Our study aims to fill the gap in estimating the impacts of political connections and bank funding diversity on the risk-taking behaviors of Vietnamese commercial banks. By employing the Bayesian methodology, our paper can overcome the small sample issues to reduce the bias in estimation results. We construct a data sample that includes 38 commercial banks in Vietnam from 2003 to 2020. Our results suggest several findings in the Vietnamese banking sector. Firstly, our findings suggest that politically connected banks have 0.4% non-performing loans higher than unconnected peers. Secondly, we find a positive relationship between funding diversity and non-performing loans of commercial banks in Vietnam. Interestingly, our findings report that the commercial banks, especially the politically connected banks, reduced non-performing loans by 0.2% and 0.4% for a year before the recent two National Congress of the Communist Party of Vietnam, respectively. It could be due to the notion that the bank managers secure their job and political promotions by reducing non-performing loans before the National Congress of the Communist Party of Vietnam. Finally, our study argues that the commercial bank had a lower level of non-performing loans during the Covid19 pandemic because the government offered stimulus supports to the local economy. Our study has substantial implications for bank managers and authorities in emerging markets.