https://www.mdu.se/

In this paper, we aim to compare Lee Carter method versus Hyndman and Ullah method for forecasting mortality rates by applying them to Swedish and West German data respectively. We find that the Hyndman and Ullah method is always more robust than its counterpart in terms of variation explanation, outlier identification and forecast accuracy. We also describe these two methods in details and strictly following the steps from the papers [16, 15]. Although this paper has a few weaknesses such as data limitation, naive assumption on data selection and disregarding the Hyndman and Ullah method extensions, our replication study contributes to confirmation that the Hyndman and Ullah method is more robust in forecasting mortality rates.