We study merger and acquisition (M&A) disclosures made during regularly scheduled earnings calls. These bundled disclosures have been understudied in the literature, which focuses on dedicated M&A calls. Bundled deals represent an economically significant share of merger transactions. Bundling occurs when earnings news is weak. Bundled disclosures produce lower announcement returns by ∼130 basis points and are associated with reduced investor attention relative to dedicated calls. Bundling is more likely when bidders suffer from agency problems, such as powerful CEOs, entrenched boards, and low institutional ownership. In bundled calls, managers discuss the transaction in an optimistic tone that contrasts with the negative tone of the rest of the call. Overall, bundling seems to be an opportunistic attempt to put a positive spin on earnings news, to which analysts respond with skepticism. Our paper contributes to the literature on voluntary disclosure and contains important practical lessons for managers, board members, and market participants.
We examine whether prominent behavioural theories – prospect theory, salience theory, and regret theory – help explain investors’ stock choices in the real world. Whereas prior studies rely on indirect tests based on the cross-section of stock returns, we study investor behaviour directly using approximately five years of comprehensive limit order book data from the Taiwan Stock Exchange. We find that aggregate investor demand, proxied by buy-sell order imbalance, is most consistent with regret theory. At the investor-type level, however, the evidence points to substantial heterogeneity: domestic individual investors’ trading is most consistent with regret theory, prospect theory has greater explanatory power for non-individual investors, and salience theory has predictive power primarily for foreign investors. Overall, our findings highlight the importance of investor heterogeneity and of accounting for investor composition when evaluating behavioural theories in financial markets.
We study whether bank digitalization reallocates consumption across Chinese households. In a dynamic stochastic general equilibrium (DSGE) model with asset-rich landlords and asset-poor renters, digital upgrades expand aggregate credit while raising the collateral value of housing. The same shock strengthens savers’ balance sheets but increases rents and debt-service burdens for renters. We test these channels using a 2010–2022 household-province-bank panel that links 57,368 household-year observations to a bank digitalization index. A one-standard-deviation increase in bank digitalization raises saver consumption by about 4 percent, with no detectable effect for non-savers. Loan volumes and bank interest income rise together, and non-savers’ housing-expenditure response is about 12 percent larger than that of savers. The evidence points to “selective inclusiveness”: banking technology creates liquidity, but the consumption gains accrue mainly to households that already hold financial and housing buffers.
The successive waves of insider selling in China's securities market highlight the need to identify and understand the factors that influence such violations. This study explores the impact of regulatory distance—defined as the geographical distance between listed firms and their regulators—on insider selling activities. The results show that a one-unit increase in regulatory distance leads to a 14% rise in the frequency of insider selling and a 12% expansion in its scale. We identify two non-mutually exclusive channels—regulatory deterrence and information asymmetry—as key drivers of this impact, with the former being dominant. In response to the insufficient deterrence caused by regulatory distance, several complementary governance mechanisms are further examined. A key finding reveals that neither investor nor analyst attention creates deterrent pressure; instead, insiders tend to manipulate media coverage to leverage investor and analyst attention, thereby facilitating their insider selling.
We analyze the role of post-issuance derivative markets in providing incentives to design opaque securities using a Bayesian persuasion framework. While opacity potentially hurts the originator in the primary market by generating adverse selection, an opaque security allows the originator to profit from informed speculation in derivatives. The originator typically confronts this trade off by choosing an intermediate level of opacity. The more liquid the derivative market, the higher the expected profits from post-issuance speculation and thus the higher the equilibrium level of opacity. Opacity is also more attractive when adverse macro shocks tend to be rare. Our model is consistent with anecdotal evidence on the behavior of originators before the financial crisis as well as with the observed increase in popularity of opaque securities at a time of major explosion in derivative trading. Finally, it has implications for the current debate about a potential repeal of the Volcker rule.
Despite increasing attention paid by local stakeholders to corporate social responsibility (CSR), little is known about whether and how firms respond to this attention. To address this gap, in this paper I examine the impact of local stakeholders’ CSR attention, measured using the state-level Google Search Volume Index (SVI) for “corporate social responsibility,” on firms’ CSR activities. I find that firms increase their CSR commitments, particularly in the community, employee, and environmental dimensions, and improve the quality of their CSR reports in response to an increase in local stakeholder attention. This effect is more pronounced among larger and more visible firms, as well as those with larger boards and higher levels of board independence, female board representation, and CEO compensation. In contrast, I find no evidence that local stakeholder attention affects firms’ other investment activities. Finally, I show that firms’ active responses to local stakeholders’ CSR attention are associated with an increase in future firm value, consistent with stakeholder theory. Together, these findings highlight the key influence of local stakeholder attention on firms’ CSR strategies and outcomes.
This paper develops a capital structure model with a financial covenant that imposes a lower limit on a firm’s interest coverage ratio. Shareholders reduce their debt level or default whenever this ratio falls below the limit. In the model, firm value, debt repayment policy, and capital structure are derived explicitly. For low levels of the limit, shareholders prefer to reduce their debt every time the ratio reaches the limit. In other words, the covenant acts as early pressure on shareholders and eliminate their incentives to default. Then, it decreases the cost of debt but also lowers equity value by constraining shareholders. Because of this trade-off, the covenant can improve firm value. With the covenant, the firm can begin with high leverage to take advantage of the decreased cost of debt. The covenant tends to improve firm value for higher bankruptcy cost and volatility because these conditions lead to high expected default costs without a covenant. It can also improve firm value for higher growth and tax rates by easing the restriction on future debt issuance. These results are consistent with empirical evidence and support the optimal contracting hypothesis for covenants.
We evaluate the relation between various anomalies/factors and market conditions using a global set of 56 equity markets. The cross-section of anomalies performs significantly better during unfavorable market states: The value-weighted daily four-factor alpha of anomaly long-short portfolios is 1.7 bps in bad times while it amounts to 1.0 bps in good times. About 75.0% of the performance gain in bad times can be attributed to the anomaly short side. Findings remain robust considering alternative definitions of market conditions, including recessions, controlling for sentiment, and across anomaly categories or regions. The results underscore the overall importance of mispricing in explaining anomalies.
We show that the Dodd-Frank Act stress tests worsened bank loan terms and reduced vacancy postings by 16% among private firms with prior relationships with stress-tested banks. The decline is concentrated in postings for less-educated workers, indicating a contraction in hiring along this margin. These effects are temporary. Firms respond to tighter credit by shifting toward smaller financial institutions. This adjustment increases loan sizes from both new and existing lenders, which mitigates the impact of stress tests on labor demand over time.
This paper examines the Reaching-for-Coupon (RFC) phenomenon in U.S. corporate bond mutual funds. We define RFC as a portfolio tilt toward higher-coupon bonds relative to peers with similar yields. Using detailed bond-level holdings data from 2002-2018, we construct a novel fund-level RFC measure and show that high-RFC funds attract larger inflows, particularly in low-interest-rate environments. Crucially, investor flows into RFC funds are less sensitive to poor performance, leading to a less concave flow-performance relationship and mitigating redemption-driven fragility. These altered flow dynamics strengthen managerial incentives to take risk. Moreover, compared to Reaching-for-Yield (RFY) funds, RFC funds provide more stable income streams and are less exposed to credit downgrades. Our results demonstrate that RFC captures a distinct channel through which income-driven investor demand shapes risk-taking and fragility in bond markets.
We introduce a novel measure for investors' Degree of Rejoicing and Regret (DRR) and test its power to explain cross-sectional stock returns. Consistent with investors demanding compensation for anticipated regret, a portfolio of low-DRR stocks outperforms that of high-DRR stocks by 16.45% annually in the U.S. market. This DRR effect is present globally across 44 markets and is stronger in countries characterized by higher individualism, greater uncertainty avoidance, and weaker investor protection. Our analysis highlights the crucial role of rejoicing, a previously underemphasized component of regret theory, and demonstrates that our DRR measure subsumes the pricing power of existing regret-only proxies.
We examine the impact of the EU Banking Union on the information production of stress tests by exploiting the institutional shift in supervisory responsibility for significant banks to the European Central Bank (ECB) under the Single Supervisory Mechanism (SSM). We hypothesize that a centralized authority with both supervisory and financial stability mandates may reduce the informativeness of stress tests to mitigate negative spillovers, particularly potential threats to financial stability arising from disclosure. Our findings support this hypothesis, showing that the information production from stress tests declined following the introduction of the SSM. This reduction is primarily driven by weakly performing banks. We find no support for alternative explanations such as supervisory leniency, market learning, or the absence of an acute crisis.
We examine methodological uncertainty in studies of the cross-section of country equity returns, analyzing 15 predictors across up to 13,824 research implementations. Varying nine key design choices, we find that many classic signals—such as momentum, beta, or idiosyncratic risk—are surprisingly fragile. Setups emphasizing small, segmented countries enhance performance, while those focused on liquid, investable markets tend to weaken it. Applying bootstrap and out-of-sample tests, only a few factors, such as market size, dividend yield, short-term momentum, and sovereign risk, consistently emerge as robust. Our evidence calls for cautious interpretation of country-level return patterns and for robustness checks across alternative research designs.
A sharp rise in mortgage rates has “locked-in” fixed-rate borrowers, hampering relocations, escalating prices, and reshaping market outcomes. This phenomenon has implications for financial institutions, mortgage lenders, policymakers, and real estate professionals. Using a representative nationwide dataset, we find the probability of a home sale declines by 18.1% with each percentage point increase in the difference between current market rates and a homeowner’s fixed rate. This lock-in effect coincides with higher home prices and an estimated 1.72 million fewer transactions between 2022Q2 and 2024Q2, disproportionately affecting first-time buyers and lower-income households. By restricting supply when rates rise, mortgage rate lock-in introduces a supply-side channel through which reduced transaction volume can dampen or partially offset the usual price response to higher interest rates. Integrating a theoretical framework that accounts for accidental landlords and uses borrower-level evidence, we show mortgage contract structure shapes equilibrium housing market dynamics during periods of rapidly changing financial conditions. The findings provide a basis for researchers and practitioners in finance and real estate to assess the broader economic and systemic consequences of interest rate lock-in.
We study how an increase to the deposit insurance limit affects households’ portfolio allocation. Using unique data on individual deposit accounts, a suitable natural experiment, along with detailed information on Canadian households’ portfolio holdings, we show that households respond by drawing down deposits and shifting towards mutual funds and stocks. These outflows amount to 2.8% of outstanding household deposits. The mechanism underlying these portfolio adjustments relies on differences in deposit betas of insured vs. uninsured deposits. More generous deposit insurance coverage, hence, may result in non-trivial adjustments to household portfolios.
We examine the asset-pricing implications of selection neglect – a failure to correct for censored information – in the fine wine market. Using a Markov Chain Monte Carlo model to account for the endogeneity of trading, we measure this bias as the difference between past observed and corrected returns. We find a strong negative relation between our measure and future returns, indicative of an upward bias in the observed prices that is eventually corrected. This effect is mitigated by investor attention, amplified by ambiguity, and proves to be a short-term mispricing effect, vulnerable to transaction and carrying costs.
This paper investigates how historically intensive irrigation systems influenced enduring institutional and cultural traits that constrain firms’ access to finance. Combining geo-climatic measures of irrigation potential with firm-level data from 174 ethnic regions across 146 countries, we find that historically irrigated societies are characterised by weaker property rights, lower trust in financial institutions, and greater reliance on internal financing. Firms in these regions report more severe financial obstacles and higher rejection rates from banks. Implementing a spatial regression discontinuity design around the Lower Rhine and using irrigation potential as an instrument, we provide evidence consistent with a long-term influence of historical irrigation on modern credit frictions. The effects are most evident among privately owned domestic firms, unaffiliated firms, and those with higher female ownership. These findings indicate that ancient irrigation infrastructure is associated with persistent imprints on contemporary financial markets.
The application of generative artificial intelligence (AI) in financial forecasting is often hindered by concerns of look-ahead bias, raising questions about whether these models genuinely learn to predict or merely memorize. This study addresses this critical issue by introducing a novel framework to rigorously evaluate the predictive power of generative AI in a setting free from look-ahead bias. State-of-the-art vision transformer models, pre-trained on ImageNet-a dataset unrelated to finance and created prior to the study's data period-are employed to predict cross-sectional cryptocurrency returns from algorithmically generated OHLCV-like images. The top-performing model yields economically and statistically significant results. Specifically, a value-weighted long-short strategy achieves an average weekly excess return of 2.24% with an annualized Sharpe ratio of 1.89, outperforming both a benchmark convolutional neural network (CNN) and traditional characteristic-sorted portfolios. Further analysis reveals that the top-performing model effectively captures cross-sectional short-term momentum and size effects, with nonlinearity serving as a key driver of its performance. This research not only validates the superior predictive capability of generative AI models in a clean experimental setting, but also underscores their substantial potential for financial applications.
