The normalized change in theoretical regression lines of the current and specified periods period of bars.
The Linear Regression Slope indicator plots the slope value of theoretical regression lines of the current bar and the previous n-1 bars (n = regression periods).
The slope values are then normalized by multiplying the raw slope values by 100 and then dividing by the price (raw slope * 100 / price). The normalization of the slope values basically shows the percent change in price per bar of the regression (best fit) line. For example, if the normalized slope is 0.15, then the regression line is rising at a rate of 0.15% per bar. Similarly, a normalized slope of -0.25 would indicate that the regression line is declining at the rate o 0.25% per bar.
Th Linear Regression Slope plotted on a daily chart of USD/ZAR 
Source: VT Trader
