A factor-based framework for stress-testing the Namibian banking sector

Abstract

Times of crises underscores the importance of guarding against deteriorations in the quality of loan portfolio through effective credit risk management. The purpose of the study is to examine the credit risk resilience of Namibia’s banking sector and forecast the quality of its loan portfolio. Methodologically, the study is hinged on the theories related to information asymmetry, moral hazard, and adverse selection. The methods include a VAR and an ARIMA out of sample dynamic forecasting model. The study employs secondary time-series data for the period 1996Q1–2021Q4 from various sources including the Bank of Namibia, the Namibia Statistics Agency, the World Bank and some others. The stress-testing results analysed via the VAR’s impulse responses show that Namibia’s banking sector is highly susceptible to various shocks with the early warnings emanating primarily from the non-performing loan itself, followed by the monetary, institutional, bank-specific, and interest rate indicators. The forecast for 2023Q4–2025Q4 obtained from the ARIMA model reveals that the riskiness of its loan portfolio is predicted to persist beyond the benchmark of 4 % set by the Bank of Namibia. The findings highlight important policy interventions, including the need to strengthen the mechanisms for monitoring the share of non-performing loans, re-evaluate existing policies, continue to ensure a sound macroeconomic and financial environment, and require banks to maintain a minimum capital adequacy ratio.

Keywords

• banking sector
• loan portfolio
• non-performing loan
• stress-testing
• forecasting
• namibia
• arima model
• var model