Quantile Regression and Dividend Heterogeneity in Emerging Markets
Dividend policy remains one of corporate finance’s enduring puzzles — not because firms are inherently unpredictable, but because their financial responses to internal and governance drivers vary across performance tiers. Traditional Ordinary Least Squares (OLS) estimation assumes that the conditional mean of dividend payout is shaped uniformly by profitability, leverage, or ownership structure. This “average effect” perspective can conceal rich behavioural differences that exist between low-performing and high-performing firms.
Quantile Regression (QR), introduced by Koenker and Bassett (1978), relaxes this restriction by estimating the impact of explanatory variables across the entire conditional distribution of the dependent variable — for instance, at the 25th, 50th, and 75th percentiles of the dividend ratio. Rather than producing a single slope coefficient, QR provides a spectrum of effects, revealing how financial and governance determinants behave at different points of firm performance.
At SASNG Econometric Services, we apply quantile regression to manufacturing sector data from Nigeria, comparing it with OLS results to highlight this heterogeneity. Our findings consistently show that profitability exerts a stronger positive effect on dividends in upper quantiles, where firms have stable earnings and free cash flow to distribute. Conversely, leverage and ownership concentration dominate at the lower quantiles, where liquidity constraints and agency costs are more pronounced. OLS would average these distinct effects into one misleading coefficient, masking the nuanced behaviour of firms under different financial pressures.
This differential insight is critical for investors, regulators, and corporate boards. Investors can identify which firms are most likely to maintain consistent payouts under varying profitability levels; regulators can design governance frameworks sensitive to firm heterogeneity; and boards can align payout strategies with firm capacity, signalling strength to the market while managing leverage prudently.
Our consulting team at SASNG provides end-to-end econometric support — from model specification and quantile decomposition to bootstrapped inference and graphical quantile process visualisations. We also benchmark results against OLS outcomes to ensure interpretative clarity and robustness. In every engagement, our objective is to transform statistical complexity into evidence-based financial strategy, helping both academics and decision-makers move beyond averages toward truly differentiated insight.
Keywords: quantile regression, dividend policy, firm value econometrics, financial consulting.
