Value investing and gambling: An asset pricing model integrating Q-theory and salience theory

Grounded in investment Q-theory, we treat industry discount rates as proxies for value-investment opportunities and use salience theory to measure gambling behaviour, integrating both into a novel asset pricing model – the ST-ICAPM – which we evaluate across Chinese industries from 2014 to 2023. Empirically, the investment factor negatively predicts returns while growth commands a positive premium; static profitability carries pricing information, but in China it operates primarily through earnings growth, so the level factor loses marginal significance once growth enters. Salience intensity robustly and negatively predicts returns, concentrating in salience measured against prominent peers (eigenvector centrality), which subsumes the market-benchmarked factor. The relation strengthens in high-friction subsamples and survives full controls, placebo, and network-threshold diagnostics, consistent with behavioural overvaluation under limits to arbitrage. Within this sample, squared Sharpe-ratio tests reject equivalence between the ST-ICAPM and each benchmark (FF3, Carhart4, FF5, HXZ-ICAPM, STCAPM). These findings underscore the importance of incentivizing value investing and tempering narrative-driven speculation to strengthen the Chinese capital market.