FOUCAULT Thierry

We study how data abundance affects the informativeness of financial analysts' forecasts at various horizons. Analysts produce forecasts of short-term and long-term earnings and choose how much information to collect about each horizon to minimize their expected forecasting error, net of information acquisition costs. When the cost of obtaining short-term information drops (i.e., more data becomes available), analysts change their information collection strategy in a way that renders their short-term forecasts more informative but that possibly reduces the informativeness of their long-term forecasts. Using a large sample of analysts' forecasts at various horizons and novel measures of their exposure to abundant data (e.g., social media data), we provide empirical support for this prediction, which implies that data abundance can impair the quality of long-term forecasts.

We model of the search for predictors by speculators (active asset managers) and use it to analyze how the improvement in data processing power and the growth in available data (“data abundance”) affect the diversity of trading signals used by speculators, the dispersion of their profits and the similarities of their holdings. Our central message is that data abundance and computing power do not have the same effects. In particular, an improvement in computing power always raises the bar for the quality of predictors that managers consider good enough to exploit while more data lower it when data becomes sufficiently abundant. When this happens, the diversity of speculators’ signals and the dispersion of their trading profits increase in equilibrium while their holdings become less correlated.

We model of the search for predictors by speculators (active asset managers) and use it to analyze how the improvement in data processing power and the growth in available data (“data abundance”) affect the diversity of trading signals used by speculators, the dispersion of their profits and the similarities of their holdings. Our central message is that data abundance and computing power do not have the same effects. In particular, an improvement in computing power always raises the bar for the quality of predictors that managers consider good enough to exploit while more data lower it when data becomes sufficiently abundant. When this happens, the diversity of speculators’ signals and the dispersion of their trading profits increase in equilibrium while their holdings become less correlated.

We analyze how computing power and data abundance affect speculators' search for predictors. In our model, speculators search for predictors through trials and optimally stop searching when they find a predictor with a signal-to-noise ratio larger than an endogenous threshold. Greater computing power raises this threshold, and therefore price informativeness, by reducing search costs. In contrast, data abundance can reduce this threshold because it intensifies competition among speculators and it increases the average number of trials to find a predictor. In the former (latter) case, price informativeness increases (decreases) with data abundance. We derive implications of these effects for the distribution of asset managers' skills and trading profits.

No summary available.

This thesis consists of three papers. The first two papers study information flows in financial markets, and the third paper studies how mutual fund families can use relatively discretionary revenue streams to channel profits to specific funds within the family. The first paper, co-authored with Daniel Schmidt, studies price and liquidity spillovers in financial markets. Using a quasi-natural experiment, we show that investors observe stock prices to extract signals, and use them to make trades. In the second paper, I show that investment funds acquire information through the securities lending market. I show that active mutual funds sell stocks borrowed from them by short sellers, while index funds-which are prohibited from trading-do not. On the other hand, index funds are able to charge higher stock lending fees to borrowers. I attribute this to the fact that they are better lenders in the sense that they cannot use the information they get to trade, and thus profit from the information of short sellers. The third paper, also with Daniel Schmidt, studies the policies of fund families in allocating securities lending and lending income among member funds. We show that fund families deviate from the claimed fair allocation, directing more securities lending and loan profits to index funds. This finding is consistent with funds substituting lower management fees with higher securities lending income.

We analyze how computing power and data abundance affect speculators' search for predictors. In our model, speculators search for predictors through trials and optimally stop searching when they find a predictor with a signal-to-noise ratio larger than an endogenous threshold. Greater computing power raises this threshold, and therefore price informativeness, by reducing search costs. In contrast, data abundance can reduce this threshold because it intensifies competition among speculators and it increases the average number of trials to find a predictor. In the former (latter) case, price informativeness increases (decreases) with data abundance. We derive implications of these effects for the distribution of asset managers' skills and trading profits.

This doctoral dissertation consists of three independent papers on household finance and asset pricing. The first two papers are closely related, use similar data, and study the role of labor income risk in portfolio choice. The third paper studies a volatility model based on the multifractal Markov-switching model. The first paper is entitled "Countercyclical Income Risk and Portfolio Choices" (with Sylvain Catherine and Paolo Sodini). Using Swedish administrative panel data on wages and portfolio choices of individuals, we show that countercyclical income risk reduces households' willingness to invest in the financial market. The second paper is titled "Seeking Skewness". Using detailed Swedish administrative household data on portfolio and labor income, this paper examines the behavior of demand-skewness investors in their portfolio choice. The third paper is "Multifractal Volatility with Shot-noise Component" (with Laurent Calvet). Based on the Markov Switching Multifractal (MSM) model of Calvet and Fisher (2004), we develop in this paper a discrete time multifractal volatility model to capture jumps and declines in the volatility process.

This thesis studies the role and functioning of regional stock markets, and in particular the Lyon Stock Exchange, during the period 1800-1945. The framework of analysis used corresponds to the literature on competition between stock exchanges. The thesis is structured around five independent chapters or papers. Chapter 1 presents the conditions of the emergence of regional exchanges in their modern form in the mid-nineteenth century, nearly 40 years after the Paris exchange. Chapters 2 and 3 address the first point of the competitive dynamics among stock exchanges in France, i.e., competition over issuers, and provide two measures of financial development: the number of companies listed on each stock market between 1870 and 1913 and the number of IPOs over a shorter period, 1898-1909. Chapters 4 and 5 focus on competition in stock market activity, i.e., orders received by stockbrokers, government-appointed financial intermediaries. While chapter 4 is devoted to the period at the end of the 19th century and in particular to the crash of the Union Générale, chapter 5 focuses on the two world wars. These two chapters propose two measures of stock market volumes.

No summary available.

We measure demand for information prior to nonfarm payroll announcements using a novel dataset consisting of clicks on news articles. We find that when information demand is high shortly before the release of the nonfarm payroll announcement, the price response of U.S. Treasury note futures to nonfarm payroll news surprises doubles. We argue that this relationship stems from the fact that market participants have more incentive to collect information when uncertainty about asset payoffs is higher, as implied by Bayesian learning models. Thus, high information demand about macroeconomic news is a proxy for high macroeconomic uncertainty.

No summary available.

No summary available.

No summary available.

No summary available.

We propose a new model of interdealer trading. Dealers trade together to reduce their inventory holding costs. Core dealers share these costs efficiently and provide liquidity to peripheral dealers, who have heterogeneous access to core dealers. We derive predictions about the effects of peripheral dealers’ connectedness to core dealers and the allocation of aggregate inventories between core and peripheral dealers on the distribution of interdealer prices, the efficiency of interdealer trades, and trading costs for the dealers’ clients. For instance, the dispersion of interdealer prices is higher when fewer peripheral dealers are connected to core dealers or when their aggregate inventory is higher.

In the first chapter, I examine whether short sellers are better informed about sovereign bond auctions than the market. I find, on average, a large increase in short-sale demand prior to the auction. Nevertheless, short-sale demand does not predict a future increase in yield. Thus, short sellers are not better informed about the outcome of auctions and do not interpret better than the market.In the second chapter, I develop and test a model explaining the gradual price decline observed in the days leading up to early asset sales such as Treasury auctions. In the model, risk-averse investors anticipate an asset sale whose magnitude - and thus price - is uncertain. I show that investors face a trade-off between hedging through a long position and speculating on the difference between the pre-sale price and the expected sale price. Due to hedging, the equilibrium price is higher than the expected selling price. As the sell date approaches, the uncertainty of the selling price decreases, the speculative short positions increase and the price decreases. Consistent with predictions, I find that the yield on Italian Treasuries increases by 1.2 basis points after the release of auction price information, relative to days without information.In the third chapter, I study the relationship between prices and repo rates during the subprime crisis. I find that Duffie's (1996) no-arbitrage relationship between prices and repo rates performs less well during the crisis. However, low repo rate bonds are 18.0% more likely to be more expensive than identical high repo rate bonds during the crisis, compared to only 9.0% before the crisis. Overall, while there are strong limits to arbitrage, prices and repo rates show greater co-movements during the crisis.

No summary available.

The first chapter shows that the returns of funds invested in illiquid stocks ("illiquid funds") are better than those of funds invested in liquid stocks. This difference stems from the ability of illiquid funds to select stocks. Stocks held by illiquid funds perform better than portfolios that have the same characteristics. Liquid funds report benchmarks against which their returns are greater. A portfolio of stocks held by illiquid funds performs better than a portfolio of stocks held by liquid funds. The second chapter documents a predictability of returns. In this chapter, opportunity periods are periods when the returns of stocks that are regularly analyzed by analysts (the tracked stocks) deviate from those of stocks that are not tracked (the neglected stocks). Subsequent returns on easily valued stocks are greater when opportunities were great, compared to periods when opportunities were limited. This behavior is consistent with a model where investors demand a premium to bear the risk of adverse selection. The third chapter explores when investment funds change their investment style (style is defined as risk exposure, taking into account the usual risk factors). Funds do not take more risk when it would be more profitable to do so. After having had bad returns, funds move towards the style of similar funds, but which have had good returns. The style of young funds deviates from the style of older funds. New fund managers deviate from the style of funds with old managers. When a fund takes more risk on one side, it does not try to systematically address the other sides of the risk.

No summary available.

No summary available.

We show that product differentiation reduces the informativeness of a firm's stock price (or its peers' stock prices) about the value of its growth opportunities. This results in less efficient exercise of a firm's growth options when managers rely on information in stock prices for their decisions. This informational cost of differentiation induces conformity in product market strategies and is larger for private firms. Hence, a firm should differentiate more after going public. We confirm this prediction empirically and show that the post-IPO increase in differentiation is stronger for firms with better informed managers or less informative peers' stock prices.

Firms significantly reduce their investment in response to non-fundamental drops in the stock price of their product-market peers. We argue that this result arises because of managers' limited ability to filter out the noise in stock prices when using them as signals about their investment opportunities. The resulting losses of capital investment and shareholders' wealth are economically large, and affect even firms that are not facing severe financing constraints or agency problems. Our findings offer a novel perspective on how stock market inefficiencies can affect the real economy, even in the absence of financing or agency frictions.

High speed market connections improve investors׳ ability to search for attractive quotes in fragmented markets, raising gains from trade. They also enable fast traders to obtain information before slow traders, generating adverse selection, and thus negative externalities. When investing in fast trading technologies, institutions do not internalize these externalities. Accordingly, they overinvest in equilibrium. Completely banning fast trading is dominated by offering two types of markets: one accepting fast traders, the other banning them. Utilitarian welfare is maximized with (i) a single market type on which fast and slow traders coexist and (ii) Pigovian taxes on investment in the fast trading technology.

This thesis consists of three separate chapters. In the first chapter, I test the hypothesis that the display format of financial information affects the decisions of individual investors. I show that a more efficient display allows individuals to better manage their limit orders by minimizing the adverse selection risk incurred by using these orders. This suggests that individual investors have bounded rationality. In the second chapter, I test whether liquidity-providing trading strategies can generate profits, after transaction costs, for the active traders who implement them. I show that only individuals in the highest performance decile can persistently beat the market using highly contrarian strategies that require the massive use of limit orders. Limit-to-arbitrage seems to explain this phenomenon. In the third chapter, I study individuals' strategies around earnings announcements. I show that round-trips that are implemented one day before an announcement generate on average higher profits and are shorter in duration than those implemented in normal times. Individuals close their winning positions on the day of the announcement, which may slow down the price adjustment following the announcement.

Short lived arbitrage opportunities arise when prices adjust with a lag to new information. They are toxic because they expose dealers to the risk of trading at stale quotes. Hence, theory implies that more frequent toxic arbitrage opportunities and a faster arbitrageurs' response to these should impair liquidity. We provide supporting evidence using data on triangular arbitrage. As predicted, illiquidity is higher on days when the fraction of toxic arbitrage opportunities and arbitrageurs' relative speed are higher. Overall, our findings suggest that the price efficiency gain of high frequency arbitrage comes at the cost of increased adverse selection risk.

Information processing filters out the noise in data but it takes time. Hence, low precision signals are available before high precision signals. We analyze how this feature affects asset price informativeness when investors can acquire signals of increasing precision over time about the payoff of an asset. As the cost of low precision signals declines, prices are more likely to reflect these signals before more precise signals become available. This effect can ultimately reduce price informativeness because it reduces the demand for more precise signals (e.g., fundamental analysis). We make additional predictions for trade and price patterns.

Peers' valuation matters for firms' investment: a one standard deviation increase in peers' valuation is associated with a 5.9% increase in corporate investment. This association is stronger when a firm's stock price informativeness is lower or when its managers appear less informed. Also, the sensitivity of a firm's investment to its stock price is lower when its peers' stock price informativeness is higher or when demands for its products and its peers' products are more correlated. Furthermore, the sensitivity of firms' investment to their peers' valuation drops significantly after going public. These findings are uniquely predicted by a model in which managers learn information from their peers' valuation.

Liquidity providers often learn information about an asset from prices of other assets. We show that this generates a self-reinforcing positive relationship between price informativeness and liquidity. This relationship causes liquidity spillovers and is a source of fragility: a small drop in the liquidity of one asset can, through a feedback loop, result in a very large drop in market liquidity and price informativeness (a liquidity crash). This feedback loop provides a new explanation for comovements in liquidity and liquidity dry-ups. It also generates multiple equilibria.

No summary available.

We develop a model in which the speed of reaction to trading opportunities is endogenous. Traders face a trade-off between the benefit of being first to seize a profit opportunity and the cost of attention required to be first to seize this opportunity. The model provides an explanation for maker/taker pricing, and has implications for the effects of algorithmic trading on liquidity, volume, and welfare. Liquidity suppliers and liquidity demanders trading intensities reinforce each other, highlighting a new form of liquidity externalities. Data on durations between trades and quotes could be used to identify these externalities.

In the first chapter, I show that traditional liquidity measures, such as market depth, are not always relevant for measuring investor welfare. I build a model of an order-driven market and show that high liquidity supply can correspond to poor execution conditions for liquidity providers and relatively low welfare.In the second chapter, I model the speed of price adjustments to news arrivals in order-driven markets, when investors have limited attention capacity.Due to their limited attention, investors imperfectly follow news arrivals. Due to their limited attention, investors imperfectly follow the arrival of news, so prices adjust to the news after a certain delay. This delay decreases as the level of attention of investors increases.The delay in price adjustment also decreases as the frequency at which news arrives, increases. The third chapter presents a work written in collaboration with Thierry Foucault. We build a model to explain how high-frequency trading can generate "mini-flash crashes" (a sudden change in price followed by a very rapid return to the previous level). Our theory is based on the idea that there is a tension between the speed at which information can be acquired and the accuracy of that information. When high-frequency traders implement strategies involving rapid reactions to market events, they increase their risk of reacting to noise and thus generate "mini flash crashes". Nevertheless, they increase the informational efficiency of the market.

The way in which securities are traded is very different from the idealized picture of a frictionless and self-equilibrating market offered by the typical finance textbook. Market Liquidity offers a more accurate and authoritative take on liquidity and price discovery. The authors start from the assumption that not everyone is present at all times simultaneously on the market, and that even the limited number of participants who are have quite diverse information about the security's fundamentals. As a result, the order flow is a complex mix of information and noise, and a consensus price only emerges gradually over time as the trading process evolves and the participants interpret the actions of other traders. Thus a security's actual transaction price may deviate from its fundamental value, as it would be assessed by a fully informed set of investors. This book takes these deviations seriously, and explains why and how they emerge in the trading process and are eventually eliminated. The authors draw on a vast body of theoretical insights and empirical findings on security price formation that have accumulated in the last thirty years, and have come to form a well-defined field within financial economics known as "market microstructure." Focusing on liquidity and price discovery, they analyze the tension between the two, pointing out that when price-relevant information reaches the market through trading pressure rather than through a public announcement, liquidity suffers. The book also confronts many puzzling phenomena in securities markets and uses the analytical tools and empirical methods of market microstructure to understand them. These include issues such as why liquidity changes over time, why large trades move prices up or down, and why these price changes are subsequently reversed, why we see concentration of securities trading, why some traders willingly disclose their intended trades while others hide them, and why we observe temporary deviations from arbitrage prices.

No summary available.

No summary available.

In this thesis, we analyze the contribution of the options market to the informational efficiency and stability of the French equity market listed on Euronext-LIFFE for the period 1996-2006, and we propose to define to what extent the merger of Euronext with Liffe in 2002 and the macroeconomic conditions of 1996-2006 influence this contribution.The study of the introduction of options on the characteristics of stocks reveals (i) the absence of impact on both volatility and systematic risk of stocks measured by beta, (ii) a negative price effect that remains statistically insignificant in the majority of sessions, (iii) a significant increase in volume, (iv) a decrease in the price range The VAR modeling shows that the introduction of the option strengthens the adjustment of volume to stock volatility significantly only at the 10% threshold. Furthermore, the proportion of informed agents in the stock market is not higher following the introduction of options.The Log-ACD (Autoregressive Conditionnal Duration) modeling augmented by the introduction of liquidity as an explanatory variable does not detect any effect of the introduction of the option on the informational reinforcement of stocks.Furthermore, our study highlights that the objective of the options market is hedging (respectively speculation and/or arbitrage) in periods of high (respectively low) volatility. The merger of Euronext with Liffe in 2002 does not introduce any significant change in the improvement of the process of price adjustment to news. The overall result of the lack of effect of the introduction of the stock option is explained by the current volatility-based trading in France. This reduces the predictive power of options.

This thesis examines the impact of the behavior of informed market makers on the NASDAQ when market makers and analysts are affiliated with the same investment bank. The objective is to test the assumptions of existing models regarding the role of informed market makers in information dissemination and the impact that their behavior may have on market quality using sophisticated econometric models. In addition, the thesis empirically addresses the issue of anonymity and examines the role of the SIZE identifier in price formation around the affiliated analyst's public announcement and intraday. This component, now named NSDQ, allows holders to anonymously quote their prices in the book. The results show that before the announcement, the aggregate range is lower when analysts and dealers are linked and the informational shares of the identified quotes are higher than those quoted under SIZE. In intraday, quotes under SIZE are abundant when market conditions change. The topic is of institutional and practical interest. Indeed, the SEC is in favor of regulating information and increasing market transparency. However, the results of the thesis suggest that privileged access to keepers can make the trading environment less costly and thus improve the quality of this market. Furthermore, pre-trade anonymity allows the makers to better manage the risk of anti-selection in the presence of informed investors during the day.

The thesis is part of the "microstructure of financial markets" framework. Its objective is to study the impact of market organization on the supply of liquidity to market participants, and thus on market performance. The first test focuses on the quoting strategies of market makers in multiple markets. We consider two risk-averse market makers competing for order flow on two markets simultaneously. We show that the coexistence of an alternative trading system influences the market makers' reserve prices, which take into account the possibility that their quotes are affected either on the same or on opposite sides of the two markets. Moreover, unlike the traditional paradigm, the market maker with the most extreme position does not always place the best price. Our results thus suggest several empirical predictions about the interactions between liquidity bids in multiple markets. They also allow us to study the impact of several reforms on market performance. In particular, we show that requiring market makers to place identical quotes in multiple systems increases the ranges of best prices quoted in the dominant market. The second essay focuses on liquidity supply strategies in an anonymous order-driven market and a non-anonymous market. The model is based on the idea that some limit order placers have privileged information about the size of future price changes. We show that the move to anonymity affects market liquidity and the information content of the order book. In light of these results, we study the switch to anonymity on the Paris stock exchange, which took place on April 23, 2001. We first find that the order book contains information about future volatility, which corroborates our postulate. We then show that the change in transparency significantly reduced not only the price ranges, but also the information content of the order book, consistent with the predictions of the theoretical model. In the final essay, we study liquidity supply strategies in an opaque order-driven market that allows hidden limit orders, and in a transparent market that does not. The model is based on the idea that some limit order placers possess private information about the long-run value of the asset. Since the order book partially reveals this information, which decreases the likelihood of execution of informed orders, an informed liquidity provider can use these hidden orders not to mitigate the impact of his order. The model allows us to suggest new empirical predictions. Furthermore, we find that allowing hidden orders improves efficiency. On the other hand, counter-intuitively, it increases the transaction costs of liquidity seekers, and may decrease the expected profits of an informed liquidity provider.

The purpose of this thesis is to contribute to a better analysis of the impact of market practices on competition between market participants. The first essay focuses on the practice of private information exchange. We consider a market in which brokers have access to (i) 'fundamental' information on the net asset value of the risky asset and (ii) non-fundamental information on the volume of orders placed by uninformed investors. We show, in the framework of the Kyle (1985) model, that it can be profitable for two differently informed brokers to share information. Information sharing has an ambiguous impact on liquidity, but it increases informational efficiency and price stability. It also reduces transaction costs, suggesting a beneficial impact of this practice on market performance. The second essay analyzes how the practice of preferencing affects the quoting strategies of two risk-averse market makers. Preferencing allows a market maker - the so-called 'preferred' market maker - to execute orders exclusively, outside the central market. Regardless of the market structure (centralized or fragmented), we show that this practice reduces the incentives of market makers to quote competitive prices. It therefore leads to an increase in the price ranges and profits of market makers. However, the position risk involved in preferencing can generate losses for the executing market maker. In the third test, we show that the preferred market maker may therefore find it profitable to communicate the status of his preferred order flow to his competitor when he is not observing it. The communication helps to steer the posted prices in a direction that favors loss reduction. Using pre-opening data from the Nasdaq, we find that the market makers known to be the most favored (the "wholesalers") are among those who participate the most in the communication game between market makers. However, wholesalers' signals contribute the least to daily price discovery and are concentrated in stocks that are strongly linked to preferencing. Moreover, at the Nasdaq opening, we find that wholesaler prices and best market prices have a significant transitory component, supporting the idea of position risk communication.

The objective of our research is to contribute to the theory of the microstructure of financial markets. The first essay is devoted to the revelation of information by prices in a fixing market. We propose a model to analyze the role of the assumptions traditionally used in microstructure models studying price formation in the presence of information asymmetries. We show that the assumptions concerning the origin of the noise that prevents the equilibrium from being perfectly revealing are particularly important. Thus, depending on whether we consider this noise to be exogenous or endogenous, completely different results are obtained with respect to the properties of rational expectations equilibria (existence, uniqueness, quantity of information revealed by prices). In the second essay, we study the impact of transaction costs on the revelation of information through prices. In the framework of a Grossman-Stiglitz model, we show that an increase in transaction costs is always at the expense of informational efficiency. For this reason, transaction costs increase the value of information and induce agents to inform themselves. Thus, in some cases, an increase in transaction costs may result in an increase in the proportion of informed agents. On the other hand, transaction costs reduce the amount of trade that informed agents want to make. This makes it possible to obtain a rational expectations equilibrium even when these agents are risk-neutral or when they receive information of infinite precision. The third essay proposes a model of the auction mechanism involved in an order-driven market. We analyze, in a dynamic framework, the way agents determine their order placement strategies (best-order/limit-order choice). Furthermore, we characterize the bid and ask prices of agents placing limit orders. We show that there are no strategies strictly preferred by all agents. Furthermore, the imbalance between the number of buy orders and the number of sell orders is found to be a key determinant of bid and ask prices. Finally, we explain the price range between the best bid and ask prices by the strategic behavior of buyers and sellers and the need to compensate agents who place limit orders for the risk of non-execution This dissertation is devoted to the theory of financial markets microstructure.

The objective of our research is to contribute to the theory of the microstructure of financial markets. The first essay is devoted to the revelation of information through prices. We propose a model that allows us to analyze the role of the assumptions traditionally used in models studying price formation in the presence of information asymmetries. We highlight the decisive role of the assumptions concerning the origin of the noise that prevents the equilibrium from being perfectly revealing. In the second essay, we study the impact of transaction costs on the revelation of information by prices. We show that an increase in transaction costs is always at the expense of informational efficiency. For this reason, transaction costs increase the value of information and can increase the proportion of informed agents. The third essay proposes a model of the auction mechanism involved in an order-driven market. We analyze in a dynamic framework how agents determine their order placement strategies (best-order choice limit orders). In addition, we characterize the bid and ask prices of limit order givers.