Test-Driven Development for Spreadsheet Risk Management
Kevin McDaid, Alan Rust · IEEE Software · 2009
Spreadsheet technology is central to the functioning of the financial sector, but the spreadsheets themselves can have a high level of error that requires innovative supporting processes and tools. Several large-scale international studies conducted in response to the Enron-inspired Sarbanes-Oxley Act confirm the importance and complexity of spreadsheet systems in financial reporting, with some companies using 200+ spreadsheets in manually controlled interconnected "webs." Unfortunately, spreadsheets frequently have an unacceptably high number of faults. Early research indicated that between 2 and 5 percent of cell formulas can be incorrect, but a more recent study found that of 50 real-world operational spreadsheets audited, 94 percent contained errors with almost 1 percent of formula cells found to be incorrect. Spreadsheet developers don't appreciate this high risk and thus use little or no formal software development processes.