Covariance is a way to measure how “in sync” (or N’sync, if that’s your thing) two variables are. If it’s positive, then on average, they move in the same direction. If it’s negative, they move in opposite directions, on average. The magnitude of covariance is difficult to interpret without a reference, which is why we can use correlation or beta to get a better idea of what the magnitude of the covariance means. Also, if the covariance is positive (negative), the correlation between the variables is positive (negative).
so, cov i,j = beta i * beta j * var(m) = cov(i, m) / var (m) * cov(j, m) / var (m) * var(m) = cov(i, m) * cov(j, m) / var (m) = (corr i,m * std i / std m * corr j,m * std j / std m) / var(m)