On the optimality of target volatility strategies