Browsing by Subject "Ambiguity"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Information Acquisition and Revelation in the Financial Markets
    (2018-06) Chu, Yinxiao
    Information plays an important role in financial markets. In this dissertation, first, we consider how traders choose different information. Second, we ask when traders acquire information under competition. Finally, we analyze how ambiguous information affects traders' incentives to trade and reveal their private information. There is information not only about the payoff but also concerning the supply and demand of an asset. In Chapter 1, we study how traders choose to process different information while asset prices are conveying some information. We show that traders decide to process different types of information depends on their initial belief and the informativeness of asset prices. In particular, when the return to each type of information is increasing, traders choose to learn only one type of information. Those who have more precise initial belief about the asset payoff (supply) choose to learn more about the asset payoff (supply). In Chapter 2, we study when traders decide to acquire information under competition. Traders consider two effects of competition in information acquisition: one is that an informed trader's profitability is affected by the presence of another informed trader, the other is the spillover of the information from the informed trader to the uninformed. We show that, when the former effect dominates, then traders tend to acquire information earlier. If the otherwise, then traders tend to delay their information acquisition. In Chapter 3, we study traders' behavior when information is ambiguous, which gives rise to multiple probability models to describe uncertainty. We demonstrate that ambiguity will reduce traders' incentive to trade and reveal their private information. When there is a moderate level of ambiguity, informed traders start to trade randomly, whereas they trade for sure when there is no or a little uncertainty. When ambiguity is sufficiently large, informed traders choose not to trade any more, and no additional information will be revealed in the market.
  • Thumbnail Image
    Item
    Robustness and Games with Linear Best Replies: Theory and Applications
    (2019-08) Marku, Keler
    This thesis consists of three chapters on games with linear best replies. In the first chapter we show how in the context of a common agency game, when principals seek robustness, then linearity in total output emerges as an equilibrium outcome. More specifically we consider a game between several principals and a common agent, where principals design contracts that are robust to misspecification of the agent's technology. The principals know a subset of the actions available to the agent, but other unknown actions could exist. Principals demand robustness and evaluate contracts on the worst-case performance over all possible actions of the agent. Despite the complexity of the game, we show that a pure strategy equilibrium always exists, by constructing a pseudo-potential for the game. Equilibrium contracts are linear in total output and imply that all players (the principals and the agent) receive a share of total output. The higher the share of total output accruing to the agent, the more efficient the outcome of the game. The crisp characterization of the equilibrium allows us to revisit the classical question of the efficiency of competitive outcomes relative to collusion among principals. We also consider a game where principals collude and offer a joint contract. The efficiency of the competitive outcome depends crucially on the ability of principals to offer side-payments to one another through the agent. In the second chapter we consider several applications of the framework introduced in the first chapter. Linearity allows for sharp predictions of the model in several contexts. The main application of the model in the first chapter is in analyzing the taxation of multinational firms where we study the effects of tax competition among countries. We show that a flat tax on domestic and foreign profits with a full deduction of foreign taxes provides the best worst-case guarantee for each country's revenues. Furthermore we consider a procurement auction setup, as well as an application of the model to private provision of public goods. In the third chapter we depart from the robustness framework and focus on the network structure of games with linear best replies. Games played on fixed networks capture a variety of economic settings including public goods, peer effects, and technology adaption. Bramoulle et al. (2014) analyze a large class of one dimensional linear best reply games and provides general results on how the network affect social and economic outcomes. In this paper we first provide an isomorphism between games with linear best replies and the threshold-linear recurrent networks used in neuroscience to study the encoding of memory patterns in the brain, connecting two seemingly unrelated literatures. Inspired by the isomorphism we extend games of linear best replies in understanding Lindahl equilibria and to games with multidimensional actions. In particular we show how Cournot competition among several firms leads to specialization in production. We show how the network structure of competition in demand for consumers shape the decision of firms in which goods to specialize.
  • Thumbnail Image
    Item
    School Psychologists’ Decision Making in Evaluations for Emotional Disturbance
    (2016-07) Sadeh, Shanna
    For decades, there has been a persistent national trend of public schools disproportionately qualifying more Black students relative to White students for special education under the category of serious emotional disturbance (ED). Such disproportionality suggests but does not prove racial bias in ED evaluations. I experimentally tested how much, if at all, school psychologists’ racial bias impacted eligibility determinations using a vignette methodology and between-group design with three conditions that varied by level of data ambiguity: (a) low-ambiguity data that do not meet ED criteria; (b) low-ambiguity data that meet ED criteria; and (c) highly ambiguous data. The hypothetical student in each vignette was a fifth grade male who had primarily externalizing problems. Participants completed one vignette in each ambiguity condition; student race (Black versus White) was experimentally manipulated. Participants were 60 practicing school psychologists in a northeastern state that adopted the federal regulations for ED eligibility. For each vignette, participants decided whether the student qualified as ED, rated their confidence in their decision and the diagnosticity of data included in the evaluation, and had the option to describe additional data they wish had been included in the results. Chi-square analyses indicated there were no statistically significant differences based on race between students qualified and disqualified as ED across ambiguity conditions, providing some evidence against the racial bias theory of disproportionality. Under the highly ambiguous data condition, there was no statistically significant difference between students qualified as ED and those not qualified – i.e., regardless of race, all students had a coin-toss chance of qualifying as ED. This finding makes sense in light of the numerous ambiguous key terms in the ED criteria, which allow for more than one reasonable interpretation. Results also showed that most school psychologists were at least moderately confident in their determinations across ambiguity conditions. Their confidence in the low-ambiguity conditions makes sense because those vignettes were designed to be relatively easy. Their confidence in the highly ambiguous data condition may illustrate the potency and frequency of confirmation bias in decision making under conditions of high uncertainty. Across ambiguity conditions, participants frequently identified behavior rating scales and infrequently identified achievement and intelligence scores as highly diagnostic. They identified interviews, family information, and observations with varying frequency across conditions, demonstrating that the diagnosticity of data can fluctuate depending on the presenting problems and evaluation results. Finally, school psychologists who opted to describe additional data they wish had been included in the evaluation results primarily requested more information about interventions that had been attempted and consultation with outside mental health providers. Implications for practice and further research opportunities are discussed.
  • Thumbnail Image
    Item
    Three essays on decision-making under uncertainty in developing countries.
    (2012-08) Lim, Sung Soo
    People face substantial risk and uncertainty throughout most developing countries. Economists have provided an array of economic theories to explain how decision agents make choices under uncertainty. Decision theory distinguishes between risky prospects and uncertain prospects. Under the classical theory of decision under risk, the utility of each outcome is weighted by its probability of occurrence. Expected utility theory was developed to explain attitudes toward risk, namely risk aversion and risk loving. Experimental studies of decision under risk have shown that people often violate the expected utility model. This study has been prompted by two questions about decision making under uncertainty: (1) Do agents in developing countries conceptualize the uncertainty in a way that is consistent with expected utility theory? (2) How do agents cope with uncertainty that lasts a long period of time? The first question is motivated by a notion that studies in the field of development economics often confound the two concepts of risk and uncertainty under the paradigm of expected utility theory. The second question is motivated by another notion, which is that there are a limited number of studies that investigate coping strategies of households during a prolonged period of uncertainty, such as morbidity shocks. To address these two questions, the first essay revisits the neoclassical theory of migration. The second and the third essays examine the impact of prime-age adult morbidity on intrahousehold resource allocation.