An Empirical Analysis of the Causal Relationship Between Stock Prices and Trading Volume: Evidence from Stock Splits on African Stock Exchanges

This paper investigates the causal relationship between trading volume and stock price around stock splits in African stock markets, which are characterized by low liquidity. The sample comprises 223 split stocks listed on seven exchanges (BRVM, Egypt, Morocco, Mauritius, Nigeria, South Africa, Tunisia) over the 2014–2024 period. We combine linear Granger (1969) causality tests with the nonlinear causality test of Diks & Panchenko (2006), after a BDS test confirming nonlinear dependence. Before the split, volume predicts price: linearly in the BRVM and Tunisia, and nonlinearly in Egypt, Tunisia and South Africa (at the 5% significance level). After the split, the relationship is reconfigured: price becomes predictive of volume, particularly in the BRVM, Tunisia, South Africa and Mauritius, while Morocco displays bidirectional causality (at the 5% significance level). These findings suggest that, once a stock is split, investors become sensitive to price levels, which affects the tradability of the security. Managers should therefore use stock splits carefully and target an optimal price range.