This study
investigates the factors influencing return on equity (ROE) for the five
largest U. S. companies by market capitalization, namely NVIDIA, Apple,
Alphabet, Microsoft, and Amazon, using an extended DuPont analysis framework.
The study uses 10 years of firm- level financial information from audited Form
10- K filings to decompose ROE into its components: profitability (profit
margin), operational efficiency (asset turnover), and financial leverage
(equity multiplier). Rather than relying exclusively on the classic
multiplicative method, the research employs a multiple regression model that
accounts for both main and interaction effects among these elements. Diagnostic
assessments confirm the model' s validity, and heteroskedasticity- consistent
(HC 3) standard errors are used to ensure robust inference. The regression
findings reveal that asset turnover is a significant positive predictor of ROE
(B = 0. 66, p <. .001), whereas profit margin (B =- 1. 18, p <. .001) and
the equity multiplier (B =- 0. 14, p <. .001) have negative main effects.
Notably, the interaction between profit margin and the equity multiplier shows
a positive and highly significant relationship (B = 1. 56, p <. .001),
indicating that financial leverage enhances the impact of profitability on
shareholder returns. The model accounts for approximately 97. 8% of the
variability in ROE (R ² =. .978). The important interaction between profit
margin (PM) and equity multiplier (EM) illustrates that profitability' s effect
on the dependent variable is contingent on financial leverage. When leverage is
low, the profit margin exerts a minimal effect on ROE. As leverage rises, the
marginal effect of profitability shifts in a positive direction. From the
estimated coefficients, the marginal effect of the profit margin is- 1. 184 +
1. 557 ·EM. This indicates that when leverage is lower, increases in profit
margin correlate with declines in the dependent variable. As the equity
multiplier (EM) increases, the negative influence of profit margin gradually lessens.
Ultimately, the effect turns positive once EM surpasses roughly 0. 76. This
research demonstrates that DuPont analysis remains a valuable and effective
tool in modern financial analysis. It enhances the clarity of performance
evaluations and facilitates more informed managerial and investment choices by
pinpointing the true determinants of shareholder returns.
