This paper aims to provide a theoretical underpinning of the dynamic efficiency model pioneered by [Ahn, S.C., Good, D.H., Sickles, R.C., 2000. Estimation of long-run inefficiency levels: A dynamic frontier approach. Econometric Reviews 19, 461â492]. In the context of a quadratic loss function this paper formulates a multi-period forward-looking rational expectations model on the evolution of the technical inefficiency level, which correctly produces a dynamic panel data model. The model is illustrated using panel data of 112 French banks. Encouraging evidence of superiority in favor of the model is reached. Substantial cost inefficiency prevails in this industry, where the constituent banks are characterized as having volatile adjustment speeds toward their long-run steady states. The sample banks exhibit increasing returns to scale and product-mix economies.