The G-Score did not originate in a literature review. It originated in filings — specifically, in the filings that a minority shareholder receives only through legal action, across years of direct practice in Korean shareholder litigation.
A consistent pattern emerged from that practice: the failure modes that ended in real value destruction rarely looked like what generic governance scorecards measured. They looked like three separate, independent patterns — not one composite “governance quality.” A company could disclose everything honestly and still be controlled by a single family. It could have a pristine independent board and still funnel value through opaque subsidiaries. Each was a different kind of failure; each required its own detection. That field observation is the first half of the framework.
The second half is academic. The decomposition into transparency, balance of power, and conflict-of-interest risk has clear antecedents in three bodies of research that had, until the G-Score, been kept separate.
T
Law & finance tradition. The disclosure-regime and legal-enforcement literature of the late 1990s and 2000s established that the same accounting number carries different informational weight depending on the jurisdiction that produced it. The T-axis inherits this insight and applies it at the filing level rather than the country level — reading disclosure behavior through its anomalies, not through its statutes.
B
Agency theory. The structural tension between insiders and outsiders is the classical object of corporate-governance research. Its U.S. form emphasized the manager-vs-shareholder relationship. Asian capital markets, where ownership is concentrated rather than diffuse, require the controlling-vs-minority frame. The B-axis carries the agency intuition into that ownership reality.
R
Tunneling research. The body of work on value extraction through affiliated structures — related-party flows, intercompany transfers, pyramidal ownership, pledge-based financing — informs the R-axis directly. The R-axis reads the structural precursors of tunneling rather than the tunneling events themselves. Detection runs ahead of discovery.
But the assembly is different. Traditional governance indices ask compliance-scale questions and aggregate them into a single number. The G-Score reads the failure precursors directly from the regulatory filings that would otherwise require litigation to surface — and it decomposes rather than aggregates, so that the component which will eventually fail is not averaged into the components that will not.