Abstract. This study employs Rescaled Range (R/S) analysis to investigate the non-linear dynamics and long-range dependence of returns in the Romanian capital market, specifically analyzing the benchmark BET index from 2000 to 2024. By calculating the Hurst exponent over Fibonacci-based subseries, the research examines the extent of persistence and non-random behavior within the index. Additionally, the Lyapunov exponent was estimated to assess the sensitivity of the BET index to initial conditions, providing insights into potential chaotic dynamics. The results reveal significant long-term memory, persistent trends, and marginally chaotic behavior, indicated by a positive yet near-zero Lyapunov exponent. These findings challenge the assumptions of weak-form market efficiency and align with earlier studies highlighting structural and behavioral inefficiencies in emerging financial markets. The implications are significant for investors, suggesting that trend-following strategies might offer superior predictive capacity compared to traditional random-walk models. For policymakers and regulators, the evidence underscores the need for enhanced market transparency and institutional robustness to mitigate structural inefficiencies. Overall, this study contributes to the growing literature that applies complexity science to financial markets, emphasizing the necessity of integrating nonlinear analytical frameworks for a deeper understanding of market dynamics in emerging economies.

Keywords: #rescaledrange; #hurstexponent; #nonlinearity; #stockmarket; #chaostheory.