Risk is endogenous when an individual is able to undertake mitigation or self protection actions that reduce the risk that he faces. Most risky environments studied in economics involve endogenous risk. This dissertation studies the conceptual and behavioral implications of introducing endogeneity in the controlled environment of the laboratory. The dissertation consists of three different experiments designed to examine how endogeneity affects risk attitudes and risk perceptions in simple experimental set ups. All three experiments employ a virtual reality scenario where the subject is able to form his own beliefs, based on naturalistic cues provided by the virtual reality experience. In the first experiment, a 'short run' individual experiment, subjects experience several forest fires that allow them to form beliefs about the probability of a house in the simulated forest being destroyed by fire. The evidence suggests that endogenous risk settings do cause subjects to employ different subjective beliefs than they use in an exogenous risk setting, although risk attitudes appear stable across these settings. Typically, the risk of natural disaster in any area is very small, and an adverse event like a forest fire occurs only once in a couple of decades. This has implications for self-protection expenditure where risk is endogenous. A 'long run' individual experiment with several rounds of decision making allows the estimation of subjective beliefs about the risk of the property burning when a fire may occur. This design allows for the study of the effect of an actual experience of forest fire on a subject's beliefs. Several mitigation options are collective in nature and require group contributions for the self-protection action to be provided. In an extension of the long run design, we study the effect of an actual experience of fire on beliefs when the risk is faced by a group rather than an individual. This framework also allows us to compare behavior in a public goods game involving risk, with the standard public goods game.

The main focus of this thesis was to gain a better understanding about the dynamics of risk perception and its influence on people’s evacuation behavior. Another major focus was to improve our knowledge regarding geo-spatial and temporal variations of risk perception and hurricane evacuation behavior. A longitudinal dataset of more than eight hundred households were collected following two major hurricane events, Ivan and Katrina. The longitudinal survey data was geocoded and a geo-spatial database was integrated to it. The geospatial database was composed of distance, elevation and hazard parameters with respect to the respondent’s household location. A set of Bivariate Probit (BP) model suggests that geospatial variables have had significant influences in explaining hurricane risk perception and evacuation behavior during both hurricanes. The findings also indicated that people made their evacuation decision in coherence with their risk perception. In addition, people updated their hurricane evacuation decision in a subsequent similar event.

Nous étudions des stratégies d'investissement dont l'utilisation s'est généralisée sur les marchés financiers, et leur impact sur le prix des actifs et le risque de marché.Dans le premier chapitre nous nous intéressons aux stratégies de diversification de portefeuille. Nous montronsau travers d'un modèle théorique que si la diversification a un effet positif au niveau individuel pour l'investisseur,elle crée également des liens entre les différents investisseurs et titres, qui peuvent se révéler dangereux d'un pointde vue systémique. Nous mesurons les deux effets afin de discuter de la désirabilité globale de la diversification.Le second chapitre considère les stratégies d'investissement basées sur le groupement de titres financierspartageant certaines caractéristiques en différentes classes, ou styles. Nous postulons que ces stratégies créentun co-mouvement excessif entre titres d'un même style, qui seront vendus et achetés ensemble au sein d'une mêmeclasse. Appliquant cette intuition aux notes des agences sur les obligations, nous montrons qu'une obligation quichange de note se met en effet à varier comme sa nouvelle note, même quand les fondamentaux économiques ne lejustifient pas.Dans le troisième chapitre nous étudions trois types d'investisseurs opérant sur le marché des changes : les carry traders, les chartistes et les fondamentalistes. Notre modèle théorique suggère que l'interaction entre cestrois règles d'investissement peut expliquer la déconnexion bien documentée entre le taux de change et sa valeurfondamentale, ainsi que provoquer un effondrement endogène des taux de change. / We study how popular investment rules in financial markets may induce endogenous movements inasset prices, leading to higher market risk.In the first chapter, we focus on portfolio diversification. We show through a theoretical model that this strategyis beneficial at the individual investor level, but also creates endogenous links between assets and investors, whichcan be dangerous from a systemic perspective. We measure both effects in order to discuss the overall desirabilityof diversification.The second chapter considers strategies based on grouping assets that share common characteristics intodifferent classes, or styles. We postulate that these strategies create excess comovement between assets of asimilar style, as they are traded together as part of the same class. Applying this reasoning to bond credit ratings,we show that bonds joining a new rating class indeed start comoving more with the bonds of this rating, evenwhen fundamental factors suggest otherwise.In the third chapter, we study three investors who operate in the foreign exchange market: carry traders,chartists and fundamentalists. We provide a theoretical model which suggests that the interaction between thesetrading rules may explain the well documented exchange rate disconnect from its fundamental value, and lead toendogenous currency crashes.

The architecture, engineering and construction industry faces challenges when dealing withprocurement contract design and risk-handling. The optimal contracting practices have beenworldwide studied in areas of contract theory which studies how the optimal incentivemechanism (“contracts”) can be designed to encourage the parties to behave more efficiently.The parties usually consist of a principal and an agent, where the principal hires an agent todeliver goods or services. However, the complexity of contract theory calls for a morepracticable approach in an attempt to understand the procurement problem in the industry andincrease knowledge-sharing between projects. The purpose of this study is to propose a model based on contract theory that can be used inpractice to investigate the effects of project endogenous risks in three different types ofprocurement contracts: fixed-price, time and material, and incentive. Thus, this study usesquantitative methods with the aim to explain the current procurement problem in the Swedisharchitecture, engineering, and construction industry, compare theory and practice, andcontribute to knowledge about the linkage between endogenous risks, optimal risk sharing andcontract design. The conclusions from this study are that the current contracting practices in the industry arenot aligned with the optimal contract design described by the theory. The theory in this researchshowed that, given endogenous project risks, the optimal incentives vary in the agent’saversion to risk resulting in a non-monotone relationship between optimal contract power andproject risk. Further, a contract becomes optimal and efficient when cost savings and qualityincentives are aligned. However, the analysis of real-world projects presented no clearrelationship between contract power and project risk, i.e. some projects with fixed-pricecontracts or time and material contracts presented the same risk level. Hence, this researchproposes a method for computing the optimal incentive contract which can be used in manycases where the other two types of contract are currently being used. Based on the theory, theoptimal incentive contract may add valuable benefits for both parties involved since it aims toefficiently share the project risk between them while providing the agent the right incentivesto work more efficiently to reduce costs and deliver high-quality services or goods. / Byggbranschen står inför utmaningar när det gäller kontraktsdesign och riskhantering. Deoptimala upphandlingsmetoderna har studerats över hela världen inom områden avkontraktsteori som i sin tur studerar hur den optimala incitamentsmekanismen ("kontrakt") kanutformas för att uppmuntra parterna att agera mer effektivt. Parterna består vanligtvis av enprincipal och en agent, där principalen anställer en agent för att leverera varor eller tjänster.Men komplexiteten i kontraktsteori kräver ett mer praktiskt tillvägagångssätt i ett försök attbättre förstå upphandlingsproblemet i byggbranschen samt att öka kunskapsutbytet mellanprojekt. Syftet med denna studie är att föreslå en modell baserad på kontraktsteori som kan användas ipraktiken för att undersöka effekter av endogena risker i tre olika typer avupphandlingskontrakt: fastpris, rörligt pris (time and material) och incitament. Denna studieanvänder därmed kvantitativa metoder i syfte att förklara det aktuella upphandlingsproblemeti den svenska byggbranschen, jämföra teori och praktik, och bidra till utökad kunskap omsambandet mellan endogena risker, optimal riskdelning och kontraktsdesign. Slutsatsen från denna studie är att den nuvarande upphandlingspraxisen i branschen inte är ilinje med den optimala kontraktsdesignen som beskrivs av teorin. Teorin i denna studie visaratt, förutsatt endogena projektrisker, så varierar de optimala incitamenten med agentensriskaversion vilket resulterar i ett icke-monoton förhållande mellan optimal kontraktsdesignoch projektrisk. Dessutom blir ett kontrakt optimalt och effektivt när kostnadsbesparingar ochkvalitetsincitament är i linje med varandra. Men analysen av verkliga projekt visar inget tydligtsamband mellan kontraktsdesign och projektrisk, det vill säga en del projekt med fastpris ellerrörligt pris visade sig ha samma risknivå. Därför föreslår denna studie en metod för beräkningav det optimala incitamentet som kan användas i många fall där de andra två typerna avkontrakt för närvarande används. Teorin säger att det optimala incitamentet kan ge värdefullafördelar för de inblandade parterna eftersom det syftar till att på ett effektivt sätt fördelaprojektrisken mellan dem samtidigt som agenten får rätt incitament att arbeta mer effektivt föratt sänka kostnaderna och leverera högkvalitativa tjänster eller varor.