<p>This study employs quantile graphs to analyze financial return series and proposes a structural significance analysis framework with economic interpretations. By defining different states through the quantile parameters of the distribution, graph structures are constructed to enable structural significance analysis for investigating relationships between distinct types of returns. We first analyze synthetic data generated by GARCH models, demonstrating that quantile graphs effectively characterize relationships between extreme tail returns. We validate the methodology on datasets from both stock markets (daily returns) and cryptocurrency markets (minute-level returns), confirming its analytical efficacy. Computational results reveal that specific structural patterns in quantile graphs capture return dynamics and identify state transitions. Through structural significance analysis of quantile graphs, this study identifies distinct dynamics between the Chinese and U.S. markets. Notably, the latent dynamics uncovered in the Bitcoin return series align with their respective price movement patterns. Furthermore, the structural significance exhibits robustness, offering fresh analytical perspectives and holding potential value for financial market analysis.</p>