Doing QBank and I’m finding for some questions where they give you two different stock’s return over a period of say 4 years. Year Stock A Stock B 1 5% 4% 2 3% 7% 3 6% 8% 4 8% 10% I know probability covariance is generally = p(state 1)(x1-u1)(y1-u2) + p(state 2)(x2-u1)(y1-u2)…etc. That one involves no dividing by the number of time periods/observations, the data above involves dividing by n-1. My questions are: 1) Just how many different kinds of covariance formulae are there that we need to cover? 2) Do we ever use n-1? I know its generally used when estimating a sample parameter, but are we ever given a population parameter and use simply ‘n’? Thanks in advance