Sales forecasting in financial distribution: a comparison of quantitative forecasting methods
- 1 Downloads
This paper deals with the issue of forecastability of sales activities of independent financial advisers (agents). Employing the most common quantitative methods on a diverse sample of timelines from multiple advisory companies, we have found that under most settings, these methods offer sub-par performance with high relative errors and no statistical differences between them. When a more granular approach is applied (reflecting sales unit size), ARIMA and the simple moving average emerge as significantly less accurate. This outcome is true for all sales units regardless of their size, when relative error is concerned. Thus, our analysis confirms the difficult forecastability of financial sales, speaking against the utilisation of more sophisticated forecasting methods, which mostly fail when compared to their much simpler and less costly counterparts.
KeywordsFinancial sales Sales forecasting Quantitative forecasting Mean absolute percentage error (MAPE) Linear model with mixed effects
This work was supported by the Czech Science Foundation under Grant No. 16-21506S: New sources of systemic risk on financial markets (2016–2018).
- Ali, M.M., and J.E. Boylan. 2010. The value of forecast information sharing in supply chains. Foresight: The International Journal of Applied Forecasting 18: 14–18.Google Scholar
- Catt, P.M. 2009. Forecastability: Insights from physics, graphical decomposition, and information theory. Foresight 13 (Spring): 24–33.Google Scholar
- Czech National Bank—CNB. 2018a. Inflation statistics. Accessed 6 September 2018. https://www.cnb.cz/en/statistics/inflation/index.html.
- Czech National Bank—CNB. 2018b. Repo interest rate over time. Accessed 6 September 2018. https://www.cnb.cz/cs/faq/jak_se_vyvijela_dvoutydenni_repo_sazba_cnb.html.
- Czech Statistical Office—CZSO. 2018a. Employment, unemployment statistics. Accessed 6 September 2018. https://www.czso.cz/csu/czso/employment_unemployment_ekon.
- Czech Statistical Office—CZSO. 2018b. Wages and labour statistics. Accessed 6 September 2018. https://www.czso.cz/csu/czso/labour_and_earnings_ekon.
- Gilliland, M. 2010. The Business Forecasting Deal: Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions. New York, USA: Wiley.Google Scholar
- Goodwin, P. 2011. High on complexity, low on evidence: Are advanced forecasting methods always as good as they seem?. Foresight: The International Journal of Applied Forecasting 23: 10–12.Google Scholar
- Gordon, A. 2010. The boundaries of quantitative forecasting methods: Respecting the limits of determinism. Foresight: The International Journal of Applied Forecasting 19: 9–15.Google Scholar
- Hyndman, R.J., and G. Athanasopoulos. 2018. Forecasting: Principles and practice. Published by OTexts.com, full text available at: https://www.otexts.org/fpp.
- Hyndman, R.J., and A.V. Kostenko. 2007. Minimum sample size requirements for seasonal forecasting models. Foresight 6 (Spring): 12–15.Google Scholar
- Madhani, P.M. 2016. Sales and marketing integration: Enhancing competitive advantages. The IUP Journal of Business Strategy 13 (4): 50–77.Google Scholar
- Ministry of Finance of the Czech Republic—MF. 2018. Macroeconomic forecast. Accessed 6 September 2018. http://www.mfcr.cz/cs/verejny-sektor/makroekonomika/makroekonomicka-predikce.
- Peterson, R.M., G.L. Gordon, and V. Krishnan. 2015. When sales and marketing align: Impact on performance. Journal of Selling 1 (15): 29–43.Google Scholar
- Reifner, U., Neuberger, D., Rissi, R., Riefa, Ch., Knobloch, M., Clerc-Renaud, S. and Finger Ch. 2012. Study on remuneration structures of financial services intermediaries and conflicts of interest. [Working Paper, No. MARKT/2012/026/H]. Hamburg, Germany: IFF - Institut für Finanzdienstleistungen.Google Scholar