We use returns across 13 African stock markets to perform factor-spanning tests. Contradicting US-based results, HML is not redundant in the five-factor model, UMD generates a reliable alpha in the Q-based regression, and the mispricing-purged SMB does not consistently dominate the other SMB factors. Because they cannot explain HML and UMD, the Q and Q5 models do not subsume the five- and six-factor models. Since none of the other models captures the ROE factor, the Q-factor model is never subsumed. The Q model spans the Q5 model because the expected growth factor r is not priced in African markets.

中文翻译：

---

哪些因素可以解释非洲股票的回报？


我们使用 13 个非洲股票市场的回报来进行跨因子测试。与基于美国的结果相矛盾，HML 在五因素模型中并不冗余，UMD 在基于 Q 的回归中生成可靠的 alpha，并且错误定价清除的 SMB 并不始终主导其他 SMB 因素。由于 Q 和 Q5 模型无法解释 HML 和 UMD，因此它们不包含五因素和六因素模型。由于其他模型均未捕获 ROE 因子，因此永远不会包含 Q 因子模型。 Q 模型跨越了 Q5 模型，因为预期增长因子 r 并未在非洲市场定价。