Consider a player with mean-variance preferences given by u = r_P - s_P choosing a portfolio P between two stocks g1 and g2 that are perfectly negatively correlated (i.e. = -1) where the expected return of stock 1 is larger than that of stock 2 (that is r₁ > r₂) and the risk measured by the standard deviation of stock 1 is larger than that of stock 2 (that is s₁ > s₂). Carefully explain how the optimal portfolio choice of this player is affected by an increase in the risk of stock 1.