Modeling and forecasting Saudi banking stability using ARIMA and exponential smoothing technique

This research examines the key factors influencing the financial stability of Saudi banks by developing an optimal stepwise linear regression model. The research uses financial information gathered from 11 Saudi banks over the period 2014-2021. Six categories for key performance indicators (KPIs) which consist of profitability, liquidity, asset quality, capitalization, bank size and economic growth are included in the model. The <i>Z</i>-score is used as its dependent variable for all stability measures. A model with the lowest standard error should be selected as the best explanatory model among all options while also maintaining the highest adjusted R-squared value. The findings showed that the chosen model has the lowest standard error around (7.209) and the highest adjusted R-squared (71.3%), The study demonstrates that NII1 ratio and CAR statistics alongside bank asset size (log of assets) produce positive effects on stability yet the stability declines when banks use investment ratio statistics or loan impairment ratio indicators. Economic growth (GDP) shows no significant influence. The second phase of this research uses ARIMA and exponential smoothing models which are selected to produce <i>Z</i>-score predictions through 2030. The chosen forecast validation metrics include RMSE, MAE, MAPE and E-square. The standardized forecasts enable banks to compare resulting data with each other. The financial performance data shows different trends. Studies indicate that Arab National Bank and National Commercial Bank will provide consistent financial outcomes. Saudi Investment Bank and Bank Al - Jazira have moderate trends with high forecast precision. Al Rajhi Bank, Samba Financial Group and Saudi British Bank continue to operate steadily. The empirical findings offer support to stakeholders and regulatory authorities in decision-making processes that enable alignment with the Vision 2030 objectives.