By eliminating the influence of statistical noise, stochastic frontier techniques permit the estimation of the best-practice value of a firm’s investment opportunities and the magnitude of a firm’s systematic failure to achieve its best-practice market value – a gauge of the magnitude of agency costs. These frontiers are estimated from the performance of all firms in the industry and, thus, capture best-practice performance that is, unlike Tobin’s q ratio, independent of the managerial decisions of any particular firm. Using the frontier measure of performance applied to 2007 data on top-tier, publicly traded US bank holding companies, the authors obtain evidence on market discipline: they find that higher managerial ownership at most banks tends to align the interests of insiders with those of outside owners and to be associated with improved financial performance; at most banks, higher blockholder ownership is associated with improved financial performance obtained from blockholders’ monitoring; and, at most banks, higher product-market concentration is associated with poorer financial performance and the so-called managerial quiet life. Using the frontier measure of investment opportunities, the authors find evidence that banks with relatively higher-valued investment opportunities achieve less of their potential market value, while banks with lower-valued opportunities achieve more of their potential value. In spite of their lower-valued opportunities, these banks, on average, achieve the same Tobin’s q ratio and, thus, appear better able to exploit their less valuable investment opportunities. The authors’ results suggest that higher-valued opportunities may reduce managers’ performance pressure and provide a stronger incentive to consume agency goods.