indicator construction

Most technical indicators fall into a group called momentum indicators in that they attempt to measure the velocity of directional movement in whatever is being evaluated. One reason momentum indicators are so popular is that, unlike price-based indicators, which typically generate buy and sell signals after-the-fact, momentum indicators actually give an indication of a change in trend direction before the actual change in price itself To better understand this concept, consider a simple momentum oscillator of todays price minus the price 10 days ago, or the 10-day rate-of-change.

As Figure 2.2 illustrates, we begin on day 10 when the closing price is 48.50. The price 9 days prior (on day 1) was 50.75. So when we subtract todays price (48.50) from the price 9 days ago (50.75) it results in a value of-2.25, which is plotted below the zero line. By following this simple procedure each following day, we develop a momentum curve that we then plot directly above or below the price curve for the same time period. As the price moves down by the same increment each day, the curve becomes a straight horizontal line between days 10 and 14. Then on day 15 the price turns up by 25 points while the momentum plot turns up by 50 points. Thus price momentum is accelerating twice as fast as is the price. The momentum curve continues this rate of change until day 23 when the change becomes constant once again, even though price continues to advance at the same rate.

On day 28 price begins to level out at 50.75, yet the momentum curve begins to drop. Should price continue to move horizontally, the momentum curve will continue to fall until the 38th day (10 days later) at which time both the momentum curve and

Figure 2.2 Price versus momentum, bottom half shows example price

price will be moving horizontally. When plotted in this manner, momentum always changes direction in advance of price. That is because the momentum curve is reflecting the rate of change, or velocity, of price movement.

Momentum indicators can be constructed to be either oscillator based or threshold based. The momentum curve in an oscillator-based indicator revolves (or oscillates) above and below a zero line. Buy signals are generated when the oscillator curve reaches an extremely low (oversold) level and turns up. Sell signals are generated when the oscillator curve reaches an extremely high (overbought) level and turns down.

How high is high and how low is low depends upon what is being measured as well as the number of periods in the calculation. As with all technical analysis, interpretation is an art form and requires a substantial amount of experience in order to get a "feel" for each indicators individual behavior.

Threshold-based momentum indicators are constructed so that the momentum curve travels between an absolute scale of 0 and 100. When the momentum curve travels to the high end of the range, the issue is considered overbought. Conversely, when the curve travels toward the low end of the range, the issue is considered oversold. Buy and sell signals are generated when the issue reaches an extreme, changes direction, and penetrates a predetermined threshold. Here again, sensitivity of the indicator is determined by what is being measured as well by the time parameter selected.

Weight of the Evidence

As mentioned, to build an effective timing system, it is not enough just to know about the existence of technical indicators, one should thoroughly understand how they are constructed and what they measure. One of the biggest traps most aspiring technicians fall into is following several indicators with different names, but that all measure the same thing (e.g., volatility, cyclicality). That is like posing the same question to several people of different languages. Although their responses may sound different, they are still basically answering the same (single) question. And whether you receive 5 or 50 responses to the question, it still only addresses one part of the equation.

A better approach would be to follow a handful of indicators that are very well understood, reliable, and selected to measure different aspects of market behavior. This way the technician can evaluate the "weight of the evidence" from a number of perspectives before making an investment decision. The following are several popular indicators that, when combined, can provide a meaningful assessment of an issues underlying strength.

Bollinger Bands (BB)

One aspect of a stocks behavior that the technician will want to know is the issues current volatility compared with its volatility over the past x number of days. To measure such volatility-based price action John Bollinger of Bollinger Capital Management developed a technical indicator called Bollinger Bands (see Figure 2.3).

Bollinger Bands are trading bands based on the volatility of prices. However, they differ from ordinary trading bands in that they are based on a standard deviation above and below a simple moving average as opposed to a fixed percentage. As a result, the spacing of the bands is determined by the volatility over the past x number of periods tusually days). Thus, the greater the volatility the broader the Bands, and the lower the volatility the narrower the Bands. Normally, Bollinger Bands are constructed using 2.0 standard deviations above and below a 20-day simple moving average. Bollinger set forth the following rules for interpreting Bollinger Bands:

• Sharp moves in the market tend to occur after the Bands tighten, and the closer to the average the better. Since reduced volatility denotes a period of consolidation, the first increase in volatility after a consolidation tends to mark the start of the next move.

• When prices move outside the Bands (either upward or downward), it signals a continuation of the move until prices move back inside the Bands.

• Moves starting at one Band tend to move to the opposite Band before reversing.

• Rallies and reactions that temporarily take prices outside the Bands usually represent exhaustion and are associated with trend reversals.

The parameters for Bollinger Bands are the number of periods used in the simple moving average and the standard deviations that define the upper and lower Bands. The greater the number of periods in the moving average, the less reactive the Bands. Conversely, the fewer the number of periods in the moving average, the more reactive the

Figure 2.3 Price chart with Bollinger bands above and below price.