Program

Anheuser-Busch Dining Hall, Knight Center, 3rd Floor

Emerson Auditorium, Knight Hall

Audience Questions

Refreshments available right outside the classroom.

Transparency as a Double-Edged Sword: Evidence from Social Score Disclosure in the Agency MBS Market

ABSTRACT

Classic disclosure theories predict that transparency improves market liquidity by reducing information asymmetry. Yet for assets that trade on fungibility, greater information may instead increase adverse selection and impair market liquidity. We examine this dual role of transparency through the lens of social score disclosure in the agency mortgage-backed securities (MBS) market. This market provides a unique setting because the same MBS can be traded as individual bonds in the specified pool (SP) market or pooled with similar MBS and traded using a standardized contract in the to-be-announced (TBA) market. We first show that social scores are informative about MBS prepayment risk. We then contrast the impact of social score disclosure across the SP and TBA markets. In the SP market, MBS with higher scores (i.e., lower prepayment risk) enjoy a price premium and improved market liquidity after the disclosure. However, since social score disclosure allows sellers to better identify the riskiest eligible MBS to deliver for a given price and higher-quality bonds are more likely to be traded in the SP markets, the TBA market experiences lower prices and liquidity. Our findings provide novel evidence that transparency improves price discovery in the SP market but exacerbates adverse selection in the TBA market, highlighting the detrimental role of transparency in the market depending on fungibility and depth.

Audience Questions

Refreshments available right outside the classroom.

Scroll to the Top: Visual Cues and the Primacy Effect in Consumer Lending*

ABSTRACT

We examine whether visually salient cues can improve borrower decision-making in consumer lending. We conduct a randomized controlled trial on an online lending platform where retail borrowers receive a personalized list of loan recommendations based on their application criteria. We test whether labels highlighting favorable loan attributes, such as lowest interest rate, fastest processing time, or highest approval rate, can mitigate the primacy effect—a tendency to select the first-listed loan on the recommendation list. We find that the label treatment does not meaningfully mitigate this effect. Although visual labels increase the likelihood that a loan is selected, their benefits are largely confined to labels attached to loans appearing first, and to a lesser extent, second on the recommendation list. We further show that a significant subset of borrowers does not apply for labeled loans, particularly when these loans appear lower on the recommendation list. These borrowers exhibit a stronger tendency to rely on simple heuristics, and their loan choices are also more likely to be inferior to those of control borrowers. Overall, our findings underscore the nuanced role of visual cues in shaping financial decision-making.

Audience Questions

Anheuser-Busch Dining Hall, Knight Center, 3rd Floor

Can AI Do Financial Research? LLM-Guided Hypothesis Discovery in Asset Pricing∗

ABSTRACT

We study whether an AI research agent can autonomously execute the hypothesis discovery loop in empirical asset pricing. We place a large language model inside a human designed research environment comprising a symbolic language of interpretable accounting formulas, an automated validation layer, and a fixed empirical evaluation pipeline. The agent searches over interactions between economic themes such as profitability, investment, valuation, and quality, and updates its proposals over successive generations, where each generation is a new round of proposal, testing, and revision based on standardized empirical feedback. Across eight theme pairs and seven generations, the system proposes and evaluates 280 candidate signals on a microcap-excluded universe; 159 clear a conventional significance screen in the predicted direction, and 38 survive multivariate horse races designed to isolate signals with independent predictive content. We then subject these survivors to a full battery of modern asset pricing tests, including multiple testing corrections, multi model factor spanning, and novelty tests against 209 published anomalies, and identify a small set that carries genuinely incremental information and survives the strictest filters. The paper introduces a transparent architecture for AI-guided hypothesis discovery in finance in which human researchers design the laboratory environment and the AI research agent autonomously carries out the discovery loop.

Audience Questions

Refreshments available right outside the classroom.

The Use of Debt Covenants to Curb Risk Shifting*

ABSTRACT

“The Use of Debt Covenants to Curb Risk Shifting”

Debt contracting invites excessive risk taking on the part of borrowers. We study how accounting-based debt covenants affect a borrower’s project choice if there is uncertainty about the borrower’s economic state. In general, covenants serve a dual role: cutting losses on loans expected to go bad and disciplining the borrower’s risk-shifting incentive. By increasing the expected recovery, the liquidation option lowers the face value of debt, thereby reducing the borrower’s incentive to take risks. We show that such covenants can curb risk shifting on projects seen through to completion, but never for borrowers in the bad state. Moreover, the value of covenants is non-monotonic in the financing needs (or leverage). Covenants become more effective the more informative the accounting system and the more uncertain the project quality (e.g., in R&D-intensive firms). We also allow the borrower to exert effort at the outset to increase the value of projects seen through to completion. Contrary to standard incomplete contracting (“hold-up”) arguments, this can strengthen the case for covenants: the benefit of committing to safer projects can outweigh the risk of upfront effort being wasted due to inefficient liquidation. The model generates empirical predictions relating the incidence of covenants to accounting quality, the ex-ante uncertainty about borrower characteristics, financial leverage, and growth options.

Audience Questions

Refreshments available right outside the classroom.

Geopolitical Tensions and Supply Chain Disclosures: Evidence from Shipment Redactions

ABSTRACT

We examine whether geopolitical tensions between the U.S. and foreign countries influence U.S. firms’ disclosure of suppliers in those countries. U.S. firms sourcing supplies from foreign countries facing higher tensions with the U.S. may face more intense scrutiny at home over these business ties and, as a result, be more reticent in disclosing ties to these foreign suppliers. Using granular shipment-level data covering around 148 million U.S. import shipments from 150 countries, we find that supplier identities are more likely to be concealed in import shipments from countries facing higher geopolitical tensions with the U.S., within the same product in the same year. Consistent with the incentive to avoid attention from the government, the effect is significantly more pronounced in industries facing higher regulatory scrutiny and when the foreign suppliers are more likely to have ties to their own governments. The effect is more pronounced in lower valued imports likely attributable to small and medium enterprises (SMEs) that face weaker information demand from capital markets. The effect is significantly more prevalent for more geopolitically sensitive products, for capital goods, and for more complex products. We highlight geopolitical tensions as an important force shaping supply chain disclosures in today’s global economy.

Audience Questions