Cross-sectional governance predictability: what the literature says
A curated review of peer-reviewed studies and working papers examining whether boardroom
structures forecast returns, multiples, and risk metrics.
Academic research on governance predictability has accelerated over the past two decades,
yet the conclusions remain contested. We summarize the most rigorous studies, note their
limitations, and explain how our own framework builds upon and departs from prior work.
Gompers, Ishii, and Metrick (2003) established an early empirical link between
governance provisions and firm valuation using a broad cross-section of United States
firms. Their governance index demonstrated that firms with stronger shareholder rights
traded at higher multiples, though subsequent replication attempts produced mixed
results depending on sample period and control specification.
Bebchuk, Cohen, and Ferrell (2009) refined this approach with an entrenchment index
focused on six specific provisions, finding more robust cross-sectional predictive power
for operating performance and valuation. Their work influenced our own component
selection by emphasizing that not all governance variables carry equal informational
content.
International extensions by Durnev and Kim (2005) and Klapper and Love (2004)
demonstrated that governance effects vary significantly across legal regimes and
disclosure environments. These studies inform our jurisdiction-specific calibration and
our skepticism toward one-size-fits-all scoring approaches.
More recent working papers using machine learning methods have identified non-linear
governance interactions that traditional linear models miss. We monitor this literature
closely but remain cautious about overfitting claims, requiring any new technique to
demonstrate out-of-sample stability across multiple market cycles before incorporation
into our published scores.
Our own contribution lies in combining cross-sectional regression frameworks with
transparent weight calibration and continuous forward testing. We publish prediction
errors quarterly and invite external replication, believing that methodological honesty
strengthens rather than weakens the practical value of governance research.
Our governance framework has matured through deliberate iteration, failed experiments, and honest reassessment. These milestones mark the key stages of that journey.