"In investment analysis, an oscillator is a mathematical tool used to identify turning points in prices and is useful in timing the purchase and sale of a security. The regression oscillator (RO) fluctuates around a moving linear regression line of closing prices. It can be used to screen a database of stock charts for stocks in a long-term upward trend, determine the appropriate time to buy an upward-trending stock, and give an early warning signal that the long-term upward trend may be reversing."
"Here's how to develop the RO using technical analysis software and also by using a computer spreadsheet if technical analysis software is not available. Let me emphasize two points, however. First of all, the RO may be used in conjunction with any method an investor uses to select a particular stock. The oscillator's purpose is to better time the purchase after the company has been selected, but because calculating the RO requires calculating the slope of the regression line, an investor may want to use the regression slope to select the stock itself. Second, although future price patterns may not repeat the past, stock prices tend to fluctuate within a range. An oscillator is useful in maximizing gains and minimizing losses regardless of the future direction of prices."
APPLYING THE RO
The value of the RO can be best illustrated by applying it to a specific company. Figure 1 shows daily price bars, a graph of Williams' %R (a traditional oscillator), the RO, and the slope of the moving regression line divided by the close (slope/close) for California Amplifier [CAMP] between May 1995 and June 1996.
Horizontal lines have been drawn on the RO graph at 14%, 0% and -14%. The RO does not drop significantly below -14% until June 11, 1996, but in 1995 comes close on May 30 (-13.2%), October 26 (-11.8%), November 22 (-13.8%) and December 18 (-14.1), and in 1996 on April 3 (-12.3%).
Assume that it is April 4, one day later. You see that the RO has a normal lower limit of 14% by looking at this graph. The RO dropped to -12.27% on April 3 and has increased to -12.25% on April 4 (point A). The regression slope on April 4 is 0.92% per day (point B) and has been rising from 0.15% since January 11. A rising slope/close shows that the trend is not only upward, but that the steepness of the trend is increasing, a desirable characteristic for a stock being considered. This would be an ideal time to buy the stock at $24.75 (point C). The stock rises to $38.75 on April 25 (point D) before turning down, a 57% profit (excluding commissions) in 14 trading days.
