The Effect of AI-Powered Data Integration Capability on Client Investment Returns of Investment Banks in Kenya

Investment banks in Kenya face persistent difficulty converting growing volumes of market data into superior, risk-adjusted client returns, and it remains unclear whether AI-powered data integration capability offers meaningful leverage for closing this gap. This study examined the effect of AI-powered data integration capability (ADIC) on client investment returns (CIR) among licensed investment banks in Kenya. A correlational survey design was adopted, and a structured questionnaire was administered to a census of 170 portfolio management, risk, technology and senior management officers drawn from all 17 licensed investment banks; 136 usable responses were returned (80% response rate). Data were analysed using descriptive statistics, Pearson correlation and linear regression. ADIC recorded a strong, positive bivariate association with CIR (r = .37, p < .001) and remained a significant predictor of CIR in regression analysis (B = .340, β = .347, t = 4.773, p < .001). The regression model was statistically significant, F(4, 131) = 15.64, p < .001, and explained 32.3% of the variance in CIR. Automated data-quality management was the most developed data-integration practice, while predictive analytics for investment decisions was the least developed, revealing a gap between data infrastructure and its analytical exploitation. The findings, interpreted through Modern Portfolio Theory, suggest that Kenyan investment banks seeking the highest return on AI investment should prioritise data integration capability, particularly predictive analytics.