Big Data as Audit Evidence: Utilizing Weather Indicators

• In Event: SET Audit Data and Data Standards

Abstract

Using daily sales revenue accounts from a multi-location retail firm with homogeneous operations in the US, we study whether big data plays an important role in improving audit analytical procedures. Specifically, this study examines two questions: 1) how the summation of disaggregated store level expectations provides more accurate and precise expectations of the account balance than expectations derived from the aggregated firm level data and 2) whether big data contributes to generating more accurate and precise expectations of the account balances than expectations derived from only financial information. Especially, as relevant audit evidence from big data, weather indicators are adopted since previous literature indicates that retail sales are likely to be influenced by daily weather conditions (e.g., precipitation). The preliminary results from this case study show that when auditors derive expectations of the company-wide sales account, the disaggregated models including peer store account observations and weather indicators provide more accurate and precise values than the aggregated model.

Authors

• Kyunghee Yoon, Rutgers Business School

• Alexander Kogan, Rutgers University - Newark