Custom Search

During the late 1950s, George C. Lane was actively involved with researching different types of oscillators to use as trading indicators. The name of each indicator was simply a letter from the alphabet, such as %A. The research took him through the alphabet twice before he decided to use two in particular, called the %K and %D. The stochastics oscillator was the result of this research. The term "stochastics" is actually a misnomer, as stochastics is a synonym for "random."

One component of a market top is that the daily closes begin to cluster near the highs of the daily range. The opposite occurs for a market bottom. Typically, the daily closes happen at the lows for each day. The stochastics oscillator is designed to forewarn of market reversals by using this information. The oscillator compares the current closing price with the highest high and the lowest low during a time period. The observation duration is anywhere from five to 14 days, while some traders go as far out as 28 days. The oscillator calculation is on a scale of zero to 100. Two lines, %K and %D, are produced to generate the market signals.

The calculation of the stochastics oscillator is a two-step process. First, %K:

%K = Raw Value = 100 {Current Close - Lowest Low14 / Highest High14 - Lowest Low14}

In this case, we are using the highest high and lowest low during the last 14 days. The second step is %D, which is a three-day smoothed version of %K:

This oscillator has two popular uses. First, the oscillator can be used to define an overbought or oversold state of the market. If the %K and %D are above the 80% level, an overbought situation usually exists in the market. An oversold condition is signaled if the %K and %D are below 20%.

A sell signal is not necessarily generated until a divergence exists between the market and the oscillator. For example, during early March IBM was edging higher while the oscillator was actually declining (Figure 1, A and B). This divergence between the price of IBM and the %K and %D lines indicated that the price was extremely vulnerable to a decline. During May, the price of IBM (Figure 1, C and D) was trending toward support (the low made in late March) when the oscillator began to diverge in oversold territory. The price of IBM advanced $20 from this oversold signal.

There is another version of the stochastics oscillator that is often used. Because the %K value can be quite volatile, traders have elected to plot %D along with a three-day moving average of the %D (Figure 2). This version is called the slow stochastics oscillator. Many traders prefer to use this version, as it produces less sensitive signals.

Overall, the stochastics oscillator is a popular tool to assist traders on recognizing turning points in the market. Like all oscillators it can reach overbought or oversold levels and prematurely warn of a market reversal. The observation of divergence between the market and the stochastics indicator can help reduce the number of false trading signals.

Thom Hartle is STOCKS & COMMODITIES' Editor.

FIGURE 1: Sell and buy signals are generated when the %K and %D reach extreme levels and begin to diverge from the market. The price of IBM (A) edged higher in March, while the indicator declined During May, the stock declined (C), while the indicator moved sideways (D), which successfully warned of a rally.

FIGURE 2: The slow version of the stochastics oscillator uses a three-day moving average of %D (dotted line) along with the normal %D.