Holiday Dummy Regression

Problem

Fit a regression of returns on a holiday dummy with an intercept and return the dummy coefficient: sum((d - mean_d) * (r - mean_r)) / sum((d - mean_d)^2).

Return 0.0 if the lists are empty, differ in length, or the dummy has zero variance. The coefficient is the mean return difference between holiday and non-holiday periods.

Hint

With a balanced dummy this equals the difference of group means.

Test cases

case 1

in: [[0.01,0.02,0.03,0.04],[0,1,0,1]]

expected: 0.010000000000000002

case 2

in: [[0.05,0.06],[1,1]]

expected: 0

case 3

in: [[0.01,0.02,0.03],[0,0,0]]

expected: 0

case 4

in: [[0.02,-0.01,0.04,0.01],[1,0,1,0]]

expected: 0.030000000000000002

Starter codePython

def holiday_dummy_regression(returns, holiday_flags):
    # Your code here
    pass

4 test cases · run your code against them in the editor