EXCHANGE RATE MOVEMENTS AND CPI CATEGORY PRICES IN RWANDA: PASS-THROUGH MAGNITUDE, TIMING AND ADJUSTMENT DYNAMICS
Exchange rate pass-through is conventionally studied through aggregate inflation, yet its magnitude and timing may substantially differ across CPI categories. This study examines how exchange rate movements are transmitted across Rwanda’s CPI categories and whether their relative exposure changes over time. Using quarterly data and category-specific autoregressive distributed lag error correction models, the analysis distinguishes both long-run pass-through from short-run responses and horizon-specific adjustment dynamics. The long-run estimates reveal positive but incomplete pass-through, ranging from 0.08 for clothing and footwear to 0.28 for food and non-alcoholic beverages and 0.54 for housing, water, electricity, gas and other fuels, with supplementary transport estimates of 0.32. HWEGO-CF difference survives a conservative worst-case covariance screen at the unadjusted 5% level, but not a three-comparison Bonferroni adjustment, so familywise-error-controlled evidence of a cross-category difference is not claimed. The dynamic paths are nonetheless sharply heterogeneous in their point estimates: HWEGO has the largest contemporaneous response and builds gradually, FNAB responds rapidly but oscillates, and CF remains small and settles quickly. The point estimates imply a horizon-dependent change in ranking, with FNAB above HWEGO in the first three quarters and HWEGO above FNAB thereafter. The findings establish the importance of examining pass-through as a horizon-dependent, category-specific process rather than a single aggregate coefficient. By distinguishing the extent, speed, and temporal ordering of price adjustment, the study extends African pass-through evidence and offers a more differentiated basis for inflation monitoring and forecasting.
Read Article