Measurement Sensitivity of the Determinants of Financial Flexibility Among JSE-Listed Non-Financial Firms

This study examines the sensitivity of empirical findings in financial flexibility research to alternative measurement approaches. Financial flexibility is widely recognised as a critical determinant of corporate financing and investment decisions, yet there is no consensus on its operational definition. Existing studies employ different proxies, in-cluding spare debt capacity, cash holdings, leverage positioning, and composite measures, which may capture distinct dimensions of financial flexibility. Using panel data from 106 non-financial firms listed on the JSE over the period 2000–2019, this study compares two widely used proxies: spare debt capacity derived from predicted leverage models (Marchica and Mura 2010) and a composite low-leverage–high-cash measure (Arslan-Ayaydin, Florackis, and Ozkan 2014). Fixed- and random-effects models, along with logistic regression, are employed to assess whether the determinants of financial flexibility remain consistent across these measures. The findings reveal substantial vari-ation in coefficient signs, statistical significance, and economic interpretation across proxies. Classification-overlap analysis shows limited agreement between the proxies, with only 24.1% of firm-year observations classified as flexible under both measures and Cohen’s kappa indicating only fair agreement. Growth opportunities, retained earnings, dividend payout, and selected firm-specific variables exhibit strong sensitivity to meas-urement choice, while some relationships remain relatively stable. The results demon-strate that financial flexibility is not a measurement-invariant construct and that proxy selection can materially alter empirical and theoretical conclusions. The study contributes to corporate finance literature by highlighting the importance of construct validity, proxy transparency, and robustness testing, particularly in emerging market contexts.