<![CDATA[This study investigates the determinants of very low leverage (VLL) among publicly listed non-financial firms in Malaysia, Indonesia and Singapore. The study evaluates whether family ownership and firm-level financial fundamentals shape firms' decisions to maintain extremely low debt levels. An unbalanced panel of 10,160 firm-year observations from 2015 to 2024 is analysed using the two-step System GMM estimator to address endogeneity, dynamic persistence and unobserved heterogeneity. Key explanatory variables include family ownership, profitability, liquidity and operating cash flow, with alternative leverage thresholds used for validation. Results show that capital structure persistence is the strongest predictor of VLL, with firms maintaining a 55 to 57% likelihood of staying in VLL positions across periods. Family ownership does not significantly influence VLL behaviour, challenging agency- and socioemotional wealth-based expectations. Financial fundamentals only matter in Malaysia; their effects disappear in Indonesia and Singapore once dynamic endogeneity is controlled. The findings reveal that several relationships identified in static models are artefacts of endogeneity bias. The strong persistence of VLL suggests that initial financing decisions have long-term effects, highlighting the need for continuous capital structure reassessment. Policymakers should consider institutional differences: Malaysia's relationship-based banking environment reinforces reliance on internal liquidity, while Indonesia requires stronger market infrastructure and creditor protections. Investors should interpret low leverage cautiously, as it may reflect historical path dependence rather than current firm performance. This study provides one of the first comparative dynamic-panel analyses of VLL behaviour in ASEAN markets using System GMM. It advances the capital structure literature by demonstrating that persistence dominates firm fundamentals and that family ownership does not determine extreme leverage choices. The study also clarifies methodological distortions found in static capital structure research]]>
