This paper examines the relationship between gross return and relative bid-ask spread, extending Amihud and Mendelson's (1986) model to incorporate both their clientele effect (concave relation) and Jacoby et al.'s (2000) level effect (convex relation). The study demonstrates that the relationship between expected return and relative spread is concave for highly liquid assets with low relative spreads and convex for thinly traded assets with higher relative spread ratios. Using NASDAQ data, the paper provides empirical evidence supporting these findings. The results are consistent with Brennan and Subrahmanyam's (1996) empirical results, which show a concave relation between variable transaction costs and return premiums, and a convex relation between return premiums and fixed costs. The paper also finds that the relationship between return and relative spread is linear when controlling for convexity, supporting Jacoby et al.'s argument that liquidity adjustments in the CAPM eliminate non-linearity in the model, particularly for thinly traded assets. The study concludes that highly liquid NYSE stocks tend to lie in the concave range, where the clientele effect dominates the level effect, while less liquid NASDAQ stocks lie in the convex range where the level effect dominates the clientele effect. The findings support the extended Amihud and Mendelson model and highlight the importance of liquidity in asset pricing.This paper examines the relationship between gross return and relative bid-ask spread, extending Amihud and Mendelson's (1986) model to incorporate both their clientele effect (concave relation) and Jacoby et al.'s (2000) level effect (convex relation). The study demonstrates that the relationship between expected return and relative spread is concave for highly liquid assets with low relative spreads and convex for thinly traded assets with higher relative spread ratios. Using NASDAQ data, the paper provides empirical evidence supporting these findings. The results are consistent with Brennan and Subrahmanyam's (1996) empirical results, which show a concave relation between variable transaction costs and return premiums, and a convex relation between return premiums and fixed costs. The paper also finds that the relationship between return and relative spread is linear when controlling for convexity, supporting Jacoby et al.'s argument that liquidity adjustments in the CAPM eliminate non-linearity in the model, particularly for thinly traded assets. The study concludes that highly liquid NYSE stocks tend to lie in the concave range, where the clientele effect dominates the level effect, while less liquid NASDAQ stocks lie in the convex range where the level effect dominates the clientele effect. The findings support the extended Amihud and Mendelson model and highlight the importance of liquidity in asset pricing.