Time Series Models for Short-Term Forecasting Performance Indicators

Inhaltsverzeichnis (Table of Contents)

• Introduction
• Outline of the Historical Background of Forecasting
• Motivation
• Methodology
• Review of Literature
• Description of Data
• Analysis
• Building a Model for Forecasting Cost of Goods Sold
• Building a Model for Forecasting Net Sales
• Computation
• Assumptions of the Classical Linear Regression Model
• Validation of Assumptions
• Assumption 1
• Assumption 2
• Assumption 3
• Assumption 4
• Assumption 5
• Assumption 6
• Assumption 7
• Weighted Least Squares Regression
• Time Series Analysis
• The Autoregressive Process
• The Moving Average Process
• The Autoregressive Moving Average Process
• The Autoregressive Integrated Moving Average Model
• Model Identification
• Model Estimation
• Diagnosis
• Conclusion
• Forecasting
• Outlook

Zielsetzung und Themenschwerpunkte (Objectives and Key Themes)

This thesis aims to develop and validate statistical models for short-term forecasting of key performance indicators, namely net sales, Cost of Goods Sold (COGS), and net contribution, using existing company data from Dalian Chemson Chemical Products Co; Ltd. (DCCP).

• The impact of raw material prices, particularly lead prices, on the COGS.
• The effect of sales price adjustments on net sales.
• The application of classical linear regression and time series analysis techniques to forecast financial performance.
• The use of simplified models that are practical and easy to implement in a business setting.
• The importance of validating model assumptions and assessing model adequacy for reliable forecasting.

Zusammenfassung der Kapitel (Chapter Summaries)

• **Introduction:** This chapter provides a brief history of forecasting and its evolution, highlighting the significance of probability theory and modern statistical methods. It introduces the motivation behind the thesis, which is to address DCCP's need for accurate short-term forecasts of key financial indicators. The methodology employed in the thesis, involving various time series models for forecasting net sales, COGS, and net contribution, is outlined.
• **Review of Literature:** This chapter discusses the extensive literature on regression and time series analysis. It emphasizes the lack of similar published studies on predicting net sales, COGS, and net contribution for DCCP specifically, as well as the abundant resources available for understanding these statistical techniques.
• **Description of Data:** This chapter provides a detailed description of the data used in the analysis, including a list of variables, their definitions, and their characteristics. The data consists of 6 quantitative variables observed over 40 consecutive months, from January 2004 to April 2007. The author also calculated several additional series from the existing data using the SPSS Data Editor, such as average cost per unit sold, average net sales per unit, and lagged variables.
• **Analysis:** This chapter delves into the construction of two models, one for forecasting net sales and the other for predicting the Cost of Goods Sold. The models are designed to predict the average value per unit sold, which is then multiplied by the expected quantity sold to obtain the overall forecast. The chapter then explores the assumptions underlying the Classical Linear Regression Model (CLR) and examines the validity of these assumptions in the context of the two models.
• **Conclusion:** This chapter summarizes the key findings of the thesis, emphasizing the performance of the OLS version of model 1 and the ARIMA version of model 2 in forecasting net sales, COGS, and net contribution. The chapter highlights the practical applicability of these models and emphasizes the importance of balancing model complexity with forecasting accuracy and efficiency. The chapter also discusses potential areas for future research, including the exploration of more sophisticated variance-stabilizing transformations and dynamic models.

Schlüsselwörter (Keywords)

This thesis focuses on forecasting key performance indicators using time series models. Key concepts include linear regression, autoregressive integrated moving average (ARIMA) models, model identification, model estimation, and model validation. Other important themes are the impact of raw material prices, particularly lead prices, on the Cost of Goods Sold (COGS), and the effect of sales price adjustments on net sales. The study utilizes statistical software such as SPSS for data analysis and model building.