Pricing Risk in Thin African Equity Markets: A Systematic Review of Models, Measurement and Nigerian Evidence

Abstract

We examine African listed-equity pricing models in settings where returns, benchmarks and investable universes are subject to consequential measurement error. We systematically review the models tested, assess how thin-market measurement choices shape findings and derive requirements for a Nigeria-adapted research design. Searches identified 2,249 records; after deduplication and screening, we included 154 reports representing 143 study families and 163 analyses, of which 54 contain Nigerian evidence. CAPM and conditional variants account for 66 analyses and characteristic factor models for 39, with preferred specifications varying across markets, periods and samples. Measurement weaknesses are pervasive: only 20 analyses clearly use total returns; 138 score weakly on non-trading or stale-price treatment; 113 score weakly on stability and validation; and one reports a formal out-of-sample test. Thirteen analyses meet a pre-specified measurement-and-inference sensitivity floor, including six with Nigerian evidence. Model rankings vary across markets, periods and samples. Benchmark choice, trading continuity, survivorship, factor construction and investability materially alter reported performance. We therefore, propose a review-derived, nested research architecture for Nigeria: construct survivorship-aware total returns; estimate a corrected local benchmark; add mechanism-based characteristic, liquidity and external-risk candidates; and require pricing-error, stability, investability and holdout tests. Our review treats market-data construction as part of identification and sets minimum standards for future African asset-pricing studies.