External indicators in demand forecasting - Case passenger car tyres

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
School of Business | Master's thesis
Ask about the availability of the thesis by sending email to the Aalto University Learning Centre oppimiskeskus@aalto.fi
Degree programme
Information and Service Management (ISM)
81 + 13
In the global business environment, adjusting to the constant changes in demand that take place in multiple markets is a common problem from the manufacturing companies. The uncertain conditions are handled by increasing the sizes of safety stocks, which cause unnecessary costs and tie up capital assets that could used more effectively elsewhere. Decreasing the size of the forecasting error should be deemed important by the businesses as it is likely to reduce these costs. Literature shows that judgmental forecasting methods are too often preferred over quantitative ones, even though they suffer from various cognitive biases and often lead to inferior results. This study investigated the effectiveness of different external indicators in improving the demand forecasts. Several publicly available economic indicators and confidence indicators were studied in addition to a couple weather-based indicators. The study is industry-specific, as it used the passenger car replacement tyre sales in several European countries between the years 2008–2016 as a source dataset. The study revealed that time series models that included multiple external indicators as external factors can outperform the best available models that are only based on the historical demand data. The benefits were greater with short six-month forecasts that longer forecasts that went one or two years into the future. Results also showed that coincident indicators with lag times close to zero were most likely to improve the predictive performance, while the intuitively useful leading indicators performed much worse. It was found that there were major differences between the different studied markets, concerning which indicators performed the best and how much they were able to improve the forecast quality. The indicators and methods that work best in most of the countries might not always be the best fit for a certain market. This study improves the understanding of the ways that the external economic indicators and confidence indicators can be used to predict future changes in demand levels. To inspect whether the findings of the study are universally applicable for manufacturing companies, similar studies should be executed with different industries and different time periods.
Thesis advisor
Malo, Pekka
Kuosmanen, Timo
external indicators, economic indicators, demand forecasting, quantitative forecasting, seasonal ARIMA, tyre industry
Other note