Measuring Momentum
One of the harder jobs in technical analysis is figuring out whether a trend is genuinely running out of steam or whether the market is merely taking a short rest before continuing. Momentum is one of the most useful tools for making that distinction. In its simplest form, momentum measures the speed at which price is changing. The calculation is remarkably uncomplicated: compare today's price with the price a specified number of periods ago. If prices are advancing rapidly, the momentum reading rises and the indicator line slopes upward. Prices can continue moving higher while momentum is falling, however, because momentum is concerned with the rate of change rather than the direction of price itself. When the rate of advance begins slowing, the momentum line flattens. Once prices begin declining, the reading falls and the line turns downward.
That is the elementary version, but there is considerably more to momentum than dividing one price by another. You want to know whether the rate of change is accelerating or slowing, rather than relying entirely on your eyes to judge the slope of a line. That brings us to rate-of-change calculations. You also want to reduce some of the noise that can obscure the underlying movement and, in many cases, judge the momentum relative to the security's normal price range. Once you start asking those questions, more arithmetic becomes necessary.
Momentum also appears in indicators that do not look like traditional momentum indicators. For instance, consider MACD. Its ingredients are moving averages, yet what the finished indicator depicts is the changing relationship between prices and their recent history. In that sense, MACD is fundamentally a momentum tool. Many momentum indicators are used to confirm a trend that has already been identified by something else, such as a moving-average crossover, but there are also systems in which momentum itself becomes the primary decision-making mechanism.
Is Momentum Really a Leading Indicator?
There is a legitimate argument over whether momentum should be classified as a leading indicator. After all, how can an indicator calculated entirely from historical prices actually lead the market? It cannot know the future. Yet momentum indicators sometimes appear to turn before other indicators because traders respond to certain conditions in remarkably repetitive ways.
When a large number of traders regard a security as overbought, they tend to respond according to the rule associated with that condition and begin taking profits. When the same security reaches an oversold condition, buyers become interested. Momentum therefore appears to anticipate a reversal, but there is no mystical force involved. Momentum is measuring changes in the behavior of the crowd. Traders become increasingly enthusiastic, newcomers join them, the rate of advance increases, and eventually that enthusiasm begins to fade. Once the crowd starts protecting its profits or abandoning positions, the rate of price change weakens. Momentum is observing crowd behavior through the arithmetic of price.
Doing the Math: Calculating Momentum
The word momentum carries a very precise meaning in physics, and that precision can become misleading when the concept is transferred to financial markets. In physics, momentum can be measured as mass multiplied by velocity. A heavy truck moving rapidly down a hill provides an intuitive example. A substantial counterforce is required to stop it.
Financial prices are not trucks, billiard balls, or particles moving through a physical system. They are the result of human decisions. You can use trading volume as a metaphor for mass and price change as a metaphor for velocity, but you should not confuse the analogy with a scientific measurement. The forces behind a price movement are people, institutions, news, expectations, fear, greed, and thousands of other influences that cannot be reduced to a physical equation.
This distinction matters because many influential technical analysts and system designers have engineering backgrounds. It is tempting to borrow the terminology and mathematics of engineering and assume that doing so gives market momentum the same scientific certainty that physical momentum possesses. It does not. A runaway truck can be stopped by applying a measurable mechanical force. There is no comparable mechanical force that can be applied to a runaway stock price. Short of closing the exchange or otherwise preventing the security from trading, the market continues to operate according to the decisions of its participants.
Markets can also be hit by sudden external shocks that completely alter the previous momentum. A flash crash, a war, or some other unexpected event can change the direction and speed of prices in a matter of minutes. The defining feature of a shock is that you do not know precisely when it will arrive or what its ultimate effect will be. The same event can have dramatically different consequences in different markets. A major shock affecting the dollar and Swiss franc need not produce the same price response as an identical shock affecting the Shanghai stock market. You can measure the resulting movement afterward. You cannot know its exact influence beforehand.
Simple Momentum
The basic momentum calculation is still a useful indicator, although it has become less fashionable as traders have developed more elaborate versions. The essential idea is all you really need to remember: compare today's price with the price a selected number of periods ago. You are not trying to predict the price directly. You are measuring how much it has changed over that interval and, more importantly, whether that change is accelerating or weakening.
You can choose almost any look-back period. Five days is useful for illustration, while 12 and 14 days are common settings in charting programs. Those parameters became popular because the designers of the indicators found them useful, not because the market contains some immutable law requiring every trader to use exactly 12 or 14 days. A shorter period will respond more quickly to recent price changes but will also produce more noise. A longer period will smooth some of that noise while making the indicator slower to react. The choice should therefore be related to the time frame and behavior of the security rather than treated as a sacred number.
The older method subtracted the earlier price from today's price. Modern momentum calculations more commonly divide today's price by the price from the selected number of periods ago. The result is a ratio that tells you how today's price compares with its earlier level. The important point is the relationship, not the particular arithmetic. When today's price is lower than the reference price, momentum moves into negative territory. When today's price is higher, momentum moves above its reference level. The indicator can therefore begin changing character before the price itself produces an obvious reversal.
Consider a situation in which a security has been declining for several days. Momentum drops into negative territory because today's price has fallen below the price five days earlier. The negative reading can appear before the actual price decline becomes dramatic. In one such sequence, momentum turns negative a day before the price opens with a gap lower, providing an early indication that the rate of change has weakened. The indicator later recovers toward its center line, but fails to remain on the other side for very long. The price is still declining, so nothing about the chart necessarily looks bullish yet.
Eventually momentum crosses upward through its center line even though the price itself is still moving lower. This is the more interesting development. The price has not yet announced that the decline is over, but the rate at which price is changing has begun to improve. Shortly afterward, the security turns higher, reaches a previous high, and eventually breaks above it. The momentum indicator did not magically predict the reversal. It simply responded to an improvement in the price relationship before that improvement became obvious in the price trend itself.
This is one of the reasons momentum can be useful as a secondary indicator. Price tells you where the market is. Momentum tells you something about how quickly the market is getting there compared with where it was previously. A declining price accompanied by improving momentum can therefore deserve attention, particularly when the price is approaching an important support level or when another indicator is beginning to confirm the change. It is not a guarantee of a reversal, but it can provide an early warning that the current movement is losing some of its previous force.
The same principle works in the opposite direction. A security can continue rising while momentum begins to weaken. If price is still making higher highs but momentum is no longer making corresponding highs, the advance may be losing acceleration. That does not mean you should automatically sell. Strong trends can continue after momentum has weakened. The value lies in recognizing that the character of the move may be changing and then looking for additional evidence before making a decision.
Momentum is therefore best understood as a measure of change rather than a crystal ball. It can sometimes turn before price, but an early turn is useful only if the market eventually confirms it. When momentum and price begin telling different stories, pay attention. The disagreement may be nothing more than temporary noise, or it may be the first indication that the current wave is approaching its end.
Momentum Versus Momentum Investing
Do not confuse the technical indicator called momentum with momentum investing. Momentum as an indicator is an arithmetic calculation. Momentum investing is the practice of buying a security primarily because its price is already rising.
We have had plenty of examples of momentum stocks and commodities over the years. During the late 1990s, Internet, telecommunications, and high-tech shares rose to extraordinary valuations as investors continued piling into the same group of securities. In the mid-2020s, artificial-intelligence stocks such as Nvidia and commodities including gold, silver, platinum, and copper provided examples of assets attracting substantial momentum-driven demand. The basic momentum-investing idea can be reduced to a very simple rule: if the price is going up, buy it.
That is not the same thing as momentum-based technical analysis. A momentum trader using technical analysis still needs defined entry and exit conditions. Buying something merely because it is rising is not a complete trading system. Comparative relative strength, which compares the momentum of different securities, is another legitimate use of momentum. It can be used to identify which security in a group has displayed the strongest relative performance. William O'Neil incorporated this type of comparison into his CANSLIM stock-selection methodology.
Using the Rate-of-Change Method
The momentum calculation used by many traders today takes the basic idea and expresses it as a percentage rate of change. One common version works as follows:
Subtract the price from the selected number of periods ago from today's closing price.
Divide that difference by the earlier closing price.
Multiply the result by 100.
The advantage is that you now have the change expressed in a percentage rather than merely as a raw price difference. If today's price is identical to the price five days ago, the rate-of-change reading is zero because there has been no percentage change. A positive number tells you the current price is higher than it was at the beginning of the measurement period, while a negative number tells you that it is lower.
Consider a simple example. Suppose a security was $10 five days ago and is now $15. Simple momentum gives you 15 divided by 10, or 1.5. If the price rises another dollar the next day, the calculation becomes 16 divided by 11, or about 1.45. That is an odd result because the stock is still climbing at exactly the same dollar-per-day pace, yet the momentum number has declined.
The percentage calculation makes the situation easier to interpret. The first reading is a 50 percent increase from $10 to $15. The next reading is approximately 45.45 percent from $11 to $16. The absolute price is still increasing at the same rate, but the percentage gain relative to the earlier price is declining. The move is therefore decelerating. That is information you can compare from one period to another in a meaningful way.
A momentum line can remain flat while prices continue moving. This is one of the most important points to understand. A flat momentum reading does not mean that price has stopped moving. It means the relative rate of change has stopped accelerating or decelerating. Prices can continue climbing while momentum remains horizontal if they continue changing at the same relative pace.
Rate of change also provides context. A 30 percent advance over ten days means something different from a 3 percent advance over ten days, and the percentage figure lets you make that distinction immediately. If you discover that a particular security rarely moves 30 percent in ten days, then a 30 percent reading becomes noteworthy. Traders who have seen that condition before may begin taking profits, producing the pullback that the momentum indicator is warning about. Depending on your trading plan, you might exit before the reversal, participate in the reversal, or simply recognize the condition and wait for additional evidence.
By studying momentum over long periods, you can establish the range of movement that a particular security has historically produced over a selected period. You might examine three-day, ten-day, 30-day, or other look-back periods. Most charting programs allow you to test different parameters against historical data, allowing you to determine which period would have produced the strongest results under a specified trading rule. Remember, however, that a parameter that performed well in the past is not guaranteed to remain optimal.
Pondering the Trickier Aspects of Momentum
When you put a momentum indicator underneath a price chart, your eye naturally tries to match the two. You see a price line rising and a momentum line rising and conclude that the two must be saying essentially the same thing. Often they are related, but they are not measuring the same thing. Momentum measures the speed of the price change, and that distinction produces several traps.
The easiest trap is a horizontal momentum line. You might look at it and conclude that momentum has stopped. What has actually stopped is acceleration or deceleration. The price itself may still be moving.
Smoothing Price Changes
If you compare a momentum line with the price series, you will notice that the two often have a family resemblance. Momentum frequently produces highs and lows around the same important areas where price produces its own turning points, although the momentum movement may occur somewhat earlier or later. This makes sense because the indicator is derived directly from price. It is not operating independently of the market. It is measuring the rate of change in the market and therefore tends to reflect many of the same advances, declines, peaks, and troughs that appear in the price itself.
Momentum compares the current close with a close from a fixed number of periods ago. The length of that look-back period has a major effect on the appearance and behavior of the indicator. A shorter period reacts quickly and can jump sharply when price changes suddenly, while a longer period generally produces a smoother series because each new observation is being compared with a price farther back in time. Momentum is also constructed differently from a moving average. A moving average incorporates the closing prices throughout its calculation period, whereas momentum is primarily concerned with the relationship between today's price and the price at a specific point in the past. That difference can cause the two indicators to react quite differently even when they are being applied to the same security.
A one-day price spike demonstrates one of the weaknesses you must understand. Suppose a stock suddenly jumps several dollars higher. Momentum responds because today's price is now dramatically different from the price used as its reference. But several days later, when that old price spike moves outside the look-back period, the momentum reading can change sharply again. Nothing particularly important may have happened to the stock on that second day. The apparent change in momentum may simply be the mechanical result of an old price entering or leaving the calculation. This is one reason you should continue looking at the actual price bars instead of allowing an indicator to become your entire view of the market. An indicator can tell you something useful about price, but it can also create movements of its own that have more to do with the mathematics of the calculation than with a new development in the security.
Momentum is not a trend indicator in quite the same sense as a moving average, yet it often appears to follow the trend. The reason is largely behavioral. When prices begin rising, traders notice the movement and some begin joining it. Their buying creates additional demand, which can cause the rate of advance to increase. Price rises more rapidly, and momentum responds to that increasing rate of change. Later, the behavior can change. Traders who already own the security may become less willing to chase it, while new buyers may become increasingly selective about paying higher prices. The price can continue producing higher closes, but each new gain may be smaller than the previous one. The trend remains upward, but the rate at which it is progressing is slowing.
This distinction between direction and speed is the real value of momentum. Price can still be moving in the same direction while momentum is telling you that the movement is losing force. Imagine a stock that advances $2, then $1.50, then $1, and then only 50 cents over successive periods. The stock is still going higher, but the character of the advance has changed. Momentum can begin reflecting that deterioration before the price actually turns down. That does not mean a reversal is guaranteed. Strong trends can slow down and then accelerate again. But a loss of momentum tells you that the trend is no longer behaving with the same urgency, and that is information worth having.
The opposite can happen during a decline. A stock may continue making lower closes while the size of those declines begins to shrink. The price has not yet turned higher, but the selling pressure may be losing some of its force. Momentum can recognize this change because it is measuring the rate of price movement rather than merely asking whether the latest close is higher or lower than the previous one. In this way, momentum can sometimes provide an early warning that the character of a move is changing even while the trend itself remains intact.
The important word is sometimes. Momentum does not know why price is moving, and it does not know what will happen next. A sharp change in the indicator can be caused by genuine acceleration in the market, but it can also be caused by the mechanics of the look-back period. You therefore want to understand both the indicator and the price that produced it. The more familiar you become with that relationship, the easier it becomes to distinguish a meaningful change in momentum from a temporary mathematical disturbance.
Filtering Momentum
There is always a trade-off between responsiveness and stability. A longer momentum period produces a smoother line, but a shorter period reacts more quickly to changes in price. Unfortunately, greater sensitivity also means greater vulnerability to small fluctuations and whipsaws.
One way to deal with this is to introduce a filter. Instead of treating every move through the zero line as an immediate signal, you might require momentum to move a specified amount beyond the threshold. For example, a buy might require a reading 2 percent above the center line, while a sell might require a reading 2 percent below it. Another alternative is to wait one or more days before accepting the signal. Historical testing can tell you whether either approach improves the results for the security you are trading.
Do not assume that upward and downward movements behave symmetrically. Securities have personalities. One may produce rapid, explosive advances followed by long, sluggish declines. Another may fall violently but recover slowly. If a security typically moves between momentum readings of plus 130 and minus 130, that tells you something about its usual rate of change over the selected period. Another security may lumber between plus 120 and minus 120 over many months, while a much more volatile issue may swing between plus 150 and minus 150 in only a few weeks. One behaves like an old pickup truck. The other behaves more like a sports car.
You will also hear commentators describe momentum cycles as though they were permanent laws. Treat that language cautiously. You have to decide for yourself whether cycles are sufficiently regular to be useful in your trading. On longer time frames, particularly weekly and monthly charts, you may occasionally see surprisingly regular oscillations. But a pattern appearing regularly in the past does not guarantee that the market will continue producing it.
Applying Momentum
The simplest momentum trading rule is straightforward: buy when the indicator crosses above its center line and sell when it crosses below. The zero line represents the point at which today's price is equal to the price from the selected number of periods ago. Above zero means today's price is higher than that earlier price. Below zero means it is lower.
This rule is easy to understand, but momentum has several unusual characteristics because it measures the rate of change rather than the dollar price itself. Those characteristics become especially important when you begin looking at divergence.
Discovering Divergence
Momentum can be confusing because traders are accustomed to interpreting an upward-sloping line as an increase in dollar value. With momentum, an upward slope means that the rate of price change is increasing. The price can therefore rise while the momentum indicator falls. There is no contradiction.
Divergence describes a situation in which momentum fails to confirm what price is doing. The clearest example occurs when prices establish a new high while momentum establishes a lower high. Technically, both may still be moving upward, so calling this a divergence is not mathematically perfect. The more useful interpretation is that momentum is failing to confirm the strength of the new price extreme.
A useful way to recognize weakening momentum is to watch what happens when price continues making new highs but the momentum indicator does not. Imagine a security advancing steadily and printing one higher high after another. At first, momentum rises along with it and establishes its own new highs. Eventually, however, the price reaches another high while momentum fails to do the same. The momentum line turns downward, may flatten for a period, and eventually crosses back through its center line. The price can still be advancing during much of this process. What has changed is the character of the advance. The gains are becoming smaller and less forceful, even though the trend has not yet visibly reversed.
This is an important distinction. A trader waiting for price itself to break down may have to give back a considerable portion of the move before receiving confirmation from a slower trend indicator. Momentum can sometimes provide the warning earlier because it is measuring the speed of the price movement rather than simply its direction. When momentum rolls over after failing to confirm a new price high, the trader has information that the advance is losing strength. In the right circumstances, that signal can be used to exit a position rather than waiting for another indicator to catch up. It is not a guarantee that the top is in, but it can provide a practical early warning.
Volume can make that warning more meaningful. Mature advances frequently attract a burst of activity as traders and investors rush to participate, while existing holders may use the increased demand to distribute shares. If volume suddenly expands at the same time that momentum is unable to confirm another price high, the two indicators are telling you something similar from different directions. Price is still rising, but the rate of advance is deteriorating and participation is becoming unusually heavy. That combination deserves attention.
Do not interpret it as an automatic reversal signal. A strong security can continue higher even after momentum has weakened, particularly when fresh buying appears. The point is that you now have evidence that the character of the move has changed. Momentum is no longer confirming the same strength it showed earlier, and abnormal volume provides another reason to question whether the advance can continue at its previous pace. When several independent pieces of evidence begin pointing in the same direction, you should pay attention before the price chart finally makes the change obvious.
Confirming Trend Indicators
Momentum is particularly useful as a confirming indicator. The basic confirmation principle is simple: require two or more independent pieces of evidence to agree before committing to a trade. A moving-average crossover combined with a momentum indicator provides one straightforward example.
Adding momentum confirmation can reduce the number of trades, eliminate some whipsaws, and improve the proportion of trades that meet your profitability criteria. You are giving up some early entries in exchange for additional evidence. That is a recurring trade-off throughout technical analysis.
Determining the Relative Strength Index
You may reasonably ask why you should wait for momentum to cross the zero line. Why not act as soon as the momentum indicator turns around? If momentum is already falling from a peak, shouldn't that be enough to tell you that the trend is weakening?
J. Welles Wilder Jr., one of the important early developers of modern technical analysis, approached the problem differently. Rather than waiting for momentum to cross the center line, he compared the average size of recent advances with the average size of recent declines. If average upward movement was becoming greater than average downward movement, the balance favored the buyers. If the relationship shifted the other way, the sellers were gaining control. This relative relationship gave the indicator its name: the Relative Strength Index, or RSI.
RSI generally reacts more quickly than basic momentum when a price move begins changing character. That makes it particularly useful for timing profit-taking. The drawback is that faster signals are not automatically more reliable signals. RSI can produce premature buy and sell indications, especially when a strong trend remains intact. For that reason, many traders use RSI as confirmation while relying on another indicator or price condition to establish the actual trade.
The mathematics can look intimidating when written out, but the important thing is understanding what the calculation is doing. Your software can handle the arithmetic.
Calculating the RSI
RSI measures the relative speed of price changes by comparing average gains with average losses over a selected number of periods. Unlike simple momentum, which can compare two individual prices, RSI works with averages of the intervening movements. It still uses a ratio, but the ratio is designed to produce an oscillator confined to a range from 0 to 100.
The first calculation is relative strength, or RS, which compares the average upward movement with the average downward movement. The RSI is then calculated as:
RS = Average Gain ÷ Average Loss
RSI = 100 − [100 ÷ (1 + RS)]
The exact calculation used by a particular implementation can involve additional smoothing conventions, but the purpose is the same. The resulting oscillator moves between 0 and 100. Readings near the lower end indicate that downward movement has dominated the recent period, while readings near the upper end indicate that upward movement has dominated.
An oscillator takes an indicator's movement and puts it onto a bounded scale so that you can identify unusually strong or weak conditions. RSI rarely spends much time at the absolute extremes. Traders commonly use 30 and 70 as the oversold and overbought reference points, although 20 and 80 or even 10 and 90 may be more appropriate for particular securities.
RSI also inherits a limitation from the period over which it is calculated. Imagine a 75-day uptrend interrupted by several RSI sell signals because the RSI is using a 14-day window. A long-term trend follower might regard those signals as false alarms. A short-term trader, however, could use those same temporary overbought conditions to take profits and then reenter. The usefulness of the signal depends partly on the time frame of the trader.
Picturing RSI
RSI becomes particularly useful when you compare its movement with both price and another momentum measure. Imagine the price series in the main window, with RSI and momentum displayed underneath it. At the beginning of the advance, all three are moving in the same general direction. Prices are climbing, momentum is strengthening, and RSI is rising along with them. Eventually, however, RSI reaches the 70 area and begins turning downward shortly after price establishes its highest close. Momentum also begins to roll over, but it takes roughly another two weeks before it crosses its center sell line. RSI has therefore provided the earlier warning that the character of the advance has changed.
The same principle applies at the other end of the oscillator's range. A reading near 30 is traditionally considered oversold, but that word can be misleading if you treat it as a command to buy. Oversold means that selling pressure has become unusually strong relative to recent gains. It does not mean that the security has finished declining. A powerful downtrend can drive RSI into the oversold area and keep it there while price continues to make lower lows.
Consider what happens when RSI reaches the oversold zone well before the eventual price bottom. The security continues declining for roughly another two and a half weeks, establishing new lows even though the average size of those declines is becoming smaller. The market is still going down, but the rate of decline is losing some of its force. RSI recognizes this change because it is comparing the relationship between recent gains and losses rather than simply asking whether the latest price is lower. In this situation, the initial oversold reading would have been a poor reason to buy immediately. It was better understood as a warning that the character of the decline might be changing.
The traditional overbought condition occurs when RSI reaches 70 or higher. The reasoning is straightforward. The security has advanced far enough, and often quickly enough, that traders may become increasingly willing to take profits. You can turn the 70 crossing into a direct trading rule, or you can use it as confirmation for a signal coming from another indicator. When RSI crosses into the overbought region near the end of an advance, it can provide a useful indication that the move has reached an area where caution is appropriate.
The oversold condition around 30 works in much the same way, but in reverse. The traditional interpretation is that selling has become sufficiently intense that many of the willing sellers have already acted, leaving the security relatively depressed and potentially more attractive to buyers. That interpretation has value, but it becomes dangerous when you turn it into a mechanical rule. Strong downtrends can remain oversold for surprisingly long periods. A trader who buys every time RSI reaches 30 can therefore find himself buying repeatedly while the security continues to fall.
This is why the behavior following an oversold reading matters as much as the reading itself. RSI can spend considerable time near the lower boundary while price continues making new lows. The indicator is already detecting an improvement in the relationship between average gains and average losses, even though the price has not yet turned upward. For a trader who is already holding a long position, that improvement can provide a reason for patience rather than an immediate reason to abandon the trade. Eventually, the price does turn higher and RSI works its way back toward the overbought region. In this particular sequence, RSI and the eventual price high occur at roughly the same time, while momentum reaches its peak earlier but has not yet crossed its center line by the end of the available price action.
The larger lesson is more important than either the 70 or 30 level. An oscillator can sometimes function as a direct trading signal, while in another situation the same reading is merely an early warning. The difference is context. You must consider the direction and strength of the underlying trend, what price is actually doing, and whether other indicators are confirming the message. An extreme oscillator reading tells you that something unusual is happening in the relationship between recent gains and losses. It does not, by itself, tell you that the market must immediately reverse.
Filtering RSI
Tushar Chande has spent considerable time developing and refining momentum indicators. His Chande Momentum Oscillator calculates the difference between the total of recent gains and the total of recent losses and then relates that difference to the total amount of price movement over the selected period. A common default is 20 periods. The resulting oscillator ranges from plus 100 to minus 100, with readings above plus 50 considered overbought and readings below minus 50 considered oversold. Charting programs commonly place a nine-day moving average over the oscillator to provide a signal line.
The Chande Momentum Oscillator differs from RSI in the way gains and losses enter the calculation. Both upward and downward movements contribute to the numerator and denominator, which permits the indicator to occupy a negative range. Because the strength of the move is measured against all of the price movement in the period, strong advances and declines can appear somewhat sooner than they do with RSI. Sometimes the difference is only a day or two, but in a fast market a day or two is not trivial.
Another Chande indicator is Aroon, a name derived from the Sanskrit word for dawn. Aroon uses two lines. One measures how many periods have passed since the most recent high, while the other measures how many periods have passed since the most recent low. The standard calculation commonly uses 20 periods and converts the results to an index.
When Aroon Up remains above 70, the upward trend is considered to have strength. When Aroon Down falls below 30, the downward trend is dominant. Aroon was designed primarily as a confirmation tool, but the crossing of its two lines can also be used as a buy or sell signal.
Another Way to Filter Momentum
Wilder also developed another method for separating meaningful trends from minor price movements. Larry Connors and Linda Raschke discuss the approach in Street Smarts: High Probability Trading Strategies for the Futures and Equity Markets. This technique involves more arithmetic than the simple momentum filters, so pay attention to what the calculation is actually trying to accomplish.
The process begins with what Wilder called directional movement. Plus Directional Movement, or +DM, is identified when the current high has moved farther above the previous high than the previous low has moved upward relative to the current low. Minus Directional Movement, or -DM, is identified when the decline from the previous low is greater than the upward movement represented by the highs. Negative values are treated as zero in these calculations.
The resulting framework leads into the Average Directional Index, or ADX. ADX is technically a trend-strength filter rather than a momentum indicator, but it serves a similar practical purpose. It helps distinguish a meaningful directional movement from the minor pullbacks and sideways fluctuations that can otherwise trigger too many signals. The objective is the same one that has appeared repeatedly throughout this discussion: participate when the market is behaving directionally and avoid allowing small, ordinary fluctuations to dictate every trading decision.
Using the Rest of the Price Bar: The Stochastic Oscillator
There is considerably more information contained in a price bar than simply whether the market went up or down. Where the security closes within its daily range can tell you something about the character of the move. A close near the high, particularly when each successive high is above the preceding high, describes more than an ordinary uptrend. It suggests that the advance is gaining force. The same principle works in reverse. During a sell-off, you expect the closing price to remain near the bottom of the day's high-low range. During a healthy rally, you expect the close to appear toward the upper end of that range. The relationship between these pieces of information is what gives the stochastic oscillator its purpose.
Two relationships are especially useful. The first is the size of the high-low range over a specified number of days. The second is the location of the closing price within that range over the same period. When the calculation uses the low as its reference point, the resulting indicator is the stochastic oscillator. When the calculation is turned around and uses the high, you get Williams %R, named for its developer, Larry Williams. The terminology surrounding the stochastic oscillator is unfortunate enough. "Stochastic" refers to randomness, which seems almost backward when the whole purpose of technical analysis is to find some order in what otherwise looks like random price movement. Then George Lane, who developed the indicator, assigned letters to its components. The first became %K and the second became %D. %K is generally referred to as the fast stochastic, while %D is the slow stochastic. Fortunately, you do not have to carry this terminology into every other corner of technical analysis. These names belong almost entirely to the stochastic oscillator.
Step 1: Putting a Number on the Fast Stochastic, %K
The %K calculation compares today's closing price with the lowest low of the selected lookback period and then places that difference in relation to the total high-low range during the same period. Five days is the traditional setting. The calculation is a normalization process because the result is converted into a common scale regardless of the actual dollar price of the security. The final ratio is multiplied by 100, producing an oscillator that moves between 0 and 100. Readings are often interpreted using broad zones such as 30 to 70 or 20 to 80, depending on the convention being used. Your software will normally allow you to change the lookback period if five days does not suit the way you are studying the market.
The basic calculation is:
%K = (Current Close - Lowest Low) ÷ (Highest High - Lowest Low) × 100
The purpose is straightforward. %K tells you where the current closing price sits within the recent range. If today's close is higher than yesterday's and neither day establishes a dramatically different range, today's close is farther above the recent low than yesterday's close was. The arithmetic therefore produces a larger %K reading, and the stochastic line rises. The indicator is attempting to measure the energy behind the movement by examining where the market is closing in relation to the territory it has recently traveled.
There is, however, an important trap here. Suppose the five-day range extends from $5 to $12 and today's closing price is $12. The numerator becomes $12 minus $5, which is $7. The denominator is the highest high, $12, minus the lowest low, $5, which is also $7. Seven divided by seven equals one, and one multiplied by 100 produces a stochastic reading of 100. The indicator is therefore telling you that the close is at the very top of the recent range. Mathematically, that is perfectly correct. Practically, it can be rather useless if you interpret the reading too literally.
Why can the stochastic oscillator give you a misleading signal? Because a reading of 100 is classified as overbought, yet the security may simply be making a powerful new high and showing no meaningful sign of weakness. You already know that the price has closed at a new high. The stochastic oscillator has not necessarily discovered some hidden reversal that price failed to reveal. It has simply converted the location of that close within the recent trading range into a numerical reading. When the current close is at the top of the recent range, the calculation can reach 100 even though the underlying trend remains strong.
Consider a security that has been advancing sharply, with several gaps appearing during the move. As the advance continues, the %K line eventually reaches 100 percent. Viewed mechanically, the indicator now labels the security overbought. A trader who treats %K as a simple buy and sell mechanism might therefore sell the position as soon as that reading appears. The problem is that the security can continue advancing despite the extreme reading. In this example, the price moves another $10 higher after the stochastic reaches 100. The supposed sell signal would have removed the trader from a profitable position while the strongest part of the advance was still taking place.
This is one of the limitations you must keep in mind when working with stochastic readings. An oscillator can reach an extreme precisely because the trend is behaving well. A strong advance naturally pushes the closing price toward the top of its recent range, which can force the stochastic toward 100. A powerful decline does the same thing in reverse, pushing the indicator toward its lower extreme. What appears to be an overbought or oversold condition may therefore be nothing more than evidence that price is moving decisively in one direction.
For that reason, the stochastic oscillator generally has more value when the market is moving sideways than when it is locked into a powerful trend. In a range-bound market, an extreme reading can be more useful because price is repeatedly moving between established upper and lower boundaries. In a strong trend, however, the oscillator can remain extreme while price continues moving in the same direction. Williams %R has essentially the same weakness because it is closely related to the stochastic calculation, with the scale displayed in the opposite direction. The terminology changes and the numbers are inverted, but the underlying problem remains.
Step 2: Refining %K With %D
The second component of the stochastic oscillator is %D. Rather than following every movement of %K, %D smooths the faster line by applying a short-term simple moving average to it. The traditional calculation uses a three-day simple moving average of %K. Because a moving average smooths and consequently slows a series, %D is known as the slow stochastic or the smoothed indicator. When both lines are displayed together, their crossings become the principal trading signals.
The formula is:
%D = Three-Day Simple Moving Average of %K
The interpretation is familiar. When %K crosses above %D, the crossover is treated as a buy signal. When %K crosses below %D, it becomes a sell signal. A crossover occurring near an overbought or oversold region may appear especially attractive because the two pieces of information seem to reinforce one another. But do not assume that every crossover will occur at one of those extremes. The lines cross frequently in the middle of the range, and those signals still have to be considered in the context of the price movement.
There are plenty of variations available. You can introduce a slowing factor into %K, alter the number of days used in the calculation, and make other adjustments depending on the software and methodology you are using. This flexibility can be useful, but it also creates an opportunity to spend far too much time adjusting the indicator until the historical chart looks appealing. A parameter that makes the past look wonderful is not automatically a parameter that will tell you anything useful about the next market move.
Fiddling With the Stochastic Oscillator on the Chart
The stochastic oscillator can reveal patterns of its own, and this is where it becomes more useful than simply labeling a market overbought or oversold. Suppose the indicator produces three successive higher highs while the %D line repeatedly rises above the %K line on the right side of those peaks. The price bars themselves may not make the change particularly obvious, yet the oscillator is showing that something has changed in the underlying movement. Instead of simply measuring where the latest close sits within its recent range, you are now watching the behavior of the indicator from one swing to the next.
The sequence matters because the stochastic is beginning to display a pattern that can provide information about momentum. Three progressively higher highs in the oscillator suggest that the internal strength of the movement is improving, while the repeated crossovers between %K and %D provide another indication that the relationship between recent price movement and its rate of change is shifting. This does not mean the price must immediately reverse or accelerate. It means the indicator is giving you evidence that may not yet be obvious from the price bars alone.
This is one reason experienced traders study the shape and progression of an oscillator rather than concentrating only on whether it has reached an arbitrary extreme. The useful information can sometimes appear in the sequence of highs, lows, and crossovers before price makes the corresponding move. The oscillator is not predicting the future in any mystical sense. It is simply organizing the recent price action in a way that can expose changes in momentum that are easy to overlook when you examine the bars by themselves.
Look closely, however, because the pattern is not as mysterious as it might first appear. Beneath the first rise in the stochastic, only two days actually produce lower closes, with another close simply repeating the previous one. You could draw a support line beneath those lower closes, or beneath the series of "knees," which is the term Lane used for these turns. In this particular example, the final downside crossover occurs five days before the price breaks the support line. That is useful information, but it should not be turned into a rule that every stochastic crossover will precede a support or resistance break by five days. Sometimes it will. Sometimes it will not.
It is easy to become fascinated by the stochastic oscillator because there always seems to be another pattern to discover. Traders particularly like to search for divergence. A bullish divergence occurs when the price establishes a new low while the stochastic fails to make a corresponding new low and instead forms a higher low. The reverse condition is bearish divergence, where the price makes a higher high but the stochastic makes a lower high. In both cases, the underlying idea is the same. Price and momentum are no longer confirming one another.
Once traders begin finding these divergences, they often take the analysis another step and examine where the divergence occurs on the stochastic scale. Is it above or below the 50 percent midpoint? A bullish divergence occurring above 50 percent is generally treated as a more favorable configuration within this particular framework. This is another example of how the stochastic oscillator can be used as an additional layer of information rather than simply as a machine that spits out buy and sell instructions.
In many respects, the stochastic oscillator is bar reading with the volume turned up. Instead of merely looking at where the close sits inside today's bar, you are examining that relationship across a series of bars and translating it into an oscillator. It became extremely popular when technology made short-term swing trading much easier to conduct. The appeal was obvious. The promise was that you could trade more actively and respond to changes in price behavior much as professional traders do. There was some truth behind that sales pitch, particularly in the willingness of professional traders to abandon a position when the security stops performing, but the popularity of the indicator also produced exaggerated expectations about what it could accomplish.
Before becoming enamored with the stochastic oscillator, keep its weaknesses firmly in mind. It has very little ability to identify the existence of a trend by itself, and its signals can encourage you to exit a profitable position much too early. The same characteristic that makes the oscillator interesting in a sideways market can make it misleading in a strongly trending one. When a security has been trending for an unusually long period, you may become nervous and begin wondering when the move will finally end. That is exactly when the stochastic oscillator can become dangerous. It will repeatedly reach extreme readings because the price keeps doing what a strong trend is supposed to do.
Do not confuse an extreme oscillator reading with proof that the trend is finished. A market can remain overbought while it continues rising, just as it can remain oversold while continuing to fall. The stochastic oscillator can help you examine the internal behavior of the price bar and can provide useful information about momentum and divergence, but it should not be allowed to overrule the larger trend simply because its line has reached 80, 90, or 100. The stronger the trend, the more careful you must be about treating an oscillator extreme as a reversal signal.
Estimating Volatility
Volatility is simply a way of describing how much prices move. You can look at the total distance between the high and low over a particular period, or you can measure how far prices are varying from some central reference point, such as an average. Both approaches are useful. The greater the volatility, the greater the risk, but also the greater the opportunity. Volatility is not some isolated statistic that belongs to one particular indicator. It is woven into technical analysis and shows up in momentum measures, price ranges, and many other calculations. The familiar VIX, often described in terms of market fear and greed, is another well-known application of the concept.
For practical trading purposes, a change in volatility should make you think about a change in the range of prices that may lie ahead. A highly volatile security can produce a wide assortment of possible outcomes, while a quiet security generally produces a narrower and therefore more predictable range of outcomes. This is why volatility deserves your attention. You are not measuring it merely because the number looks interesting. You are measuring it because changing volatility can affect where you place a profit target and where you put a stop loss. If the market's behavior changes, your expectations for both gain and loss should change with it.
We will deal with three methods of estimating volatility and the advantages and shortcomings of each. From there, the discussion turns to the most widely recognized application of volatility in technical trading, the Bollinger Band. Another type of band, based on average true range, is also introduced. There are two broad varieties of volatility worth distinguishing. Historical volatility describes the amount of price variation that has actually occurred in the recent past, and that is the type discussed here. Implied volatility belongs primarily to option analysis and concerns what the market is implying about future price movement. It is a separate subject and is not covered in this book.
Catching a Slippery Concept
Volatility sounds like a simple word until you try to define it precisely. Traders use it constantly, often without agreeing on exactly what they mean, and statisticians themselves can argue over the finer points of the definition. From a mathematical perspective, volatility is commonly associated with the standard deviation of price changes. Standard deviation is certainly not the only possible measure, but for most technical analysis purposes it provides a useful way to think about variation.
In ordinary trading language, volatility is often used almost interchangeably with variance, and that is the approach taken here. Variance looks at how far the individual price observations are from a central measure such as a moving average. You calculate the differences, square them so that negative and positive deviations do not cancel one another, add the results, and divide by the number of observations. The squaring process has an important consequence. Large departures from the average receive considerably more weight than small ones. The more extreme the deviations and the more frequently they occur, the greater the resulting measure of volatility.
Variance itself is not generally used as a stand-alone trading measure. The reason is that variance and standard deviation are closely connected, with standard deviation being essentially the square root of variance. If the words "square root" make you want to close the book, don't. Your charting software will perform these calculations for you. The important thing for the trader is understanding what the number represents and how to use it, not proving that you can perform the statistical calculation by hand.
Time frame is critical when discussing volatility. A security can appear quiet on one time frame and highly volatile on another, and neither observation is necessarily wrong. A daily chart gives you one picture of price variation. An hourly chart gives you another. A weekly chart gives you another. Many of the sweeping statements traders make about volatility become contradictory simply because nobody bothered to specify the period being measured. Your choice of time frame therefore influences not only the volatility number you see, but also the amount of risk you believe you are taking and the kind of trading opportunities you are willing to pursue.
Volatility and risk are closely related, but you should not make the mistake of assuming that every volatile security is automatically a poor trade. Volatility changes the amount of uncertainty you must deal with, and two traders can use exactly the same indicators while producing very different results because they respond differently to that uncertainty. One trader may be comfortable with wide price swings and establish a stop that gives the position room to move, while another may find the same movement unacceptable and exit much sooner. The indicators have not changed. The risk environment has.
The difference becomes obvious when you compare a low-variance price series with a high-variance one. The low-variance series moves in a relatively orderly fashion, with prices staying within a narrower range and producing smaller swings from one observation to the next. The high-variance series is much less restrained. Its price movements are larger, and the distance between successive highs and lows can expand considerably. You can usually see the difference before doing any calculation. One market appears calm, while the other seems to jump around.
The important point is that high variance increases the amount of room available for a trade to move against you. A security can be trending in the direction you expect and still carry substantially more risk because its normal price swings are larger. If you buy a steadily rising security that normally moves only a small amount each day, a modest decline may have little significance. The same percentage decline in a highly variable security may occur within a few hours and could be large enough to trigger your stop before the larger trend resumes. The direction of the trend has not changed, but the path taken to get there has become much more dangerous.
This is why volatility must be considered alongside the trading signal rather than treated as a separate curiosity. A trend may tell you where price appears to be going, while variance tells you something about how violently it may get there. Higher variance creates greater profit opportunity because larger movements are available, but it creates greater loss potential for exactly the same reason. You cannot take one side of that equation and ignore the other. Higher variance means higher risk, and your trade management must account for it.
How Volatility Arises
Crowd behavior is one of the major forces behind changes in volatility. When traders become excited about a developing move, volatility often expands. If prices are rising, buyers begin anticipating still higher prices. If prices are falling, sellers may rush to get out before the decline becomes worse. The result can be a rapid expansion in the daily range as participants become increasingly aggressive.
There is an interesting tendency for volatility to behave differently around turning points. Volatility can become unusually quiet before a major change in direction and then expand sharply as the new move begins its first significant thrust. This is not a guarantee, of course. One of the irritating facts about volatility is that it can sometimes become very high or very low without giving you a satisfying explanation. Markets do not always provide a clean reason for everything they do.
High volatility means that trading carries greater immediate risk, but it also creates greater profit potential. Low volatility reduces the immediate range of potential loss, but it can also limit the amount of movement available to capture. Volatility itself is therefore neither good nor bad. What matters is whether its behavior is stable enough for you to make reasonable estimates of the possible gain and loss. A relatively stable volatility pattern allows you to establish expectations with greater confidence.
Every security develops its own normal level of volatility, and that normal level can change as the fundamental situation changes and as the population of traders in the security changes. Over time, you may even begin to think of a security as having a particular "personality." That personality is really a reflection of how the participants trading the security collectively respond to risk. A stock whose traders routinely tolerate large price swings will behave differently from one whose participants react quickly to even modest changes.
Low Volatility With Trending
A low-volatility uptrend is usually easy to recognize because the price movement has a certain order to it. Prices continue working higher, but they do so without constantly producing large reversals or violent swings. Each advance and setback remains relatively contained, allowing the underlying direction to stand out clearly. The calmer the movement, the easier it is to distinguish the trend from the ordinary noise that surrounds it.
A trending security with relatively low volatility creates a useful trading environment because two favorable conditions are present at the same time. Price has established a directional bias, while the day-to-day fluctuations remain reasonably controlled. That gives you a clearer basis for expecting the position to continue moving in the direction of the trend, and it also makes stop placement more practical. Your stop does not have to sit so far away simply to survive the security's normal fluctuations, nor are you constantly forced to watch an otherwise sound position get shaken out by a sudden reversal.
There is another advantage that does not show up in a volatility calculation. Low-volatility trends are easier to live with. A position that advances steadily and experiences only modest setbacks does not demand the same emotional tolerance as one that repeatedly swings sharply against you before recovering. You are still taking risk, and the trend can still fail, but the price action gives you fewer reasons to question the trade every few minutes. When the market moves in an orderly fashion, your indicators are easier to interpret and your trading decisions are generally easier to execute.
This is why volatility should always be considered together with direction. A clear trend tells you that price has a directional tendency, while low volatility tells you that the movement is relatively controlled. Neither condition guarantees a profitable trade, but together they create a market environment in which the existing trend is easier to recognize, easier to manage, and less likely to be disrupted by ordinary price noise.
Low Volatility Without Trending
A security moving sideways with very little variation from one day to the next presents a different problem. If the price is confined to a narrow range and has no directional bias, there may be little reason to expect a meaningful gain in the time frame you are examining. The security is moving, but it is not giving you enough movement to work with.
One solution is to reduce the time frame. A daily chart may show almost nothing of interest, while an hourly chart may reveal several smaller peaks and troughs that can actually be traded. You are not changing what the security did. You are changing the scale at which you are observing it.
An even more interesting situation develops when the price moves sideways but the range of the individual bars begins changing. If the high-low ranges start contracting or expanding, you have something worth watching. Range contraction and range expansion can provide important clues about an approaching breakout. You still do not know which direction the eventual breakout will take, but you can begin preparing for the possibility.
During a sideways consolidation, price bars often maintain a uniform height before transitioning into a distinct phase where their individual high-low ranges visibly shorten and tighten. This compression reflects a sharp drop in short-term volatility as buyers and sellers reach a temporary equilibrium. Such contractions routinely precede major breakout moves, though the shrinking ranges alone do not signal whether the eventual surge will resolve upward or downward. That distinction matters: you are preparing for imminent movement, not pretending to know its direction in advance.
High Volatility Without Trending
A sideways market is already difficult enough. Add high volatility to it and you have what many traders would call a nightmare. The security is moving dramatically, but it is not moving anywhere. Bulls push prices higher, bears knock them back down, and the resulting chart becomes a succession of sharp reversals with little dependable direction.
One or two days may produce a large move in one direction, followed immediately by an equally aggressive move in the opposite direction. These sharp reversals make entries difficult because a move that looks like the beginning of a trend can disappear almost immediately. They also make systematic stops harder to use because ordinary price noise can be large enough to reach a stop even when the larger trading range has not actually changed.
When volatility becomes extremely high and the security has no recognizable trend, the sensible choices become narrower. You can stop trading that security until the conditions become more orderly, or you can move to a shorter time frame where the individual swings become more manageable. What you should not do is assume that bigger price movement automatically means better opportunity. Sometimes a large amount of movement simply means a large amount of uncertainty.
Measuring Volatility
You can measure volatility with simple calculations or with more elaborate statistical tools. In financial analysis, the term usually brings standard deviation to mind, and that is covered shortly, but standard deviation is not the only useful way to look at price variation. Before getting into the mathematics, it helps to understand some of the simpler measurements and what they can and cannot tell you.
Tracking the Maximum Move
One straightforward approach is to measure the largest price movement over a specified number of days. This is sometimes called the maximum move or gross move. Take the highest high over the period and subtract the lowest low. A ten-day period is one possible choice, but the number of days can be adjusted to suit the trading method.
The resulting number can be useful for establishing expectations. You can use it as a rough guide to a profit target, sometimes described as the maximum favorable excursion, or as an indication of the worst-case movement against the position, the maximum adverse excursion. When the maximum move is large, there is obviously room for a substantial gain. There is also room for a substantial loss. That distinction is easy to overlook when a large range appears attractive.
The bigger problem is that volatility does not remain fixed. A security may be capable of moving $10 over a particular period today and only $4 over the same period several months later. If volatility contracts, the amount of profit you can reasonably expect must contract along with it. This is why projecting the volatility of the last 30 days into the next 30 days is dangerous. Such a projection assumes that market conditions will remain unchanged, and market conditions do not remain unchanged.
A straight trend line or price channel is already a projection into the future, but the consequences of being wrong may be relatively modest. Projecting volatility is another matter. Volatility can change rapidly, and assuming that a recent level of volatility will continue indefinitely can produce an unrealistic estimate of both opportunity and risk.
Maximum Move and Trend
Do not assume that volatility and trend must move together. The trend in price can increase while the degree of volatility decreases, or volatility can rise while the underlying trend becomes weaker. At other times the two will move in the same direction. There is no requirement that the size of the price swings and the direction of the price trend agree.
This means that knowing the trend of the maximum move does not necessarily tell you anything about the trend in price itself. The reverse is also true. A security can maintain a clear directional trend while its volatility changes dramatically. Volatility and trendedness are related to trading risk, but they are not the same thing.
Maximum Move and Holding Period
Two securities can travel from an identical starting low to the exact same ultimate high over the same timeframe, yet present vastly different risk profiles along the way. An orderly price series progresses steadily with small, predictable pullbacks from its lowest point to its peak. In contrast, a disorderly price series fluctuates wildly with aggressive, erratic swings along the exact same overall path.
If you looked only at the net high-low movement over the entire period, you might conclude that both securities presented essentially the same risk. However, measuring only the total high-low range fails to account for the turbulence of the journey itself. A volatile, unpredictable path significantly increases the likelihood of being stopped out prematurely or experiencing severe drawdowns before the ultimate move completes. You should therefore be cautious about treating an expected holding period as something carved permanently in stone. The path matters, not merely the beginning and ending prices.
Considering the Standard Deviation
The maximum move measures the gross distance between the lowest low and highest high, but it misses the character of the movement between those two points. Standard deviation provides a way to describe this additional risk. It measures the dispersion of prices around an average. The farther the observations spread from the average, the greater the standard deviation. The calculation belongs to the same general statistical family as standard error, but its purpose here is to describe the actual variation of prices around a central value.
In technical analysis, standard deviation is commonly calculated around a moving average. It tells you how widely prices are distributed around that centerline. A narrow distribution produces a small standard deviation. A wide distribution produces a large one. This gives you a more useful description of the character of price movement than simply knowing the largest move from low to high.
You might reasonably expect to see a raw standard deviation indicator sitting on the chart at this point. Most charting programs provide one. In practice, however, traders rarely stare at standard deviation by itself because there are more useful ways to put the same information to work. Bollinger Bands are the obvious example. They incorporate standard deviation into a price-based framework that is much easier to interpret visually.
Using the Average True Range Indicator
Another way to estimate volatility is to examine the average high-low range over a selected number of days. The average true range, or ATR, is particularly useful because it improves on the ordinary high-low calculation by accounting for gaps. Where a gap makes the previous close more important than the current high or low, ATR incorporates that information into the range calculation. The mechanics include examples of expanding and contracting ranges.
Two conditions deserve particular attention. When the highs and lows move farther apart, the range is expanding and volatility is increasing. This creates a larger potential profit opportunity, but the potential loss expands by approximately the same principle. Greater movement cuts both ways. When the highs and lows begin moving closer together, the range is contracting and volatility is declining. It is tempting to conclude that the reduction in volatility automatically means that risk has disappeared. It has not. A contraction can continue only until the market breaks out of the compressed range.
The practical value of tracking ATR lies in its ability to highlight shifts in market participation. Following a sudden, unusually large price bar that establishes a major support level, the ATR indicator often begins falling. Even if you cannot know in advance whether that support level will hold, the behavior of the indicator reveals a crucial dynamic: volatility drops rapidly as the market settles into a tighter trading environment, failing to sustain the initial volatility spike even as prices continue to drift upward.
That divergence deserves close attention. When price continues advancing while the measure of its true range contracts, the underlying trend is losing its expansionary force. As buyers struggle to push prices aggressively higher, the price action typically develops a series of lower highs against the flat support level, forming a descending pattern until support eventually gives way. The ATR does not predict the exact timing or direction of the ultimate breakout, but it provides an early warning that price direction and volatility are no longer moving together.
Applying Volatility Measures: Bollinger Bands
The most familiar practical application of volatility is the Bollinger Band, developed by John Bollinger. The traditional construction places a 20-day simple moving average of closing prices in the center and then establishes an upper and lower band at two standard deviations from that moving average. Under the assumptions behind the calculation, the bands encompass roughly 95 percent of the variation around the average.
The usefulness of the Bollinger Band comes from putting price into context. Instead of looking at a $50 price and trying to decide whether $50 is high or low by itself, you can see where that price sits relative to its recent norm. The moving average establishes the center, while the bands expand and contract according to the amount of variation in the price. A relative high is therefore not simply a high price. It is a price that is high in relation to the security's recent behavior.
A price reaching the upper band can be interpreted as a continuation signal rather than an automatic sell signal. In a strong advance, the price may continue to "walk" along the upper band for some time. The same thing can happen along the lower band during a decline. A breakout through a Bollinger Band is still a breakout, and the first expectation should be that price may continue in the direction of the break rather than immediately reverse.
Every price thrust eventually runs out of energy, however, and Bollinger Bands can provide clues that the advance is losing force. One sign is that the price stops hugging the upper band and retreats toward the central moving average, or even farther. When a rally reaches a temporary peak, the price typically pulls back toward its central moving average as buying enthusiasm cools. If buyers make a second push to push the price higher but fail to exceed or even reach the previous high, this creates a classic double top structure. This inability to establish a higher relative high visually confirms that upside momentum is waning.
The second warning comes from the bands themselves. When the upper and lower bands begin moving closer together, the range is narrowing and volatility is contracting. Traders are becoming less willing to push the price to new extremes. Buyers are not aggressively establishing new highs, but sellers are not forcing new lows either. This narrowing is commonly called the squeeze, and it suggests that a breakout may be approaching.
Do not assume that a squeeze tells you the direction of that breakout. A consolidation phase following a squeeze often ends in a trend reversal, but a reversal is not required by the indicator. A breakout can just as easily occur in the same direction as the previous trend. When a breakout occurs, a rapid penetration to the opposite side can sometimes be a false movement—a head fake—rather than the beginning of a sustained, structural reversal.
A sudden shift across the bands often occurs when traders become overly eager to take profits after a large price advance. Following such an aggressive drop, the price will frequently experience brief upward pauses or minor bounces along its downward trajectory. These temporary recoveries represent classic pullbacks within an emerging move. While minor pullbacks often fail and allow the previous trend to resume, a recovery following a sharp, decisive break through the outer band is far less likely to be just an ordinary temporary retracement.
This is why Bollinger Bands should not be treated as a self-contained trading system. If you are trying to distinguish a genuine reversal from a head fake, bring other evidence into the decision. Momentum indicators such as the relative strength index and MACD can be particularly useful for confirmation. The band tells you that something unusual has happened. Another indicator can help determine whether the price action has the momentum necessary to make the move meaningful.
Applying Stops With Average True Range Bands
Bollinger Bands are not normally designed as stop-setting tools. Their upper and lower boundaries are equally distant from the moving average, which means that an upside breakout and a downside breakout are treated with essentially the same statistical distance. That symmetry makes sense for measuring volatility, but it is not always what you want when managing an existing trend.
Consider a strongly trending security. If you are long, a modest downward movement should not necessarily be treated as seriously as a comparable upward movement. The market has already demonstrated a directional bias. Because false breakouts are common, it makes sense to require a more substantial move against the trend before declaring that the trend has actually failed. This is where ATR bands provide an alternative.
For a band to function as a stop, you want asymmetry. During an uptrend, the lower band should be placed farther from the average so that a downward move must be more severe than the ordinary upward fluctuations before the stop is triggered. During a downtrend, you widen the upper band so that an upward move must travel farther before it is treated as a genuine reversal.
Instead of constructing the bands from standard deviation, you can use a version of ATR. The adjustment is simple in concept. In an uptrend, increase the distance between the lower band and the average. In a downtrend, increase the distance between the upper band and the average. The wider band establishes a more demanding test for a countertrend move. Ordinary corrections can remain inside the band, while a sufficiently large reversal can finally break through it.
The approach is logical, but it does require more work. Before you can decide which band to widen, you must first determine the direction of the trend. ATR by itself measures the amount of movement, not its direction. Standard deviation likewise tells you about dispersion rather than whether volatility is bullish or bearish. Direction has to come from the price structure or from another indicator.
This arrangement creates a more demanding ATR test for a move against the established trend. An upside breakout must clear the upper band to confirm the continuing trend. A downside move during the uptrend must travel farther because the lower band has deliberately been widened. In the example, the lower band is 50 percent wider than the upper band, making it 150 percent of the ATR.
The final bar breaks through that widened lower band shortly after the support line itself has been broken. That combination is much more significant than either event by itself. You have a price breakout through support and a volatility breakout through the ATR band. The evidence therefore argues against treating the move as an ordinary retracement.
The trading response depends partly on the kind of trader you are. A longer-term position trader could have entered near the low on the left side of the chart and remained in the position until the support line failed. A swing trader could have sold at breakout 1, possibly establishing a short position, covered at point 2, and then sold again at point 3. These are different ways of using the same price information.
Very few traders should make the ATR breakout the sole reason for entering or exiting a trade. The ATR band tells you that volatility has expanded enough to cross a statistically meaningful boundary. It does not, by itself, explain everything happening in the security. Price structure, momentum, trend, support and resistance, and other indicators still have their part to play.