For estimating the efficiency of the banks operating in an economy it is generally assumed that these banks lie under a common frontier but instead the banks may lie under different frontiers. The present study tests this assumption of the existence of a common frontier for commercial banks by employing the recently developed stochastic-meta-frontier econometric model. The meta frontier methodology allows us to compare the efficiencies of firms that lie under different groups such that these groups may have access to different technologies. Our dataset comprises of 27 Public Banks, 22 Private Banks and, 38 Foreign Banks operating in India and spans from 2005-2017, forming a total of 995 bank-year observations. The study estimates the cost and profit efficiencies of these commercial bank groups using a trans-log cost function and a trans-log profit function respectively. The study significantly establishes that the efficiency estimates obtained using a common frontier are generally overestimated. Further to this, the study contributes to literature that (a) instead of one common frontier for all the banks together there exist different frontiers for each ownership group of banks. (b) public banks’ group is the most efficient while foreign banks’ group is the least efficient under both cost and profit efficiency calculations. Interestingly, the observations from the cost gap-ratio and profit gap-ratio show that public banks are most innovative in reducing cost while foreign banks are most innovative in improving profits among the three ownership groups. These findings can be used by the policy makers to frame ownership- oriented policies that may help in strengthening the overall banking industry.