<p>We study the implications of ambiguity under arbitrage constraints and in a market dominated by retail investors to understand its impact on asset pricing. We propose a novel approach to empirically measure stock level ambiguity and analyse how ambiguity-averse investors respond to varying levels of market ambiguity. In contrast to the previous research, we find a positive ambiguity premium. We attribute this to investors’ underreaction caused by the presence of significant market frictions in China. The results are robust across different factor models, alternative measure of market ambiguity, and varied portfolio formation approaches. Our findings suggest that ambiguity could be an important missing factor that could explain the much-debated high equity risk premium puzzle.</p>