This study compares the performance of banks that are part of a financial holding company (FHC banks) with that of banks that are not (independent banks) using Taiwan data from 2002:Q1 to 2006:Q2. The comparisons are based on 14 performance ratios resulting from the concept of CAMEL (which is an acronym for Capital adequacy, Asset quality, Management efficiency, Earnings ability, and Liquidity sufficiency). To ensure that becoming part of an FHC is a random process, we used four matching methods to select the controlled banks so that the characteristic variables of banks in the two groups are statistically indifferent. By using the data before matching, it was found that FHC banks significantly defeat independent banks, regardless of their performance ratios. Conversely, when the sample was used after the matching, the results changed dramatically. Although FHC banks still beat the independent banks in terms of capital adequacy, asset quality, and liquidity sufficiency, FHC banks and independent banks are found to have equal profitability and management efficiency. Earlier studies that do not consider the endogeneity problem tend to overestimate the joining effect.