Communication can build trust rather than merely transmit information. I show that asset managers' voluntary transparency about risk fosters trust and encourages investors to bear risk. Among otherwise-identical S&P 500 index mutual funds, investor letters discussing risk in more detail attract higher flows. To strengthen causality, I exploit corner bunching in risk detail using a control function approach. A pre-registered, randomized experiment isolates the mechanism: transparency about risk raises trust in management even as perceived risk rises and expected returns fall, contradicting a learning-based explanation. The effect concentrates among inexperienced investors, high risk-aversion market states, and anxious fund readerships.

If learning from data were a purely statistical problem, investors endowed with an identical dataset would use it alike. They do not. In tightly controlled trading contests where investors are restricted to a common set of signals, experienced investors respond more strongly to informative signals. I also uncover a familiarity bias toward recognizable signals. Gains from predictability concentrate among experienced investors, whose actions partly converge. I interpret these findings through a model in which investors solve a prediction problem under a bias that penalizes signal use. I show such behavioral frictions can sustain return predictability in equilibrium, even when every investor adopts the signal.

We develop an empirical model of how households form beliefs about stock market returns, utilizing advances in machine learning tools. Our approach estimates interpretable experience weights that vary across households while capturing a shared expectation formation process. These estimated experience effects explain close to one-fifth of the variation in subjective return expectations extracted from survey microdata. Interpreting the estimated weights reveals four stylized facts that offer new empirical foundations for theories of expectation formation. Most notably, we uncover a role for expected future states to influence the retrieval of past experiences, and illuminate households' perceived links between asset prices and the business cycle.

We show that the positioning of fund families influences price competition between individual mutual funds. We document that investors consider a limited set of families, and model the equilibrium consequences: a competition network arises, and funds have the incentive to internalize this market structure when setting fees. Calibrating our model with prospectus download data, we confirm that fees charged by seemingly homogeneous S&P 500 index funds are predicted by their families' positions in the consideration-based competition network. Counterfactual market structure analyses reveal that fees are one-third the level of monopoly fees but three times higher than competitive benchmarks.

We develop an approach that combines the estimation of monthly firm-level expected returns with an assignment of firms to (possibly) latent groups, both based on observable characteristics, using machine learning principles with linear models. The best-performing methods are flexible two-stage sparse models that capture group-membership predictive relationships. Portfolios formed to exploit such group-varying predictions based on a parsimonious set of characteristics deliver economically meaningful returns with low turnover. We propose statistical tests based on nonparametric bootstrapping for our results, and detail how different characteristics may matter for different groups of firms, making comparisons to the existing literature.

We examine how decision-makers' (DMs') ambiguity attitudes shape trust for two different sources of financial forecasting: human or machine learning (ML). In an incentivized laboratory experiment, we measure subjects' ambiguity attitudes and optimism regarding forecast accuracy for both sources. Our results reveal that DMs are similarly ambiguity-seeking and ambiguity-generated insensitive ("a-insensitive"; i.e., they insufficiently discriminate between changes in the likelihood of prediction accuracy), regardless of the analyst type. DMs hold more optimistic beliefs about the accuracy of ML analysts, which predicts higher trust in ML analysts over human analysts. However, DMs who are more a-insensitive are less likely to incorporate their beliefs into their trust. DMs' a-insensitivity increases with financial literacy, suggesting that financially literate DMs perceive greater ambiguity in prediction accuracy. Our findings demonstrate that a-insensitivity acts as a cognitive barrier between beliefs and trust.