Using a comprehensive database of corporate news, we examine how bond mutual funds trade on the sentiment of news releases. We find that bond funds trade against the direction of news sentiment (e.g., selling after good news about a firm). The results are more pronounced in bonds that lie within a fund’s investment objective sector, and in bonds with high turnover and low information asymmetry, and in credit-rating news and news with positive sentiment. Funds that most frequently trade against news sentiment produce a higher alpha, and a source of such alpha is bond price reversals post news events. Fixed income mutual funds, dealers, and insurance companies complement each other in news trading, with mutual funds trading against news largely in the absence of dealers. Our study indicates that bond mutual funds represent a significant liquidity provider, upon corporate news events, in the market for corporate bonds.

Using a dataset of over a million local government procurement contracts in China, we study the social and economic costs of local government debt. Firms contracting with local governments with high maturing debt-to-fiscal income see an increase in accounts receivables that do not reverse, a decrease in cash balances, and an increase in the probability of litigation by creditors. These effects (1) are not driven by local economic conditions, endogenous government indebtedness, or self-selection into becoming government suppliers, (2) do not apply for government-linked firms, and (3) are larger for firms in areas with weaker labor and property rights. Affected firms are less likely to be repeat government contractors and more likely to see protests relating to non-payment of wages or pension contributions, suggesting that suppliers bear the costs involuntarily.

We use a model of a limit order book to determine the optimal tick size set by a social planner who maximizes the welfare of market participants. Our results show that when investors arrive sequentially and supply liquidity by undercutting or queuing behind existing orders, the optimal tick size is a positive function of the asset value and a negative function of stock liquidity. Intuitively, the tick size is a strategic tool a social planner uses to optimally affect investors' choice between liquidity demand and supply, thus mitigating the inefficiencies created by excessive undercutting and queuing. The policy implication of such findings is that both the European tick size regime and the 2022 SEC proposal dominate Reg. NMS Rule 612 that formalizes the tick size regime for the U.S. markets. Using data from the U.S. and the European markets we test our model's predictions.

We test for skewness preferences in a large set of observational panel data on online

poker games (n=4,450,585). Each observation refers to a choice between a safe option and

a binary risk of winning or losing the game. Our setting offers a real-world choice situation

with substantial incentives where probability distributions are simple, transparent, and

known to the individuals. Individuals reveal a strong and robust preference for idiosyncratic

skewness, which has important implications for asset pricing. The effect of skewness

is most pronounced among experienced and losing players but remains highly significant

for winning players, in contrast to the variance effect.

We find that active mutual funds owning product market competitors have superior risk-adjusted returns that are not driven by industry concentration, common selection, or stock-picking ability. These funds charge higher fees but also generate persistent net-of-fee returns for investors. Funds with higher common ownership are more active voters who are more likely to vote against executive incentives compensation and for directors with existing directorships in competitors. Our findings suggest some actively- managed mutual funds have an incentive to soften product market competition and that proxy voting could serve as one mechanism for influencing corporate policy.

Using a novel representative sample of digital payment data, we observe a robust U-shaped relationship between individual investors’ monthly entertainment-related consumption and stock market returns in the previous month. Contrary to the prediction of the wealth effect, individuals increase their entertainment-related consumption after experiencing large positive and negative stock market shocks. We show that the latter effect, termed “financial retail therapy,” is consistent with a dynamic model of Prospect Theory, and provide further evidence for it in a controlled laboratory experiment. Finally, we show that our results are not driven by income effects or wealth shock measurement errors.

Households exhibit "return chasing" behavior, so through asset reallocation channel, good stock market performance induces contractions in deposit supply. Using stock market performance as a shock to deposit supply, we trace out banks' deposit demand and identify the relationship between bank market power and the slope of deposit demand. Exploiting a fixed effects identification strategy by comparing branches with the same parent bank located in different cities within the same county, we find that bank market power makes deposit demand curve steeper. Steeper deposit demand curve attenuates the spillover effects on the local deposit market of stock market fluctuations. Counties with more bank market power also experience less contractions in small business lending when stock market performance is good. Overall, our results suggest that bank market power is important in insulating and stabilizing local deposit and lending market from the spillover effects of the stock market.

A beneficial effect of bank regulation may play out through the asset management sector. When asset managers count on a central bank to support market liquidity in a systemic event, they take on fire-sale risk that is excessive from a social perspective. However, the extent of risk-taking today also incorporates the spread that bank dealers would charge for absorbing fire sales tomorrow. If regulation constrains banks' balance-sheet space, the expected spread would be higher, reining in excesses in asset managers' risk-taking and ultimately raising welfare.

This paper investigates the impact of trading on the commodity futures risk premium. We focus on intra-commodity spreading positions and study the asset pricing implications of spreading pressure (SP), that is, spreading positions scaled by open interest, on the cross-section of commodity futures returns. We document that SP negatively predicts futures excess returns. A battery of empirical tests shows that SP helps separate commodities that trade based on economic fundamentals from commodities that are subject to market frictions introduced via commodity index investments. We propose an SP factor, a long-short portfolio based on SP that is priced in the commodity futures market, even after controlling for well-known factors, and is robust to accounting for omitted variable biases and measurement errors.

We uncover a significant relationship between the persistence of marketing employment strategy and fund performance in the U.S. mutual fund industry. Using regulatory filings, we show a large heterogeneity in fund companies' marketing employment share, which refers to the fraction of employees devoted to marketing and sales. Not only does the marketing employment share increase in family size and predict subsequent fund flows, but it is also persistent across fund families. A framework based on Bayesian persuasion and costly learning helps explain the observed strategic marketing decision. Regarding an optimal marketing plan, fund companies with different skill types commit to heterogeneous marketing employment strategies. Conditional on the skill level, fund companies' optimal marketing employment share responds to their past performance differently. Low-skill funds only conduct marketing following good-enough past performance, whereas high-skill funds maintain a high marketing employment share even with very poor past performance. Consistent with the model prediction, we show that the volatility of the marketing ratio is negatively correlated with the long-term performance of fund companies.

After a new direct board connection is formed to a product market peer, a firm's gross margin increases by 0.8 p.p. Gross margin also rises by 0.4 p.p. after a new connection is formed to a peer indirectly through a third intermediate firm. Board connections have positive profitability spillovers on the closest rivals, and the effects are stronger when the newly connected peers share major corporate customers, have more similar business descriptions, or are located closer to each other. Using retail scanner data, we further provide direct evidence that new board connections are related to higher product prices of consumer goods.

Using a sample of 101,834 cross-border deals announced between 1985 and 2018, we show that merger flows involve acquirers from more democratic countries than their targets. This result is primarily driven by a “pull” factor, that is, firms in countries with weaker democratic institutions attract cross-border deals. Democratic institutions have a fundamental influence on international merger activity because they are conducive to better corporate governance standards. Further decomposing the effect reveals that democratic institutions also play a direct role that is not due to investor protection or economic development. Our findings imply that democracy is an important, omitted determinant of cross-border mergers.

Private Equity (PE) has grown into a substantial asset class, but there remain major problems with measuring PE fund returns. Investors continue to use the internal rate of return (IRR) as a key measure of fund performance. It is well known that early returns of cash can have a substantial impact on fund IRRs, but the magnitude and causes of this effect have not previously been systematically analysed. We demonstrate that the IRR is affected by two biases: a convexity bias, and a “quit-whilst-ahead” bias arising because the returns on PE projects tend to covary with their durations. Both bias the IRRs of PE funds upwards.

Using parametric and non-parametric estimation techniques, we show that these biases boost fund IRRs by an average of around 3% per annum. This represents a substantial proportion of the amount by which the average net PE fund IRR (around 12% per annum) appears to outperform returns on listed equity indices.

Fund cash multiples and PMEs become similarly biased if they are annualized to try to make them comparable with other assets. We further demonstrate that alternative performance measures which have been suggested by practitioners are also biased, which confirms how poorly understood these effects are. Failure to take proper account of these biases is likely to lead investors into badly misinformed investment decisions.

I present a theory of investor selling behavior in which the disposition effect arises because investors are slow to update their beliefs about the values of the assets they hold. I show numerically that the theory generates a disposition effect, propensity to sell functions, and other trading statistics that are in line with empirical estimates. I also show that the theory generates a reversed disposition effect at the end of the tax year, a weaker effect for more sophisticated investors, a stronger effect in more volatile stocks, and a disposition effect in short sales.

We study the impact of investors’ benchmarking concerns on market efficiency and asset pricing. Both separative and integrative learning technologies are examined as investors allocate limited attention across assets. We show that benchmarking can increase the price informativeness of benchmarked asset as investors optimally adopt integrative learning to observe a combined signal about asset payoffs. This is contrary to the result from the existing literature that assumes separative learning. Benchmarking can also increase the overall market efficiency with either type of learning. Yet, the implications for asset prices and comovements can be qualitatively different under different learning technology.

The value factor has no incremental pricing power in the Fama-French (2015) five-factor model. Thereby, its pricing power is primarily subsumed by the investment factor. We show that the strong relationship between the two factors arises because their sorting variables—book-to-market and investment—are both driven by shocks to expected cash flows and discount rates. We document that only stocks driven by shocks to discount rates contain the factors’ cross-sectional pricing information. The value and investment premia based on these stocks are more than 50% higher than the usual value and investment premia. Importantly, adjusted versions of the value and investment factors that use only such discount rate shock-driven stocks cannot subsume each other and improve the five-factor model’s pricing power. Thus, a value factor built from stocks for which book-to-market is actually a good indicator of expected returns captures incremental pricing information and is no longer redundant. Consequently, multifactor models should include such a value factor.

We find that loans to a borrower sharing similar technologies with the bank’s prior borrowers have lower loan spreads, likely due to reduced costs in loan screening and monitoring from bank’s accumulated knowledge. Such effect cannot be explained by product market competition, technology value and innovation ability or other firm characteristics. We show that borrower technology similarity is informative about firm creditworthiness. Despite identification challenges, we use a structural bank-borrower matching model to show that the total economic surplus for banks and borrowers can be enhanced by matching banks to borrowers with a high technology similarity to the bank’s prior borrowers. This technology similarity also plays an important role in the bank’s learning-by-lending process.

Unlike labor income, human capital is inseparable from individuals and does not accrue to creditors at default. As a consequence, human capital investment should be more resilient to “debt overhang” than labor supply. We develop a dynamic model displaying this important difference. We find that while both labor supply and human capital investment are hump-shaped in leverage, human capital investment tails off less aggressively as leverage builds up. This is especially the case when human capital depreciation rates are lower. Importantly, because skills acquisition is only valuable when households expect to supply labor in the future, the anticipated greater reduction in labor supply due to debt overhang back-propagates into a reduction in skills acquisition ex ante. Using individually identifiable data, we provide empirical support for the model.

This paper compares various machine learning models to predict the cross-section of emerging market stock returns. We document that allowing for non-linearities and interactions leads to economically and statistically superior out-of-sample returns compared to traditional linear models. Although we find that both linear and machine learning models show higher predictability for stocks associated with higher limits to arbitrage, we also show that this effect is less pronounced for non-linear models. Furthermore, significant net returns can be achieved when accounting for transaction costs, short-selling constraints, and limiting our investment universe to big stocks only.

We study the extent to which actively managed mutual funds bet on the CEO. Focusing on firms with CEO turnovers in a particular month, we find significantly higher trading activity and exit rates for funds holding this stock in that month compared to all other months and compared to all other firms. The trading activity and exit rates are higher for raided CEOs and serial CEOs, consistent with some funds placing larger bets on CEOs with higher perceived managerial ability. In further tests, we find strong persistency in the tendency for some funds to bet on the CEO, and show that such funds are less likely to be team managed, and have larger portfolio weights on firms in industries where managerial skills are more valuable. They charge higher fees, but despite that, their net returns are similar. Overall, our results uncover that betting on the CEO is an investment strategy of some actively managed mutual funds. We finally show that this strategy is upheld in equilibrium in a model where the motive for trade is differences in opinion about the importance of the CEO.

We develop a machine learning (ML) approach to establish new insights into how memory affects financial market participants’ belief formation processes in the field. Using analyst forecasts as proxies for market beliefs, we extract analysts' mental contexts and recalls that shape forecasts by training an ML memory model. First, we find that long-term memories are salient in analysts’ recalls. However, compared to an ML benchmark trained to fit realized earnings, analysts pay more attention to distant episodes in regular times but less during crisis times, leading to recall distortions and therefore forecast errors. Second, we decompose analysts' mental contexts and show that they are mainly shaped by past earnings and forecasting decisions instead of current firm fundamentals as indicated by the ML benchmark. This difference in contexts further explains the recall distortion. Third, our comprehensive memory model reveals the significance of specific memory features and channels in analysts' belief formation, including the temporal contiguity effect and selective forgetting.

I show that, in a DSGE setting with heterogeneous shareholders, the stock market risk premium and volatility decrease as monitoring increases. Monitoring arises because inside shareholders have an incentive to extract private benefits from firm output, whereas outside shareholders have the incentive to limit this extraction. Monitoring varies positively with the share of outside shareholder ownership, such that monitoring represents a source of cross-sectional and time-series variation in equilibrium asset pricing moments. I present empirical evidence supporting these theoretical results.

The microfinance industry serves over 140 million borrowers worldwide, and has been hailed as a means of fighting poverty and stimulating growth of small businesses in low- and middle-income countries. Yet evidence from several countries suggests a negligible average impact of microcredit on the performance of small businesses. In this paper, I explore the impact of equity-like microfinance contracts with performance-contingent payments, which provide a greater amount of risk-sharing than the standard rigid microcredit contract. I conduct artefactual field experiments with a sample of business owners who were participating in two broader field experiments in Kenya and Pakistan that had provided their businesses with large capital injections. I find that contracts with performance-contingent repayments outperform standard microcredit contracts, by stimulating more profitable investment choices, especially for the most risk-averse individuals. Loss-averse individuals also particularly value equity-like contracts, which provide downside protection in return for upside profit sharing. However, individuals who exhibit non-linear probability weighting prefer debt contracts, especially in the presence of a skewed profits distribution (where there is a low probability of very high outcomes, which such individuals overweight). I structurally estimate these three distinct dimensions of risk preferences using a prospect-theoretic model and show that relatively simple tweaks to contract design (specifically, capping upside profit sharing) can improve the feasibility of equity-like contracts. By utilising financial technology developments that improve screening and monitoring, and by taking into consideration three distinct elements of risk preferences, financial institutions that cater to small firms can unlock new forms of performance-contingent capital to provide better risk-sharing and improve client welfare.

I quantify the costs of realized flood disasters for banks and create a novel measure of bank-level flood risk exposure using expected flood risk estimates and mortgage lending data. I document that banks with large shares of mortgages in affected areas experience lower profits and capital ratios following flood disasters. In the cross-section of stock returns, small banks with high exposure to flood risk underperform other banks, on average, by up to 9.6% per year; this implies that exposure to flood is not fully priced. Underperformance persists when controlling for the negative effects of disasters on realized returns and adjusting for investors’ climate change concerns. The findings support regulatory concerns that bank equity is exposed to physical risk from climate change.

This paper uses proprietary data on self-reported employee reviews from Glassdoor.com to study the relationship between employee satisfaction and mutual funds’ performance. Using the staggered adoption of Anti-SLAPP (Strategic Lawsuits Against Public Participation) laws in the U.S. and variation from mergers between asset management companies to identify exogenous variation in job satisfaction, we find that employee satisfaction is positively linked to fund performance and size but that only performance-critical employees' satisfaction matters. A one-point increase on the 5-point scale of employee satisfaction leads to a 36bps increase in annual abnormal fund performance. Finally, while there is a positive effect of employee satisfaction on risk-taking, we cannot establish a causal relationship.

We evaluate the performance of common stock trade classification algorithms to infer the trade direction of option trades. Using a large sample of matched intraday transactions and Open/Close data, we show that the algorithms’ success to classify option trades is considerably lower than for stocks. The weak performance is due to sophisticated customers who often use limit orders instead of market orders to implement their trading strategies. These traders’ behavior varies over time and across exchanges with different pricing models. We introduce new rules that enhance existing algorithms and improve classification accuracy by 9% to 47%. Applying our new rules to construct a long-short trading strategy for stocks based on option order imbalance increases Sharpe ratios from 2.65 to 4.07.

In this paper, we provide an estimate of the ex-ante risk premia on earnings announcements based on the option market. We find that the risk premia are time-varying and have predictive power on future stock returns. The well-documented positive post-earnings-announcement drift (PEAD) is present only when the risk premia are high. After controlling for the announcement risk premia, the PEAD factor of the literature no longer has any abnormal returns. Moreover, while trading option straddles is not profitable unconditionally, conditional on high ex-ante risk premia, it becomes profitable even net of transaction costs.

ETFs add a new layer of market-making to the corporate bond market that improves the market quality of the underlying bonds. Dealers use the flexibility of the primary corporate bond ETF market as a warehouse to manage inventory. The face value of ETF holdings in investment grade (high yield) bonds is 9.1% (25.9%) greater on the downgrade date than thirty days prior. Bonds eligible for inclusion in ETFs with the most active primary markets overreact less than other downgraded bonds from the same issuer. This new layer of market-making leads to a negative relation between ETF ownership and idiosyncratic volatility.

We examine the causal effect of politicians’ partisan ideologies on firms’ industrial pollution

decisions. Using a regression discontinuity design involving close U.S. congressional elections, we show that plants increase pollution and invest less in abatement following close Republican wins. We also find evidence of reallocation: firms shift emissions away from areas represented by Democrats. However, costs rise and M/B ratios decline for firms whose representation becomes more Democratic, suggesting that politicians’ ideological demands can be privately costly. Pollution-related illnesses spike around plants in Republican districts, suggesting that firms’ passthrough of politicians’ ideologies can have real consequences for local communities.

Using granular auction data on 15 million item purchases in Brazilian public procurements between 2005-2021, we document a widespread pattern that the lowest bidder (``kamikaze'') does not satisfy required formalities after the auction is concluded, which allows the second-lowest bid to win the auction. Such a pattern can be observed in up to 15-20% of procurement auctions and results in 15-17% higher procurement prices as compared to similar auctions procuring the same product or service items, organized by the same government institutions, and even having the same winning firm. Kamikaze firms are smaller, younger, and tend to be co-owned by the same ultimate owner as the winning firm. Using observed kamikaze behaviour as a marker, we aim to measure how higher procured prices contribute to real outcomes by public service providers by reducing the budget available for sourcing other items. Taking the case of hospital mortality data, we see an increased number of deaths in the four quarters after an increased fraction of procurement auctions involving kamikazes.

Our study introduces an experimental framework to examine the role of attention in the decision-making process, with a particular focus on its impact on learning and belief formation. In order to identify attention, we draw upon the Salient Theory developed by Bordalo, Gennaioli, and Shleifer (2012). We conduct a two-stage online experiment and uncover several noteworthy findings. Firstly, slightly more than half of the participants show salient thinking characteristics, indicating a proclivity to overweight the standout option among a set of alternatives. Secondly, our results reveal that participants tend to overreact to salient signals and, more importantly, that overreaction is mainly driven by salient thinkers. Additionally, salient thinkers exhibit a greater degree of optimism than others when they receive positive signals, which is further amplified when these signals are infrequent. Lastly, while the salience anomaly is more pronounced in the short term, it disappears over an extended estimation period.

Using a unique experiment in the credit market, we find that anonymous loan applications mitigate racial disparities. When names are on applications, ethnic minorities are 10.7% less likely to receive online loan offers than otherwise identical majority applicants; anonymizing applications eliminates such disparities. After receiving online loan offers, applicants need to visit lenders in person for identity verification before loan origination. Despite that race is revealed to lenders, racial disparities in loan origination also decrease. We do not find significant racial gaps in loan performance either before or after anonymization. Further tests show that accurate statistical discrimination is unlikely to explain our results.

Using comprehensive issuance-level information for dollar-denominated short-term debt, we show that investments by money market mutual funds (MMFs) significantly reduce issuers’ funding frictions. Issuers without MMF funding pay approximately 10 basis points more for placing their short-term debt, even when comparing issuers with small MMF investments to issuers without MMF investments. Funding costs increased for issuers who lost their MMF investors because of an exogenous regulatory shock to the MMF industry in 2016. Issuers who lose their MMF investors reduce their outstanding short-term debt by more than 50% and issue debt with shorter durations, suggesting that MMF investments reduce an issuers’ funding fragility.

We study the effect of a central bank digital currency (CBDC) on the money market. A CBDC is equivalent to a 100% reserve requirement to fund those transactions that require CBDC, contrary to transactions that require bank deposits that only need partial reserve backing. We find that a higher fraction of transactions conducted with CBDC will drain reserves and tend to increase the interbank rate. The effect of CBDC remuneration is however ambiguous. A higher CBDC rate increases its value as a payment instrument. This leads to lower funding costs and larger investment, decreasing or increasing the demand for reserves and the interbank market rate, depending on which effect dominates. We show that a cap on CBDC will reduce the interbank rate and the deposit rate, as banks need less deposits to buy reserves. A CBDC design with tiered remuneration does not bring additional benefits relative to a single (lower) remuneration rate.

We propose a novel method to impute daily mutual fund trades in individual stocks from data on quarterly fund holdings, monthly total net assets, and daily fund returns – so that the method can be applied to standard CRSP mutual fund data. We set up an (underidentified) system of linear equations and solve the underidentification issue with an iterative method that applies random and adaptive constraints on trade incidence. The method produces daily, stock-level trade estimates with associated confidence levels. Validation and simulation analyses using proprietary daily fund trading data show good accuracy, especially for larger trades.

This paper studies the link between government guarantees for banks and bank asset concentration. We show theoretically that these guarantees, when combined with high leverage, incentivize banks to further invest in asset classes they are already heavily exposed to. We confirm these predictions using U.S. panel data, exploiting exogenous changes in banks' political connections for variation in bailout expectations. At the bank level, we find that higher bailout probabilities are associated with higher portfolio concentration. At the bank-loan class level, we find that banks respond to an increase in their bailout expectations by further loading up on loan classes that already have a high weight in their portfolio.

I study the spillover effects of removing barriers to growth in one product market on entry and growth of firms in the downstream/customer market. Constrained firms produce low quality goods and, in turn, hamper access to high quality inputs for downstream firms. I exploit the repeal of product reservation in India, whereby hundreds of products stop being reserved for exclusive production by small firms. With an increase in production of high quality goods in the input market, entry in the downstream product market increases. Entrants are not worse on observable characteristics. Productive downstream incumbents grow and less productive ones shrink. My results imply that business dynamism has positive spill-over effects along the supply chain.

While various theories stress the importance of technology for cryptocurrency valuation, empirical evidence is limited. In this paper, we study whether technology aspects of cryptocurrencies matter for their valuations, using machine learning methods to construct a technology index from initial coin offering whitepapers. We then track down the performance of cryptocurrencies from their initial offering to long-term valuations. We find that the cryptocurrencies with high technology indexes are more likely to succeed and less likely to be delisted subsequently. Moreover, the technology index strongly and positively predicts the long-run performances of cryptocurrencies. Overall, the results suggest that technological sophistication is an important determinant of cryptocurrency valuations.

In this paper, I define employee morale as employees’ attitudes toward and perceptions of the tasks the employees have in the companies they work for, toward companies’ dynamics and working conditions, and toward their interactions with fellow colleagues. I explore how employee morale affects post-merger integration and performance and post-merger merged firm employee morale using various proxies. The paper makes several novel findings. First, mergers between firms with similar employee morale are more likely to merge and achieve greater short-run and long-run post-merger synergies. Second, firms with high and similar morale achieve better post-merger integration than firms with low and similar morale and complementary morale. Third, companies with similar morale experience a more rapid rate of completion and exhibit a higher likelihood of completion. Fourth, target companies with high employee morale take less time to be integrated into acquiring companies, regardless of the acquiring companies’ employee morale. Fifth, acquiring companies value the employee morale profile of target companies and they tend to go after target companies with high level and low dispersion in dimensions of employee morale. Sixth, distance between acquirer and target employee morale is negatively associated with post-merger employee morale level and positively associated with post-merger employee morale changes. Finally, the observed acquirer price runup reflects takeover rumors generated from acquirer employees.

We study how prepayment penalties influence credit availability and borrowers' post-origination loan performance using a fuzzy regression discontinuity design that exploits exogenous variation due to idiosyncrasies of US state law. Estimates show a prepayment penalty 1) reduces the probability of delinquency by 5.1%, 2) the odds of foreclosure by 4.9%, 3) increases the likelihood of loan origination by 2.5%, and 4) lowers interest rates by 6.4%. The effect sizes are larger among riskier borrowers. The findings are consistent with theoretical models' predictions that prepayment penalties act as a borrower commitment device which acts against adverse selection as refinancing leads to a deterioration of the quality of lenders' mortgage pools.

This paper studies how banking regulators should disclose the models they use to assess banks that have reporting discretion. In my setting, assessments depend on both economic conditions and the fundamental of banks' asset. The regulatory models provide signals about economic conditions and banks report the fundamental of their asset. On the one hand, disclosing the models helps banks to understand how their assets perform under different economic environments. On the other hand, it induces banks with assets that are socially undesirable to manipulate the report and obtain favorable assessments. While the regulator can partially deter manipulation by designing the assessment rule optimally, the disclosure of regulatory models is necessary. The optimal disclosure policy is to disclose the regulatory models when the assessment rule is more likely to induce manipulation and keep them secret otherwise. In this way, disclosure complements the assessment rule by reducing manipulation in cases that harm the regulator more. The analyses speak directly to the supervisory stress test and climate risk stress test.

Why are over-the-counter markets the dominant market structure for many asset classes? We argue that non-anonymous trading and repeated interactions facilitate trading relationships that solve a moral hazard problem. Using data from a large European OTC dealer, we provide show that customers with a strong relationship obtain much more favorable bid-ask spreads, and that this effect is even stronger during market stress. We present evidence that moral hazard, not adverse selection, drive relationships. Customers with strong relationships can credibly commit to a fixed trading size, to not trade with other dealers and to not exploit the dealer when quotes are stale. We also show that the scope of OTC relationships extends across asset classes and that investment banks' organisational structure is designed to ensure that traders offer discounts to highly valued customers.

I develop a real options model to explain average returns and return volatilities of stock portfolios sorted on the book-to-market ratio. While average returns increase monotonically across portfolios, return volatilities are U-shaped. My model combines business cycle variations with countercyclical economic uncertainty. Operating leverage and procyclical growth options make both value stocks and growth stocks risky, generating U-shaped return volatilities. Growth stocks additionally load on the negative variance risk premium which reduces their expected return. Using structural estimation, my model jointly fits average returns and return volatilities, thereby solving a long-standing problem in investment-based asset pricing. Further reduced-form evidence supports the model channels.

We combine high-frequency stock returns with risk-neutralization to extract the daily common component of tail risks perceived by investors in the cross-section of firms. We find that our tail risk measure significantly predicts the equity premium, variance risk premium and realized moments of market returns at short-horizons. Furthermore, a long-short portfolio built by sorting stocks on their recent exposure to tail risk generates abnormal returns with respect to standard factor models and helps explain the momentum anomaly. Incorporating investors’ preferences via risk neutralization is fundamental to our findings.