Job Market Paper
Using corporate bond market evidence, we show that sophisticated investor attention allocation explains large variation in price underreaction to information. We hypothesize that investors have information-processing constraints so prices underreact to information, but investors optimally allocate more attention to more payoff-relevant news, so prices underreact less to such news. Consistent with this “sophisticated inattention” hypothesis, corporate bonds with higher credit risk underreact less to default risk news and bonds with longer duration underreact less to interest rate news. As the credit quality and duration of a bond changes over time, the price response speed to different news also changes accordingly, consistent with investors adjusting attention allocation to changes in circumstances. Many investors do appear to face binding attention constraints, as larger shocks from one risk consume more attention and result in slower price response to the other risk. Differences in bond liquidity cannot explain our findings, because if a bond is illiquid, it should be equally slow to respond to all news, but we find large differences in response speeds at the risk level. Finally, the amount of “money left on the table” is small enough to be plausibly explained by information-processing costs.
We demonstrate that uninformed demand shocks drive a substantial fraction of Fama-French size and value factor movements. From 1965 to 2015, retail investors frequently made large uninformed capital reallocations across mutual funds with different factor exposures, creating significant factor-level demand shocks in the stock market. These flows generate large swings in prices that reverse in the subsequent 1-3 years, explaining 30% of factor return variance. We argue that flows create such large price movements due to slow-moving liquidity provision: other investors are slow to absorb the flows. Existing rational or behavioral models with frictionless liquidity provision cannot explain observed patterns.
Trading intermediaries can execute trades as agents, taking no risk and earning commissions, or as principals, taking full risk and earning mark-ups. We argue this choice is determined by trading off intermediary holding cost against the cost of monitoring agents. In agency trading, intermediaries lack incentive to execute well so clients need to monitor. In principal trading, intermediary incentive is aligned but require compensation for holding inventory. Empirically, agency is indeed prevalent in transparent, high-turnover markets where monitoring costs are low. The spike of agency trading in corporate bonds during the financial crisis is consistent with the holding cost prediction.
Work in Progress
Recent studies use mutual fund flows to infer which asset pricing model investors use. Among the tested models, the Capital Asset Pricing Model (CAPM) was found to be "closest to the true asset pricing model." We show that, in fact, fund flow data is most consistent with investors relying on fund rankings (e.g., Morningstar ratings) and chasing past returns. We also show that investors do not adjust for market beta or exposures to other risk factors when allocating capital among mutual funds. Flows are weaker for high-volatility funds only because Morningstar penalizes funds for high total volatility.