Dynamic Combination of Automatic Forecasts for Corporate Budgeting

In this paper, an optimal dynamic combination model of automatic forecasts is proposed for corporate budgeting purposes. During the budgeting process, financial managers usually have to create forecasts for various economic quantities such as sales, the cost of goods sold, net earnings, etc. This situation is becoming exponentially more complex if we take into account the various production lines and business segments of the firms. In such an environment, it is very important for a forecasting model to have the ability to produce fast, reliable, accurate, and cost-effective forecasts on a timely basis. The purpose of this paper is to provide a forecasting methodology for budgeting purposes that incorporates all the desirable properties that financial managers expect of a forecasting model that supports a modern budgeting system. Our forecasting methodology utilizes various automatic forecasting models for univariate time series that are combined to produce a single optimal forecast for each financial quantity under consideration.