When using dummy variables to represent categorical variables in regression analysis, the general rule is:

Number of dummy variables = Number of categories - 1

Given three categories (small-cap, mid-cap, and large-cap):

• Number of categories = 3
• Number of dummy variables needed = 3 - 1 = 2

Therefore, 2 dummy variables would be required to distinguish among small-cap, mid-cap, and large-cap stocks in a multiple linear regression. One category (typically the first one) serves as the reference category, and the other two dummy variables capture the differences from this baseline.