Resource allocation with Time Series DEA applied to Brazilian Federal Saving banks

One limitation in the economic analysis of efficiency and productivity is the impossibility to determine whether a service organization has reached their optimum output-to-input configuration, i.e. whether efficient units could be more efficient or whether inefficient units have reached their maximum potential and could not improve their performance. In this work, the usage of time series data instead of cross-sectional data from different DMUs is motivated to avoid this problematic of comparing units which might significantly differ in their internal structure (production technology) even presenting similar input/output levels. From the optimum output-to-input ratio, resource lacks (with respect to projected goals) and slacks can be determined for each decision unit evaluated individually. The case of Brazilian Federal Saving banks is presented as an empirical application of the methodology.