Predicting future economic trends appropriately is essential to economic policy making. Currently, the DSGE model approach is a benchmark economic forecasting technique widely employed. However, large external shocks, such as large-scale natural disasters and COVID-19, challenge current approaches to economic forecasting. Multiple approaches will be needed in this situation, including reduced-form model and indicator-based approaches. This paper discusses different forecasting approaches, by comparing forecasts during normal times and crisis periods. The Medium-term Projection Framework (MPF), used in the Economic Outlook for Southeast Asia, China and India series, receives particular attention. The paper also examines challenges unique to developing Asia and large external shock periods. The measurement of potential output, difficulties in modelling the credit channel, and the incorporation of Big Data pose challenges regarding developing Asian countries, and large external shocks may force deviation from assumptions of traditional frameworks such as rational expectations. Finally, this paper points out that natural disasters will be a useful proxy for large shocks in Developing Asia.