Paris December 2018

This paper uses data on shareholder proposals to study how leverage affects the interaction between firms and labor unions. We find a negative association between financial leverage and shareholder proposals sponsored by unions. Our results are consistent with the idea that capital structure affects labor unions' behavior and suggest that debt deters labor unions from engaging in negotiation tactics. Additional tests indicate that the negative association between debt and union proposals is driven by governance proposals, in particular proposals on executive compensation and board elections, and more pronounced in firms in poorer financial condition.

This paper analyses the purchase and redemption behaviour of mutual fund investors and its implications on fund liquidity risk. We collect a novel set of proprietary data which contains a large number of French investors holding funds with various degrees of asset liquidity. We build a Self-Exciting Poisson model capturing fund flows' clustering effects and over-dispersion. The model improves the forecast accuracy of future flows and provides a reliable risk indicator (Flow Value at Risk.) Accordingly, we introduce the notion of liability risk where investor's behaviour increases mutual fund liquidity risk. We further decompose fund flows into investor categories. We find that investors exhibit high heterogeneous behaviour, and a lead-lag relation exists between them. Finally, we control flow dynamics for various economic conditions. We show that although flows evolve with economic conditions, investor's behaviour stays the main significant determinant of flows' randomness. Our findings encourage fund manager to adopt an ALM approach.

This paper studies the effect of the firm-journalist network on news tone and stock returns. Using a proprietary dataset on the firm's and the CEO's connections to the Wall Street Journal (WSJ) reporters, I find that such connections lead to markedly more favorable coverage of corporate M&A news and to better associated market re-actions. The effect on the financial market is larger for the deals featured on the front page of the Journal. Evidence suggests that the relationship is causal: First, using the reporters' turnover as an instrument for the connected coverage, I observe that the news slant and market effects remain significant. Additionally, using Rupert Murdoch's acquisition of the WSJ as an exogenous shock to journalistic independence, I find that firms previously connected to Mr. Murdoch received better coverage and more positive stock returns after the ownership change.

We develop a continuous-time intertemporal CAPM model that allows for risky beta exposure, which we explicitly specify. In the model, the expected return on a stock depends on beta's co-movement with market variance and more generally with the stochastic discount factor and deviates from the standard security market line when beta risk is priced. When estimating the model on returns and options we find that allowing for beta risk helps explain the expected returns on the low and high beta stocks, which are challenging for standard factor models.

We develop an asset pricing model to analyze the joint impact of liquidity costs and market segmentation. The freely traded securities command a premium for liquidity level and global market and liquidity risk premiums whereas securities that can only be held by a subset of investors additionally command a local market and liquidity risk premiums. Based on a new methodology, we find that the liquidity level premium dominates the liquidity risk premiums for our sample of 24 emerging markets. Whereas the local liquidity risk premium is empirically small, the global market liquidity risk premium dramatically increases during crises and market corrections.

Low-quality securities class action lawsuits disproportionally target firms with valuable innovation. We establish this fact using data on lawsuits against U.S. corporations between 1996 and 2011 and the private economic value of a firm's newly granted patents as a measure of valuable innovation. We find that securities class actions impose a substantial implicit "tax" on highly innovative firms. Regarding the channel, our findings suggest that changes in investment opportunities and corporate disclosure induced by the innovation make successful innovators attractive targets of low-quality litigation. Overall, our results provide new evidence consistent with the view that the current class action litigation system has adverse effects on the competitiveness of the U.S. economy.

What is the impact of negative interest rates on bank lending and risk-taking? To answer this question we study the changes in lending policies using the Euro area Bank Lending Survey and the Spanish Credit Register. Banks whose net interest income is adversely affected by negative rates are concurrently lowly capitalized, take less risk and adjust loan terms and conditions to shore up their risk weighted assets and capitalization. These banks also increase non-interest charges more. Importantly, we find no differences in banks' credit supply and credit standard setting, neither in Euro area nor in Spain, suggesting that negative rates do not necessarily contract the supply of credit, which can be interpreted in the sense that the so-called "reversal rate" has not been reached yet.

We offer a novel perspective on the negative relation between idiosyncratic volatility (IVOL) and expected returns. We show that the IVOL puzzle is largely driven by a mean-reversion behavior of the stocks' volatilities, which is not captured by a simple historic measure of IVOL. In doing so, we make use of option implied information to extract the expected mean-reversion speed of IVOL in an almost model-free fashion. Together with the current level of IVOL this method allows us to identify stocks' expected IVOL changes in a very general setting. Under the assumption of IVOL carrying a positive price of risk (Merton (1987)) we offer a explanation for the puzzle. In a horse race we show that the mean-reversion speed is superior to the most prominent competing explanations. All our findings are robust to different measures of IVOL and various stock characteristics.

Firms are connected if they borrow from the same lenders in the credit market. This study examines whether firm-level idiosyncratic shocks propagate in borrowing networks which are built on this type of firm link. Through increasing credit demand or loan defaults, borrower-level idiosyncratic shocks may drag the lenders into credit constraints, and then will negatively affect the subsequent credit supply to other borrowers that share the same lenders but not directly affected by these shocks. I identify idiosyncratic shocks with the occurrence of major natural disasters in the U.S. for almost 30 years. I find that disaster-affected relationship-borrowers receive more loans after the disaster, and impose substantial loan declines, output losses and equity value drops on their connected peers who are not affected by the natural disasters. This spillover effect is more severe for these connected peers in weaker relationship with the common lenders. My estimates are economically large, suggesting that firm linkages in the credit markets are an important determinant of the propagation of idiosyncratic shocks in the economy.

We propose a 2-country asset-pricing model where agents' preferences change endogenously as a function of the popularity of internationally traded goods. We determine the effect of the time-variation of preferences on equity markets, consumption and portfolio choices. When agents are more sensitive to the popularity of domestic consumption goods, the local stock market reacts more strongly to the preferences of local agents than to the preferences of foreign agents. Therefore, home bias arises because home-country stock represents a better investment opportunity for hedging against future fluctuations in preferences. We test our model and find that preference evolution is a plausible driver of key macroeconomic variables and stock returns.

We study the extent of cross-asset learning in financial markets by examining spillover effects around mutual fund fire sales. We find that the well-documented impact-reversal pattern for the returns of fire sale stocks (e.g., Coval and Stafford, 2007) spills over onto the stock returns of economic peers with a magnitude that is around one fifth of the original effect. These spillovers extend to liquidity and are not explained by common funding shocks or the hedging activity of liquidity providers. We conclude that they represent information spillovers due to learning from prices, thus identifying cross-asset learning as an important driver for the commonality in returns and liquidity.

We document a new empirical phenomenon in which the positions of managed money (MM) traders, who are sophisticated speculators in the commodity futures market, as disclosed by the CFTC Disaggregated Commitment of Traders (DCOT) reports, can predict the cross-section of commodity producers' stock returns in the subsequent week. Specifically, if the DCOT reports an increase in long position, a decrease in short position, an increase in net position, or an increase in the ratio of long over short position of MM, then the stock price of producers of the same commodity would increase in the following week, and the finding is robust to a variety of choices of measures and weighting schemes. The results are more pronounced in firms with higher information asymmetry. Our finding further challenges the efficient market hypothesis. We find that the positions of MM have predictive power, though positions of producers (commercial hedgers) do not.

We model a two-tiered market structure in which an investor can trade an asset on a trading platform with a set of dealers who in turn have access to an interdealer market. The investor's order is informative about the asset's payoff and dealers who were contacted by the investor use this information in the interdealer market. Increasing the number of contacted dealers lowers markups through competition but increases the dealers' costs of providing the asset through information leakage. We then compare a centralized market in which investors can trade among themselves in a central limit order book to a market in which investors have to use the electronic platform to trade the asset. With imperfect competition among dealers, investor welfare is higher in the centralized market if private values are strongly dispersed or if the mass of investors is large.

We document a new empirical finding that, in the cross-section, the information contained in bank loans' forward-looking uncertainty measure can predict firms' returns across a range of time horizons, in that firms having loans with high uncertainty measure will under-perform firms having loans with low uncertainty measure. This effect is separate from previously documented asset pricing puzzles related to idiosyncratic volatility, analyst forecast dispersion, and credit risk. We believe return predictability arises because banks have private information regarding firms' future prospects including operating performance and cash flow uncertainty, and we indeed find the predictability of proxies of these two variables. A long-short strategy based on this finding can generate a significant alpha of over 7% per annum. In addition, the return effect is more pronounced for firms with fewer analyst coverage and lower institutional ownership.

We examine banks' choice between two costly instruments used to identify good loan applicants: direct screening by acquiring borrower-specific information and collateral requirements. We show that with longer relationships the preference for screening increases both in initial and in later periods. Total welfare is enhanced as a result of more efficient selection, and access to credit can be improved. The model explains the reduced incidence of collateral, and potentially higher interest rates in later periods for safer borrowers. The results are stronger under more intense bank competition. Our findings support policies conducive to enduring lending relationships.

How do analysts form earnings expectations? This paper documents that analysts’ beliefs are influenced by the recent performance of other industries that they cover. I show that negative shocks to one coverage industry lead analysts to make more pessimistic earnings forecasts for firms in another industry. Those pessimistic forecasts are less accurate and lower than the actual earnings. Analysts are affected even if the focal firms have no relationship with the shocked industry. These findings are consistent with the notion that analysts heuristically overgeneralize other industries’ performance and incorrectly lower their expectations based on noise rather than information. Moreover, I demonstrate that analyst overgeneralization has significant effects on financial markets: the resulting increase in analyst disagreement induces higher trading volumes and larger return volatilities, and the resulting analysts’ pessimism leads to temporary underpricing.

We use French matched employer-employee data to track skilled individuals entering the labor market during the late 1990s tech bubble. The boom led to a sharp increase in the share of skilled entrants in the tech sector, which offers relative higher wages at the time. When the boom ends, however, the wage premium reverses and these skilled workers end up with a 5.5% wage discount ten years out, relative to similar peers who started in a non-tech sector. Other moments of the wage distribution of the boom, pre-boom, and post-boom cohorts are inconsistent with explanations based on a selection effect or a cycle effect. Instead, we provide suggestive evidence that workers allocated to the booming tech sector accumulate human capital early in their career that rapidly becomes obsolete.

We document an economically significant relation between director turnover and prior firm performance. This relation manifests in idiosyncratic stock returns consistent with relative performance evaluation and the monitoring of actions attributable to directors. The director turnover-performance sensitivity increases substantially throughout the 2000s, and varies with a number of governance characteristics, most notably with the presence of an active external blockholder. Directors who leave firms following poor performance are significantly less likely to obtain new directorships in the future. In sum, the threat of replacement for poor firm performance has become an increasingly significant incentive for the directors of public corporations.

We investigate the role of High Frequency Traders (HFTs) during flash crashes. By using a new methodology to identify flash crashes, defined as sudden and extreme price movements which occur in relatively short time and then reverts to the initial level, we identify 65 flash crashes episodes among 37 stocks that belong to the CAC40 traded in the NYSE-Euronext Paris market in 2013. We show that HFTs are responsible for initiating the crash in roughly 70% of the considered events, and that they strongly contribute to exacerbating the consequences of the crash, especially at his climax. In most of the cases, instead of providing liquidity, they start selling more as the crash develops. HFTs do not even contribute to recovery after the end of the crash, but they continue to initiate selling orders. This is worryingly true even for HFTs which agreed to provide liquidity under a market making agreement, especially if flash crashes occur simultaneously on several stocks. Among the HFTs, Investment Banks HFTs played the largest role and are those that are the most aggressive in selling during flash crashes.

We study an economy with demand spillovers where firms' decisions to produce are strategic complements. Firms have access to an increasing returns to scale technology and choose their operating leverage trading off higher fixed costs for lower variable costs. The choice of operating leverage determines the firm's systematic risk, that is, determines how responsive the firm's profits are to an aggregate labor productivity shock, which is the only risk factor in this economy. We show that firms take excessive risk as they do not internalize that higher operating leverage increases the likelihood of a coordination failure where output is inefficiently depressed across the economy. More generally, our analysis suggests that individual risk-taking decisions aggregate into excessive output volatility in the presence of strategic complementarities among agents.

As the recent financial crisis illustrated, the default of certain entities can have disastrous effects on the economy. This paper presents a framework aimed at analysing the asset pricing and macro implications of the existence of "systemic defaults". This framework is flexible and tractable enough to simultaneously replicate the price fluctuations of various far-out-of-the-money (disaster-exposed) credit and equity derivatives. According to our estimation results, market data imply that the default of a systemic entity is anticipated to be followed by a 4% decrease in consumption. The recessionary influence of systemic defaults implies that financial instruments whose payoffs are exposed to such credit events carry substantial risk premiums.

We estimate a dynamic no-arbitrage term structure model that jointly prices the cross-section of Treasury bonds and special repo rates. We show that special repo rates on on-the-run Treasuries can explain almost 80\% of the on-the-run premium, but only after incorporating a time-varying risk premium on the special spreads of both on- and off-the-run bonds. We show that the repo risk premium is priced in the cross-section of off-the-run bonds with very low special spreads.