We introduce a deep visualization (DV) framework that turns conventional commodity data into images and extracts predictive signals via convolutional feature learning. Specifically, we encode futures price trajectories and the futures surface as images, then derive four deep‑visualization (DV) predictors, carry ($bs_{DV}$), basis momentum ($bm_{DV}$), momentum ($mom_{DV}$), and skewness ($sk_{DV}$), each of which consistently outperforms its traditional formula‑based counterpart in return predictability. By forming long–short portfolios in the top (bottom) quartile of each DV predictor, we build an image‑based four‑factor model that delivers significant alpha and better explains the cross‑section of commodity returns than existing benchmarks. Further evidence shows that the explanatory power of these image‑based factors is strongly linked to macroeconomic uncertainty and geopolitical risk. Our findings reveal that transforming conventional financial data into images and relying solely on image-derived features suffices to construct a sophisticated asset pricing model at least in commodity markets, pioneering the paradigm of image‑based asset pricing.

It is diffffcult to understand the risk-return trade-off in option market with observable factormodels. In this paper, we employ a latent factor model for delta-hedge option returns over a varietyof important exchange traded options in China, based on the instrumented principal componentanalysis (IPCA). This model incorporates conditional betas instrumented by option characteristics,to tackle the diffffculty caused by short lifespans and rapidly migrating characteristics of options. Ourresults show that a three-factor IPCA model can explain 19.30% variance in returns of individualoptions and 99.23% for managed portfolios. An asset pricing test with bootstrap shows that there isno unexplained alpha term with such a model. Comparison with observable factor model indicatesthe necessity of including characteristics. We also provide subsample analysis and characteristicimportance.

Using comprehensive account-level data and 192 asset pricing anomaly signals, we investigate whether retail investors and institutions trade on anomalies in China. We find that retail investors tend to trade contrary to anomaly prescriptions, suggesting that they have a strong tendency to buy (sell) overvalued (undervalued) stocks. In contrast, institutions trade consistent with anomalies, indicating that they buy (sell) undervalued (overvalued) stocks. Regarding the information content of anomalies, we find that small retail investors trade contrary to trading-based anomalies, whereas institutions trade consistent with both trading- and accounting-based anomalies. Additionally, lottery stock preference and return extrapolation help explain investors’ trading behavior on anomalies.

This paper employs a proprietary data set on commodity producers’ profit margins (PPMG) and establishes a robust positive relationship between commodity producers’ profitability growth and future returns of commodity futures. The spread portfolio that longs top-PPMG futures contracts and shorts bottom-PPMG futures contracts delivers a statistically significant average weekly return of 36 basis points. We further demonstrate that profitability is a strong SDF factor in commodity futures market. We theoretically justify our empirical findings by developing an investment-based pricing model, in which producers optimally adjust their production process by maximizing profits subject to aggregate profitability shocks. The model reproduces key empirical results through calibration and simulation.

Beyond a bias toward local stocks, investors prefer companies in certain cities over others. This study uses the geographic network of investor-followed stocks from stock watchlists to identify intercity investment preferences in China. We measure the city-pair connectivity by its likelihood of sharing an investor in common whose stock watchlist is highly concentrated in the firms of that city pair. We find that a higher connectivity-weighted aggregate stock demand-to-supply ratio across connected cities is associated with higher stock valuations, higher turnover, better liquidity, and lower cost of equity for firms in the focal city. The effects are robust to controls for geographic proximity and the broad investor base, are stronger among small firms, extend to stock return predictability, and imply excess intercity return comovement. Our results suggest that city connectivity revealed on the stock watchlist helps identify network factors in asset pricing.

According to prospect theory [Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk, Eco- nometrica, 47, 263–292], gains and losses are measured from a reference point. We attempted to ascertain to what extent the refer- ence point shifts following gains or losses. In questionnaire studies, we asked subjects what stock price today will generate the same utility as a previous change in a stock price. From participants’ responses, we calculated the magnitude of reference point adapta- tion, which was significantly greater following a gain than following a loss of equivalent size. We also found the asymmetric adap- tation of gains and losses persisted when a stock was included within a portfolio rather than being considered individually. In studies using financial incentives within the BDM procedure [Becker, G. M., DeGroot, M. H., & Marschak, J. (1964). Measuring utility by a single-response sequential method. Behavioral Science, 9(3), 226–232], we again noted faster adaptation of the reference point to gains than losses. We related our findings to several aspects of asset pricing and investor behavior.

Beyond a bias toward local stocks, investors prefer companies in certain cities over others. This study uses the geographic network of investor-followed stocks from stock watchlists to identify intercity investment preferences in China. We measure the city-pair connectivity by its likelihood of sharing an investor in common whose stock watchlist is highly concentrated in the firms of that city pair. We find that a higher connectivity-weighted aggregate stock demand-to-supply ratio across connected cities is associated with higher stock valuations, higher turnover, better liquidity, and lower cost of equity for firms in the focal city. The effects are robust to controls for geographic proximity and the broad investor base, are stronger among small firms, extend to stock return predictability, and imply excess intercity return comovement. Our results suggest that city connectivity revealed on the stock watchlist helps identify network factors in asset pricing.

Investors' previous experiences with a stock affect their willingness to repurchase it. Using Chinese investor-level brokerage data, we find that investors are less likely to repurchase stocks that have increased in value since they were sold. We then construct a novel measure of Regret to capture investors' repurchase regret and investigate its asset pricing implications. Stocks with higher Regret experience lower buying pressure from retail investors in the future, leading to lower future returns. In terms of economic magnitude, portfolios with low Regret generate 12% more annualized abnormal returns. Further analyses show that the pricing effect of Regret is more pronounced among lottery-like stocks and those in which investors have previously gained profit. The results are robust to alternative estimations.

We apply various test portfolios and alternative statistical methodologies to evaluate the performance of eleven prominent asset pricing models. To compile the test portfolios, we construct 105 anomalies in China and apply the 23 significant anomalies as test assets for model comparison. The results indicate that in the time-series test and anomalies explanation, the Hou et al. (2019) five-factor q model exhibits the best overall performance. The pairwise cross-sectional R^2s and the multiple model comparison tests affirm that the Hou et al. (2019) five-factor q model, the Fama and French (2018) six-factor (FF6) model and the Kelly et al. (2019) five-factor Instrumented Principal Component Analysis (IPCA5) model stand out as the top performers. Notably, the performance of the five-factor q model is insensitive to variations in experimental design.

We utilize ridge regression to create a novel set of characteristics-based "ridge factors". We propose Bayesian Average Stochastic Discount Factors (SDFs) based on these ridge factors, addressing model uncertainty in line with asset pricing theory. This approach shrinks the relative contribution of low-variance principal portfolios, avoiding model selection and presumption of a "true model". Our results demonstrate that ridge factor principal portfolios can achieve greater sparsity while maintaining prediction accuracy. Additionally, our Bayesian average SDF produces a higher Sharpe ratio for the tangency portfolio compared to other models.