Abstract:
Non-Performing Loans (NPLs) are a problem in banks as they lead to stagnation and bank
failure. NPLs are more serious in Community Banks (CBs) as they do not have adequate
resources to track down defaulters. Using explanatory sequential research design, this study
examined drivers of NPLs in CBs. It applied unbalanced panel data gathered from Bank of
Tanzania (BOT) and CBs' audited accounts. It also made use of key informant interviews to
validate some findings from panel data analysis. Data were analysed descriptively and
inferentially using Stata, while a random effect panel regression model was carried out to
analyse drivers of NPLs. The findings establish that NPLs ratios had an upward swing, with a
sharp increase in 2007-09 associated with effects of the global financial crisis. Overall, the
paper establishes a strong relationship between NPLs ratio, bank level and GDP factors.
Statistically the relationship was significantly positive with net interest margin but negative
with capitalization and loan to assets ratios. The association with a co-operative banking factor
was also positive, defying the joint liability contract theory but consistent with agency theory.
The relationship of NPLs with GDP was positive, thus contradicting the paradigm that a
healthy economy is associated with lower NPLs. Generally, the CBs in Tanzania did not
practice due diligence in financing a hotly growing economy. It is recommended that CBs
should revisit their pricing policy to control the incidence of NPLs. Moreover, regulators need
to enforce capital adequacy regulation in order to control excessive risk-taking behaviour in
thinly capitalized banks. Further, BOT should enforce prudential lending practices to curb
NPLs in periods of high lending tempo. Further, credit bureau facility should be availed to
smaller banks to track down multiple borrowers.
