On a holistic view of supply chain financial performance and strategic position

Abstract

Abstract Measuring corporate financial performance is an essential task in many supply chain decisions, such as supply chain strategic positioning and partner selection. This study introduces an analytical approach that can quickly scan financial data of many companies and produce a summary measure for each company. The approach offers organizations a less wearing way to obtain a holistic view of all target companies’ financial performance patterns, which imply the underlying supply chain strategies. The strategy map, a two-dimensional representation of the summary, provides a comprehensible means to apprehend the relative strategic position and measure the similarity between companies. The approach relies on three popular machine learning models, forecasting, clustering, and classification. It takes multi-year, multi-variate financial time series from the three standard financial statements, learns the patterns from the data, and tunes model parameters to configure the final settings for future applications. The input data needed are relatively easy to obtain and the self-learning modules only require modest domain knowledge to apply the approach. Its noise reduction, outlier detection, and feature selection functions ensure a consistent and robust performance. The empirical test using data from all US manufacturers and traders listed on NYSE and NASDAQ demonstrates the efficacy of the approach.