10 days the lows have been violated.

6. Find the Average Downside Penetration by dividing the sum of all downside penetrations during the lookback period by their number. Go to cell I13 , enter the formula =F13/H13 , and copy it down the length of that column. It will show the Average Downside Penetration for each day, that is, the normal level of downside noise in that market.

7. Place your stop for today at a multiple of yesterday's Average Downside Penetration below yesterday's low. Multiply yesterday's Average Downside Penetration by a selected coefficient, starting at 2 but testing as high as 3, and subtract the result from yesterday's low to obtain today's stop. Go to cell J14 , enter the formula =C13 - 2 * I13 , and copy it down the length of that column. It will place a stop two Average Downside Penetrations below the latest low. If today's low penetrates yesterday's low by twice the normal range of noise, we bail out.

8. Refine the formula to prevent it from lowering stops in uptrends. If the above formula tells us to lower our stop, we simply leave it at the previous day's level. Go to cell K16 , enter the formula

= MAX(J14:J16) , and copy it down the length of that column. It will prevent the stop from declining for three days, by which time either the uptrend resumes or the stop is hit.

Rules for Shorts in Downtrends When the trend is down, we calcu late SafeZone on the basis of the highs because their pattern deter mines stop placement.

1. Obtain at least a month of data for your stock or future in highlow-close format, as shown in Table 6.1 (highs are in column B with the first record in row 3).

2. Test whether today's high is higher than yesterday's. Go to cell L4 , enter the formula = IF(B4 > B3,B4 - B3,0) , and copy it down the length of that column. It measures the height of the upside penetration above the previous day's range, and if there is none, it shows zero.

3. Choose the lookback period for summarizing upside penetrations. Begin with 10 days and experiment with higher values. Go to cell M13 , enter the formula = SUM(L4:L13) , and copy it down the length of that column. It will summarize the extent of all upside penetrations for the past 10 days.

4. Mark each bar that penetrates above the previous bar. Go to cell N4 , enter the formula = IF(B4 > B3,1,0) , and copy it down the length of that column. It will mark each upside penetration with 1 and no penetration with 0.

5. Count the number of upside penetrations during the lookback period, in this case 10 days. Go to cell O13 , enter the formula = SUM(N4:N13) , and copy it down the length of that column. It will show how many times in the past 10 days the highs have been violated.

6. Find the Average Upside Penetration by dividing the sum of all upside penetrations during the lookback period by their number. Go to cell P13 , enter the formula = M13/O13 , and copy it down the length of that column. It shows the Average Upside Penetration, the normal level of upside noise in that market.

7. Place the stop for your short position today at a multiple of yesterday's Average Upside Penetration above yesterday's high. Multiply yesterday's Average Upside Penetration by a selected coefficient, starting at 2 but testing as high as 3, and add the result to yesterday's high to obtain today's stop. Go to cell Q14 , enter

the formula = B13 + 2 * P13 , and copy it down the length of that col-umn. It will place a stop two Average Upside Penetrations above yesterday's high. If today's high shoots above yesterday's high by twice the normal amount, it hits our stop and we bail out.

8. Refine the formula to prevent it from raising the stop during a downtrend. If the above formula tells us to raise our stop, we simply leave it at the previous day's level. Go to cell R16 , enter the formula = MIN(Q14:Q16) , and copy it down the length of that column. It will prevent the stop from rising for three days, by which time either the downtrend resumes or the stop is hit.

SafeZone offers an original approach to placing stops. It monitors changes in prices and adapts stops to the current levels of activity. It places stops at individually tailored distances rather than at obvious support and resistance levels.

SafeZone works on the way down just as well as on the way up. Here we count each upside penetration of the previous day's range during a selected time window and average that data to find the Average Upside Penetration. We multiply it by a coefficient, starting with 3, and add that to the high of each bar.

Like all systems and indicators in this book, SafeZone is not a mechanical gadget to replace independent thought. You have to establish the lookback period, the window of time during which SafeZone is calculated. Do not go back beyond the last important turning point. If the market has reversed from down to up two weeks ago, then SafeZone for the current long trades should not look back more than 10 trading days.

Another important decision is choosing the coefficient for the SafeZone stop. Usually, a coefficient between two and three provides a margin of safety, but you must research it on your own market data. Once you have done your homework and tweaked this indicator, it will become your own private tool in the battle for survival and success in the markets. You can add it to almost any trading system, including Triple Screen.

The Chandelier Exit

When a trend speeds up, you may want to shift gears and ride it instead of trading swings. Catching swings calls for tight stops, but a longer-term position demands more breathing room. The Chandelier Exit is designed to protect such positions.

When buyers place stops, they usually count back from the lows and put stops below recent important bottoms. When traders go short, they usually count back from the highs and place stops above the recent peaks. The Chandelier Exit takes a different approach. When traders are long, it hangs their stops from the highest peak reached by that trend, like a chandelier hangs from the tallest point in the room. As prices move up, the Chandelier Exit, suspended from the highest point of that trend, also rises. It tracks volatility as well as prices, as its distance from the peak grows with the rise in volatility. We will review Chandelier Exits for up-trends, but you may reverse these rules and apply them to downtrends.

There is no telling how high a trend may go, and a Chandelier will rise until prices peel off from the ceiling and hit that stop. This method, along with several others, was presented by Chuck LeBeau in our Traders' Camp in January 2000 in the Caribbean and again in March 2001 in the Pacific.

The Chandelier Exit draws on the concept of Average True Range, described by Welles Wilder in 1966. The True Range is the greatest of the three figures-the distance between today's high and low, or today's high and yesterday's close, or today's low and yesterday's close. The True Range reflects overnight volatility by comparing today's and yesterday's prices.

The True Range is included in many software packages. The Average True Range (ATR) is obtained by averaging True Range over a period of time. How long a period? You can begin by looking back a month, but a modern computerized trader can easily test different lookback periods for Average True Range.

The Chandelier Exit subtracts the Average True Range, multiplied by a coefficient, from the highest point reached by the trend. Its formula is:

Chandelier = HP - coef * ATR

where

Chandelier = the Chandelier Exit

HP = the highest point for a selected number of days

coef = the coefficient, selected by the trader

ATR = Average True Range for a selected number of days

If an uptrend is boiling and daily ranges are wide, then stops fall a little farther away. If the uptrend is quiet and calm with more narrow ranges, the stops inch a little closer.

In Windows on Wallstreet software, the formula for the Chandelier Exit is:

Hhv(hi, 22) - 3 * ATR(22)

Hhv(hi, 22) is the highest high reached in the past 22 days and ATR(22) is the Average True Range for the past 22 days. Try to test other parameters in the markets you like to trade.

This formula multiplies the Average True Range by three before subtracting it from the high of the past 22 trading days. A serious trader is passionate about his research. Such a person will immediately notice that this formula contains three variables-the length of the lookback period for the highest high, the length of the lookback period for the True Range, and the coefficient for multiplying Average True Range. It probably doesn't pay to fiddle with the first variable, as the highest high in a runaway uptrend is likely to be close to the right edge and most lookback periods catch it. The Average True Range is only slightly more sensitive to the length of its lookback period. A more fertile ground for experimenting is the ATR coefficient. If everyone has their stops at 3 ATRs below the peak, wouldn't you like to see what happens if you set your stop at 3.5 or 2.5?

The Chandelier Exit can be used to trail profits on short positions in runaway downtrends. There the formula becomes:

Llv(lo, 22) + 3 * ATR(22)

Llv(lo, 22) is the lowest low reached in the past 22 days and ATR(22) is the True Range for the past 22 days.

Experienced traders sometimes joke that bear markets have no support and bull markets no resistance. A runaway trend can overshoot all rational expectations. The Chandelier deals with that by tying its stops to the price extreme as well as volatility.

The negative side of Chandelier Exits is that they give up a great deal of profit. Three Average True Ranges can amount to a lot of money in a volatile market. A beginning trader is better off taking his profits at the channel wall. A more advanced trader whose market is hitting its channel wall while showing great strength is more likely to switch to a Chandelier Exit. If he holds a large position, he may take partial profits at the channel wall and carry the rest using the Chandelier. The Chandelier Exit may act a backup exit strategy for experienced traders.

Gamblers lose fortunes trying to nail tops and bottoms. A good trader is a realist who wants to grab a chunk from the body of a trend, leaving topand bottom-fishing to people on an ego trip. The Chandelier Exit helps take care of that chunk.