Holt's method, although successful, has limitations like its limited use for linear trends, seasonal fluctuations, and tendency to overestimate. The Damped Trend Method (DTM) was introduced to address these shortcomings, but still struggles with optimal forecast accuracy. Hence, this study proposes a new hybrid model by integrating the classical Holt's method and the Moving Average Box-Jenkins Methodology known as Holt Integrated Moving Average (HIMA). To evaluate the performance of the new hybrid model, Malaysia Consumer Price Index data from the year 1960 to 2022 was used. The dataset is partitioned using Repeated Time-Series Cross-Validation (RTS-CV). This study has also shown that the HIMA model has improved the forecasting of classical Holt's method and the Damped Trend Method (DTM) for time-series data, which has level and trend components. The results of this study show that HIMA has produced the lowest error measure value (1.5703, 1.1811, and 1.0013 for RMSE, MAE, and MAPE, respectively) and a high percentage of forecast accuracy (99 per cent) compared to other comparative models for CPI forecasting.