Does monetary policy pass through differ across the liquidity regimes in Nepal?

Abstract

This study aims to answer whether monetary policy pass-through varies in the liquidity regime in Nepal. Using monthly data and a layered empirical strategy (dynamic pass-through modelling, Markov-switching, threshold/STAR tests, rolling validation, local projections, and a 20 A class bank’s base-rate panel), this study shows that the pass-through is present (but incomplete), persistent, and has state dependence on the banking sector. Baseline estimates suggest a high pass-through with persistence in the retail rates (0.69 for deposits and 0.49 for lending). The regime-based results show a stronger pass through in the states of tight liquidity. The post-IRC period improves the base-rate anchoring. The 12-month base-rate response rises from 0.17 before IRC to 0.56 after. In addition, the regime persistence is also backed up by predictive validation. It shows the inclusion of lagged state information boosts model performance (AUC: 0.324 to 0.904). The results imply that there’s an incomplete transition of money-market signals into retail-rate and credit adjustment, rather than the absence of the rate effects in Nepal. Overall, the regime-specific nature of the monetary transmission in Nepal is not captured by a single average pass-through coefficient.