In Chapter 2, we present a model that allows us to study the effect of increased competition among charities for donations, and show that it will result in a lower provision of public goods. When charities get donations, they must pay two fundraising costs: a travel cost and an extra cost, a "premium" in our terminology. This premium arises from the extra time, effort, or incentives a charity must provide to garner a contribution from a donor who is solicited by other charities. Increased competition raises this premium, which leads to deadweight loss, so that revenue net of fundraising costs falls after a new firm enters into the market. A problem with public goods markets is asymmetric information between charities and donors, such that donors do not know which charities will cheat. In Chapter 3, we show that honest charities can get more donations than dishonest charities by investing in a capital stock. We study a two-period model under two assumptions, one where first-period investment does not affect the provision of public goods in the second period, and one where first-period investment does affect the provision of public goods in the second period. In the first case, we prove the existence of a separating equilibrium where honest charities make an investment and dishonest charities invest nothing. Thus, donors will donate more to charities that make investments, even if the investment is not used to produce public goods. In the second case, honest charities may invest the efficient amount, overinvest, or underinvest, depending on the donors' beliefs. In Chapter 4, we borrow parts of the models in the previous two chapters in order to see what effect the signaling cost has on the number of firms and average revenue. In our model, donor utility increases when they give to a charity that matches their ideology. We are interested in the long-run equilibrium, so unlike in Chapter 2, we assume there is free entry in the market. The two important results are that the number of firms decreases and average revenue increases if the required signaling cost increases.

In recent years, extensive research has been devoted to opportunistically exploiting spectrum in a distributed cognitive radio network. In such a network, autonomous secondary users (SUs) compete with each other for better channels without instructions from a centralized authority or explicit coordination among SUs. Channel selection relies on channel occupancy information observed by SUs, including whether a channel is occupied by a PU or an SU. Therefore, the SUs' performance depends on the quality of the information. Current research in this area often assumes that the SUs can distinguish a channel occupied by a PU from one occupied by another SU. This can potentially be achieved using advanced signal detection techniques but not by simple energy detection. However, energy detection is currently the primary detection technique proposed for use in cognitive radio networks. This creates a need to design a channel selection strategy under the assumption that, when SUs observe channel availability, they cannot distinguish between a channel occupied by a PU and one occupied by another SU. Also, as energy detection is simpler and less costly than more advanced signal detection techniques, it is worth understanding the value associated with better channel occupancy information. The first part of this thesis investigates the impact of different types of imperfect information on the performance of secondary users (SUs) attempting to opportunistically exploit spectrum resources in a distributed manner in a channel environment where all the channels have the same PU duty cycle. We refer to this scenario as the homogeneous channel environment. We design channel selection strategies that leverage different levels of information about channel occupancy. We consider two sources of imperfect information: partial observability and sensing errors. Partial observability models SUs that are unable to distinguish the activity of PUs from SUs. Therefore, under the partial observability models, SUs can only observe whether a channel was occupied or not without further distinguishing it was occupied by a PU or by SUs. This type of imperfect information exists, as discussed above, when energy detection is adopted as the sensing technique. We propose two channel selection strategies under full and partial observability of channel activity and evaluate the performance of our proposed strategies through both theoretical and simulation results. We prove that both proposed strategies converge to a stable orthogonal channel allocation when the missed detection rate is zero. The simulation results validate the efficiency and robustness of our proposed strategies even with a non-zero probability of missed detection. The second part of this thesis focuses on computing the probability distribution of the number of successful users in a multi-channel random access scheme. This probability distribution is commonly encountered in distributed multi-channel communication systems. An algorithm to calculate this distribution based on a recursive expression was previously proposed. We propose a non-recursive algorithm that has a lower execution time than the one previously proposed in the literature. The third part of this thesis investigates secondary users (SUs) attempting to opportunistically exploit spectrum resources in a scenario where the channels have different duty cycles, which we refer to as the heterogeneous channel environment. In particular, we model the channel selection process as a one shot game. We prove the existence of a symmetric Nash equilibrium for the proposed static game and design a channel selection strategy that achieves this equilibrium. The simulation results compare the performance of the Nash equilibrium to two other strategies(the random and the proportional strategies) under different PU activity scenarios. / Master of Science

Thesis advisor: Ryan Chahrour / Expectations formation is central to macroeconomics. Households, firms, and policymakers must form expectations not only about fundamentals, but about what other agents’ beliefs are, because others’ beliefs will determine their actions. The three essays in this dissertation examine empirically and theoretically how agents use both public and private information to form expectations. The first two essays combine a models of optimizing behavior and forecasting with data on the macroeconomy, financial prices, and macroeconomic forecasts to examine the extent to which economic agents learn about the macroeconomy from financial prices and monetary policy actions. The third essay examines theoretically how members of a committee use public and private information to form beliefs when they care both about having accurate forecasts and coordinating actions with others. All three essays emphasize that frictions in expectations formation are a salient feature of the world, and understanding the extent and importance of those frictions is important for both positive and normative questions in macroeconomics and finance. Beliefs about the future determine the willingness of financial market participants to save and invest, and theory suggests they should value more highly assets which are expected to pay higher returns during recessionary periods when consumption is otherwise low. Hence, financial prices reflect macroeconomic expectations. In the first essay, titled "Macroeconomic Disagreement in Treasury Yields," I explore how agents with idiosyncratic, private information form beliefs about both the macroeconomy and the beliefs of other agents. Using data on United States Treasury debt, the macroeconomy, and individual inflation forecasts, I estimate the precision of bond traders’ information about the macroeconomy and how much they disagree with each other. I allow for traders to learn both from private signals and from asset prices, which aggregate the beliefs of all the traders in the market. I find that bond prices are moderately informative about macroeconomic variables, but are the source of most of the information traders have about monetary policy and the beliefs of others. In contrast to studies which assume full information, risk premia are much less important than slow-adjusting interest rate expectations for explaining the behavior of long-run yields. The most important signal for bond traders appears to be the Federal Reserve’s short-run rate, which encodes information about the macroeconomy and the central bank’s intended future policy. Nevertheless, the fact that traders held disparate beliefs about the macroeconomy, and especially about the long-run inflation target of the Federal Reserve, elevated long-term yields on average. The first essay demonstrates empirically that financial market participants learn about the macroeconomy from monetary policy actions. However, it is silent on how monetary policymakers form beliefs about the macroeconomy, or how the information in monetary policy rates endogenously affects macroeconomic outcomes. In the second essay "Your Guess is as Good as Mine: Central Bank Information and Monetary Policy," I use data on private sector forecasts and forecasts from the Federal Reserve Board staff to examine the typical assumption of common information between firms and monetary policymakers. Using forecasts from a survey of professional forecasters and from the Federal Reserve Board staff, I show evidence against the typical assumption of common information between monetary policymakers and the private sector, and also that policymakers are, at best, only weakly better at forecasting than private forecasters. Based on this evidence, I augment an otherwise standard monetary policy model by relaxing the common information assumption. Instead, I assume there is idiosyncratic, private information among price-setting firms, and between firms and the central banker. Firms combine private information about aggregate conditions with the observed monetary policy rate to form expectations about fundamentals and the beliefs of rival firms. The central banker must form expectations about firms’ beliefs because those beliefs will determine inflation and overall economic activity. But as a result of their differences in information sets, firms must form expectations about other firms’ expectations, and what the central banks’ expectations of their expectations are. I examine the ability of this model to fit the data and find that the model can capture features of both firm and central bank inflation expectations, but in the absence of imperfect information among households, it is difficult to simultaneously match the forecast data and data on real activity. This result points to the sensitivity of models with dispersed information to the underlying assumptions about how central bankers will respond to exogenous shocks. The second chapter emphasized how the assumptions economists make regarding monetary policymakers’ information is critical for understanding their actions. Motivated by this example, my third chapter "Information Investment in a Coordination Game" explores theoretically how members of a committee who are uncertain about others’ beliefs decide on a binary action, and how their decision to pay close attention to public or private signals is related to their desire to accurately forecast versus coordinating their behavior with others. I show that when it is assumed that information decisions among committee members are symmetric - everyone pays the same amount of attention to the same things - there is a unique outcome of the coordination game. However, I further show that it is difficult to guarantee that committee members will all choose a symmetric allocation of information. Aside from the direct cost of acquiring better information, allocating attention to more accurate signals can harm welfare when coordination motives are dominant. In a set of numerical exercises, however, I show that it is possible for a unique equilibrium to exist, and that actions that do not have a large impact on the payoffs of committee members (such as changing the size of the committee) may nevertheless have large impacts on the accuracy of the committee’s forecasts. This suggests a possible tension between the welfare of the committee, which benefits from consensus, and the welfare of those affected by the committee’s actions, which likely depends on whether the committee takes the objectively correct action. My dissertation has important implications for both academic economists and policymakers. Understanding the sources of business cycle fluctuations and the determinants of asset prices requires grappling with the fact that people have differences in beliefs. Empirical evidence suggests that agents’ beliefs are shaped by both idiosyncratic forces and by public announcements and policy decisions, and economists’ models need to reflect these features of the world. Policy, too, is affected by the information available to policymakers, and to understand how policymakers have acted in the past and should act in the future, it is necessary to take seriously the ways their belief formation deviates from the full information rational expectations benchmark.

This thesis examines two closely related issues: (1) the ability of imperfect information models to explain some aspects of business cycle dynamics, and (2) the interaction between central bank communications and monetary policy. These issues are related because central bank communications can only be studied in models with imperfect information. In chapter 1, I investigate the ability of a noisy rational expectations model to generate plausible macroeconomic dynamics. The model allows for imperfect, heterogeneous information, and signal extraction from endogenous variables. I find that imperfect information significantly improves the model's ability to generate persistent, hump-shaped responses to a transitory monetary policy shock. This is achieved without the need for mechanical frictions. In addition, the model generates realistic inflation forecast errors. Chapter 2 explores the relationship between central bank statements about future policy and the degree of commitment. I allow the central bank to make (possibly vague) statements about its expected future policy. I begin by assuming that the central bank adopts a loss function which internalizes the bygone costs of deviating from such a pre-announced policy action. The resulting policy is a convex combination of pure discretion and full commitment. As the precision of central bank statements increases, this policy converges to the full commitment policy. I then show that this type of commitment to internalize bygone costs is sustainable only for moderate degrees of precision. Chapter 3 studies the impact of central bank communications about the state of the economy. In particular, I examine the extent to which increased central bank transparency creates a meaningful trade-off between beneficially conveying fundamental information and adversely contaminating observed data with the central bank's opinion. This question is addressed in a variant of the model from chapter 1. In this environment, both the central bank and private agents learn about the state of the economy from observations of endogenous variables. By making the central bank learn from endogenous variables, I am able to study the impact of communications precision on the bank's signal extraction problem.

This thesis examines two closely related issues: (1) the ability of imperfect information models to explain some aspects of business cycle dynamics, and (2) the interaction between central bank communications and monetary policy. These issues are related because central bank communications can only be studied in models with imperfect information. In chapter 1, I investigate the ability of a noisy rational expectations model to generate plausible macroeconomic dynamics. The model allows for imperfect, heterogeneous information, and signal extraction from endogenous variables. I find that imperfect information significantly improves the model's ability to generate persistent, hump-shaped responses to a transitory monetary policy shock. This is achieved without the need for mechanical frictions. In addition, the model generates realistic inflation forecast errors. Chapter 2 explores the relationship between central bank statements about future policy and the degree of commitment. I allow the central bank to make (possibly vague) statements about its expected future policy. I begin by assuming that the central bank adopts a loss function which internalizes the bygone costs of deviating from such a pre-announced policy action. The resulting policy is a convex combination of pure discretion and full commitment. As the precision of central bank statements increases, this policy converges to the full commitment policy. I then show that this type of commitment to internalize bygone costs is sustainable only for moderate degrees of precision. Chapter 3 studies the impact of central bank communications about the state of the economy. In particular, I examine the extent to which increased central bank transparency creates a meaningful trade-off between beneficially conveying fundamental information and adversely contaminating observed data with the central bank's opinion. This question is addressed in a variant of the model from chapter 1. In this environment, both the central bank and private agents learn about the state of the economy from observations of endogenous variables. By making the central bank learn from endogenous variables, I am able to study the impact of communications precision on the bank's signal extraction problem.

This dissertation develops models of charitable giving in the presence of uncertainty. The model of chapter 2 studies a two-stage signaling game of charitable donations with two players: a charity manager and a wealthy donor. A representative charity manager, who is perfectly informed, collects a donation from a representative donor, who has imperfect information about the manager's types. The manager uses the donation to produce a public good, and in the process decides whether to create waste in order to obtain a personal gain. I solve for separating and pooling sequential equilibria of the game, and employ the Intuitive Criterion of Cho & Kreps (1987) as a refinement to deal with the problem of multiple equilibria. I find that there exists no fully separating equilibrium in which the donor can discern all possible manager types. In addition, the results suggest that the amount of the initial donation may help the donor to induce the manager to reveal his true type. In chapter 3, I analyze the effect of competitive pressures in the philanthropic sector. I find evidence in support of market systems acting as a disciplining device, which induces the manager to play strategies that increase social welfare. Chapter 4 uses an alternative to expected utility theory, known as Choquet expected utility, to model the interaction between a wealthy donor and a charity manager in the presence of uncertainty. / Economics

This thesis investigates the use of Bayesian analysis upon an opponent's behaviour in order to determine the desired goals or strategy used by a given adversary. A terrain analysis approach utilising the A* algorithm is investigated, where a probability distribution between discrete behaviours of an opponent relative to a set of possible goals is generated. The Bayesian analysis of agent behaviour accurately determines the intended goal of an opponent agent, even when the opponent's actions are altered randomly. The environment of Poker is introduced and abstracted for ease of analysis. Bayes' theorem is used to generate an effective opponent model, categorizing behaviour according to its similarity with known styles of opponent. The accuracy of Bayes' rule yields a notable improvement in the performance of an agent once an opponent's style is understood. A hybrid of the Bayesian style predictor and a neuroevolutionary approach is shown to lead to effective dynamic play, in comparison to agents that do not use an opponent model. The use of recurrence in evolved networks is also shown to improve the performance and generalizability of an agent in a multiplayer environment. These strategies are then employed in the full-scale environment of Texas Hold'em, where a betting round-based approach proves useful in determining and counteracting an opponent's play. It is shown that the use of opponent models, with the adaptive benefits of neuroevolution aid the performance of an agent, even when the behaviour of an opponent does not necessarily fit within the strict definitions of opponent 'style'.

<p align="justify">Computer Science is currently facing a grand challenge : finding good design practices for embedded systems. Embedded systems are essentially computers interacting with some physical process. You could find one in a braking systems or in a nuclear power plant for example. They present several design difficulties : first they are reactive systems, interacting indefinitely with their environment. Second,they must satisfy real-time constraints specifying when they should respond, and not only how. Finally, their environment is often deeply continuous, presenting complex dynamics. The formal models of choice for specifying such systems are timed and hybrid automata for which model checking is pretty well studied.</p> <p align="justify">In a first part of this thesis, we study a complete design approach, including verification and code generation, for timed automata. We have to define a new semantics for timed automata, the AASAP semantics, that preserves the decidability properties for model checking and at the same time is implementable. Our notion of implementability is completely novel, and relies on the simulation of a semantics that is obviously implementable on a real platform. We wrote tools for the analysis and code generation and exemplify them on a case study about the well known Philips Audio Control Protocol.</p> <p align="justify">In a second part of this thesis, we study the problem of controller synthesis for an environment specified as a hybrid automaton. We give a new solution for discrete controllers having only an imperfect information about the state of the system. In the process, we defined a new algorithm, based on the monotonicity of the controllable predecessors operator, for efficiently finding a controller and we show some promising applications on a classical problem : the universality test for finite automata.

In this thesis we take the Bayesian persuasion model, impose a limitation on the signal space and subsequently derive the first order conditions for the equilibrium. Then we enhance the model for several different constraints. We find that the analysis is greatly facilitated if such constraints do not relax when more signals are added to the model. The first type of constraint is imposed to eliminate the perfectly informative persuasion mechanisms. The second type of constraint limits the total uncertainty reduction caused by the persuasion. Finally the third type of constraint is a participation constraint. It ensures a certain level of reservation utility for the persuaded party. Powered by TCPDF (www.tcpdf.org)

Essa dissertação foi desenvolvida com o objetivo de responder ao seguinte problema de pesquisa: há indícios de que notas explicativas das demonstrações financeiras que sejam menos compreensíveis, na visão de profissionais de investimento, conduzem a custos de transação maiores no estabelecimento de contratos? Por custos de transação, entendem-se aqueles referentes à coleta de informações, os custos de negociação e os de estabelecimento de contratos (COASE, 1991). Uma informação de difícil compreensibilidade, portanto, seria um catalisador de novos custos de transação, trazendo ineficiências no estabelecimento de contratos futuros. Para atingir os objetivos desejados, foram distribuídos questionários aos profissionais de investimento no 22º Congresso APIMEC que foi realizado nos dias 30 a 31 de agosto de 2012. Do total de participantes, 54 pessoas responderam o mesmo. De modo geral, verificou-se que sob a situação de tempo escasso e um grande custo de decodificação das notas explicativas, o que aumenta os custos de transação, pode-se conjecturar que algumas análises estejam sendo feitas de modo ineficiente, o que poderia conduzir a ineficiências no estabelecimento de contratos futuros. Verificou-se também que há indícios de que há excesso de informações e não falta das mesmas, o que conduziria a problemas de informação imperfeita e não de informação incompleta (LOPES, 2008). / This dissertation was developed in order to answer the following research problem: Is there evidence that the explanatory notes in the financial statements that are less understandable, by the investment professionals point of view, lead to higher transaction costs in establishing contracts? For transaction costs, we understood those that are related to information collection, the costs of negotiating and establishing contracts (COASE, 1991). A difficult understandability of information would therefore be a catalyst of new transaction costs, bringing inefficiencies in establishing future contracts. To achieve the desired goals, questionnaires were distributed to investment professionals in 22º Congresso Apimec that was held on 30 to 31 August 2012. Of all participants, 54 people responded it. In general, it was found that under the situation of limited time and a large cost to decode the notes, which increases transaction costs, it can be conjectured that some analyzes are being done inefficiently, which might lead inefficiencies in the establishment of futures contracts. It was also found that there is evidence that there is too much information and not lack of them, which would lead to problems of imperfect information and not incomplete information (LOPES, 2008).