Reevaluating Industry Effects in Firm Performance
New machine learning model reveals how industry classification systems transform performance analysis.
1The Industry Effect Debate
While strategy research examines key determinants of firm performance, the industry effect has long been debated. Traditional views argued for minimal industry impact. This study challenges the classification methodology underlying such conclusions.
2Comparing Classification Systems
Researchers compared existing industry classification systems (like SIC) with a new ML model analyzing annual report texts. Findings show significant variation in how systems group firms, with some companies consistently clustered together across methods.
3Grouping Similar Firms
The study found systems that consistently group similar firms together demonstrate stronger industry effects. This suggests traditional classifications systematically underestimate industry's true impact on performance.
4ML Model's Contribution
The new model processes annual report texts to make more accurate industry assignments. This approach creates groupings that better reflect firms' true sector similarities, significantly altering analysis outcomes.
5Managerial Implications
Findings indicate understanding industry dynamics is more critical than previously thought for executives. Precise industry classification could prove vital in developing long-term success strategies.
6Broad Research Impact
This study calls for reevaluating fundamental methodologies in firm performance analysis. More accurate measurement of industry effects could profoundly influence both academic research and corporate strategy decisions.