This study examines how scholarly influence is generated and sustained in academic finance by analysing lifetime top-cited articles. Using citation data from ScholarGPS combined with systematic thematic and theoretical classification, it maps the intellectual structure of finance and traces how influence evolves across research generations and major structural breaks. Despite substantial diversification in research topics and increasing empirical sophistication, long-run influence remains tightly anchored to a small set of early-established theoretical frameworks, particularly equilibrium asset pricing and agency-based models. Rather than being displaced, these foundations persist as the core of the finance canon. Influence evolves primarily through path-dependent recombination: newer topics and methods achieve durable impact mainly when integrated into established theoretical cores rather than through paradigm replacement. Multivariate analyses show that field-defining articles receive significantly higher long-run citation impact than otherwise comparable influential articles, independent of journal prestige, authorship, or study design, although this advantage weakens markedly after 2000. Major technological, regulatory, and crisis events reweight the focus of newly influential research but do not reset the canon, with post-crisis influence concentrated in governance, behavior, and frictions through extensions of existing paradigms. Overall, the findings show that finance adapts rapidly in empirical scope and method, yet far more slowly at the level of core theory, explaining the persistence of canonical ideas amid changing research agendas.