Here's how you can use the Relative Strength Index

Last week I had an excellent example of bearish divergence in an oscillator forecast an impending decline in an individual equity. Bearish divergence is often used with oscillator analysis and often gives the best results. For those of you who are feeling a bit lost, let me give a couple of definitions before proceeding.

First, an oscillator is any indicator that is generally used to determine whether or not the entity you are analyzing is overbought or oversold. Common examples include the Relative Strength Index (RSI) and Stochastic Oscillator. They are called “oscillators” because they are able to move freely from 0 to 100 (and above and below if they aren’t normalized).

Second, bearish divergence is when the oscillator makes a lower high but the price of the entity makes a higher high. In my example from last week, both Stochastics and the Relative Strength Index made a lower high while the price continued to rise in the equity. For the purpose of this column, I will focus your attention on the RSI.

The Relative Strength Index is a measure of a stock’s recent gains over a fixed period of time (in this case 14 days) versus its’ recent losses over the same period of time. A rising RSI means that the stock has had more days that gained over the last 14 days then days that lost.



Focusing our attention on the chart above, one can see that on April 18th, the price of Allegheny Technologies, INC. (ATI | Quote | Chart | News | PowerRating) closed at $72.72 and the RSI was at 78.48. Please also note that divergences of this type are more meaningful when occurring at extremes. Traditionally, the RSI is considered to be in extreme territory when it is over 70 or under 30.

On April 24th, the price of ATI closed at $74.52 and the RSI was at 76.04. Therefore, the price rose but RSI dropped and created a “bearish divergence”. Had the price dropped and the RSI risen, then a “bullish divergence” would have been created. This bearish divergence, perhaps because other traders noticed it, or perhaps because of some extraneous force, correctly forecast a drop in this particular stock. This is not a tool I often use, but it can obviously be useful in certain circumstances.