The fundamental paradigm of asset pricing is changing fast. Over time financial economists have grown more accepting of incorporating human fallibility into formal asset pricing models. Today, many economists agree that the central task of asset pricing is to examine how expected returns are related to both risk and investor misvaluation as opposed to risk only. There is accumulating evidence that misvaluation can occur indirectly. In the presence of cross security effects, the risk and misvaluation associated with an asset are channeled not only to its own expected return but to the returns of other unrelated assets. With this in mind the focus of this dissertation is to examine such effects. Results from the dissertation suggest that pessimism associated with the returns on an asset can be channeled into the prices of and returns on an asset whose return distribution is not associated with pessimism. The price of the latter can be higher, lower, or remain unchanged as compared to its price under no pessimism. An empirical investigation of the effect of the return on a broad market index on the return on an index of gold stocks in the period January 1998 to April 2003 reveals that pessimism was directly channeled from the market into the index of gold stocks. Finally, results from the dissertation suggest that the existence of growth stocks affects the prices of and returns on value stocks. In the late 1990's the skyrocketing prices and valuations of the high growth companies formed a striking contrast to the considerably lower prices and valuations of more traditional companies. Analysis conducted with experimental data indicates that the trading prices of a value asset are generally higher in the presence of a growth asset that experiences positive updates in its fundamental value. In addition, lower prices observed in the growth market have a positive effect on the returns on value assets.

This dissertation includes three essays on decision-making by boundedly rational economic agents. The first essay deals with decision-making by firms where decision-making is costly. The other two essays deal with decision-making by individuals. Taken together, the last two argue that decision-making is stochastic, and develop a stochastic model of decision-making reflecting that argument. In particular, the second essay does a detailed review of the psychological literature on estimating probabilities, and argues that the most successful explanation of the psychological results is that people's estimates are stochastic. Decision-making is costly, and so people use various heuristics to economize on the use of mental resources. The results suggest that the choice of heuristic may be partially random. The third essay builds on the second and develops an axiomatic model of choice, in the spirit of Savage (1954), in which individual's beliefs about the probabilities of events are modeled by a stochastic process.

In the first chapter, we consider the logistic regression model with an integrated regressor. In particular, we derive the limit distributions of the nonlinear least squares (NLS) estimators and their t-ratios. It is shown that the NLS estimators are generally not efficient. Moreover, the t-ratios for the level parameters have limit distributions that are non-normal and dependent upon nuisance parameters, due to the asymptotic correlation between the innovations of the regressor and the regression errors. We propose an efficient NLS procedure to deal with the inefficiency of the estimators and inferential difficulty. The new NLS procedure yields estimators that are efficient and have asymptotically normal t-ratios. The second chapter considers a state space model with integrated latent variables. The model provides an effective framework to specify, test and extract common stochastic trends in a set of integrated time series. The standard Kalman filter is used to estimate the model, and the asymptotic theory of the Kalman filter is derived. In particular, we establish the consistency and asymptotic mixed normality of the maximum likelihood estimator, and validate the conventional inference for this class of models. Moreover, we derive a trace statistic to test the number of common stochastic trends in a system of integrated time series. The asymptotic distribution of the trace statistic is derived as normal. Chapter 3 provides an empirical illustration by investigating common stochastic trends of interest rates with different maturities.

The existence of equilibrium is a test of the internal consistency of an economic model. In any model with domestic moneys and dividend-yielding assets, the first question is: will money be dominated by assets in equilibrium? In cash-in-advance models domestic moneys always have positive liquidity value as instruments for domestic commodity transactions. If a non-dominated liquidity role for domestic currencies is posited existence of equilibrium is not usually problematic. For the case where information flows lead to binding cash-in-advance constraints an equilibrium exists in which domestic moneys have positive liquidity value. This equilibrium possesses the unit velocity property, but leads to a sharper characterization of equilibrium market and shadow prices in relation to fundamentals. That fiat money should be the unique provider of liquidity services is not necessary for equilibrium. It is possible to construct models with well-defined equilibria in which financial assets provide liquidity services. In these models the pricing equations for liquidity-providing assets contain premia for these services over and above risk premia and returns for delaying consumption. Such models can also generate new relationships between the velocity of money or the spot exchange rate and asset returns. That markets be information-efficient is, however, necessary for equilibrium. Consequently, any rejection of efficient-markets is evidence against the assumptions of the equilibrium theory. Consider, for instance, the case of the efficiency of forward exchange rates vis-a-vis spot rates. Depending on whether forward speculation is consummated through arbitrage of currencies or of assets (e.g., covered and uncovered bonds) the forward efficiency condition will or will not involve liquidity premia. In testing forward efficiency in both models we find, however, there is no material change in the results. Forward efficiency appears to be robust as well to specifications of the utility function. However there is evidence that forward efficiency is not robust to either the measurement of consumption risk, or the choice of covariance estimator of the forecast error.

Chapter one provides a new estimator for the ordered polychotomous model. The estimator is based on the use of the average of the standard normal densities with different means as a parametric approximation to the density of the error term. The method also, for the first time, provides a consistent, differentiable estimator of the distribution function of the error term. Chapter two employs the conventional interpretation of endogeneity in econometric models to develop a way of eliminating the inconsistency resulting from endogenous explanators in cross sectional models. The method first obtains an estimate of the unobserved heterogeneity responsible for the endogeneity and then creates a synthetic observation by taking a non-parametric weighted average of nearby observations. The deviations are produced from these synthetic means thereby eliminating the unobserved heterogeneity. The procedure is particularly useful for estimating models when the endogenous regressors are censored or appear non-linearly in the primary equation. Chapter three first calculates the exact distribution of Blum et al's (1961) statistic, which is based on a comparison of the sample joint CDF with the product of the sample marginal CDF's, for very small sample size and simulate the distribution quantiles of it for sample size not large enough to employ the asymptotic result. Secondly, the asymptotic distribution of the statistic constructed from residuals and/or predicted values, to test the independence of the error term and the regressors in nonlinear regression models, is obtained. Thirdly, bootstrap technique is used to obtain the distribution quantiles of the statistic constructed from residuals and/or predicted values. The test is nonparametric in that it does not specify the parametric form of distributions of the error term and the regressors.

We lay a unified foundation for a theory of general equilibrium by proving the existence of an equilibrium for a grand model which covers all the well-known general equilibrium models under the convexity and continuity assumptions. The grand model allows an economy to have an extended list of commodities including assets which can be traded on unlimited short sales. The conceptual framework we develop for the existence problem is simple. Consider an economy consisting of two agents. If there were a commodity bundle which is always desirable to one agent and always undesirable to the other agent, the economy could not reach an equilibrium because they can increase their utility through an indefinite give-and-take process. What we need for the existence of an equilibrium is to exclude the presence of commodity bundles that can bring an economy into this state of "economic symbiosis." We proceed further by taking the Closedness Hypothesis that the utility possibility set is compact. The finite dimensional findings do not hold for an economy with an infinite dimensional commodity space so that we investigate under what circumstances the Closedness Hypothesis holds. We develop sufficient conditions for the Closedness Hypothesis to hold and prove the existence of an equilibrium of an infinite dimensional economy under some spanning conditions on consumption sets.

In the first chapter; we consider nonlinear transformations of random walks driven by thick-tailed innovations with infinite means or variances. In particular, we show how nonlinearity, nonstationarity, and thick tails interact to generate persistency in memory, and we clearly demonstrate that this triad may generate a broad spectrum of persistency patterns. Time series generated by nonlinear transformations of random walks with thick-tailed innovations have asymptotic autocorrelations that decay very slowly as the number of lags increases or do not even decay at all and remain constant at all lags. We also discuss nonlinear regression asymptotics when the regressor is observable and an alternative regression technique when it is unobservable. We use our model to analyze two empirical applications: exchange rates governed by a target zone and electricity price spikes driven by capacity shortfalls. In the second chapter of the dissertation, we show that typical tests for purchasing power parity using target zone exchange rates are inherently misspecified. The real exchange rate cannot be mean-reverting by construction, since the nominal exchange rate is generated by a nonlinear transformation of a nonstationary economic fundamental. As an alternative, we propose basing the real exchange rate (and thus the test) on conditional expectations of the fundamental itself. Real exchange rates based on the true fundamental may in fact exhibit mean reversion, if the long-run purchasing power parity hypothesis holds. The final chapter investigates the statistical properties of the Kalman filter for state space models including integrated time series. In particular, we derive the full asymptotics of maximum likelihood estimation for a prototypical class of such models: those with a single latent common stochastic trend. We demonstrate the utility of the state space model by extracting a common stochastic trend in three empirical analyses: interest rates, stock return volatility and trading volume, and a stock price index.

This dissertation develops efficient semiparametric estimation of parameters and expectations in dynamic nonlinear systems and analyzes the role of environmental factors in productivity growth accounting. The first essay considers the estimation of a general class of dynamic nonlinear systems. The semiparametric efficiency bound and efficient score are established for the problems. Using an M-estimator based on the efficient score, the feasible form of the semiparametric efficient estimators is worked out for several explicit assumptions regarding the degree of dependence between the predetermined variables and the disturbances of the model. Using this result, the second essay develops semiparametric estimation of the expectation of known functions of observable variables and unknown parameters in the class of dynamic nonlinear models. The semiparametric efficiency bound for this problem is established and an estimator that achieves the bound is worked out for two explicit assumptions. For the assumption of independence, the residual-based predictors proposed by Brown and Mariano (1989) are shown to be semiparametric efficient. Under unconditional mean zero assumption, I proposed an improved heteroskedastic autocorrelation consistent estimator. The third essay explores the directional distance function method to analyze productivity growth. The method explicitly evaluates the role of undesirable outputs of the economy, such as carbon dioxide and other green-house gases, have on the frontier production process which we specify as a piecewise linear and convex boundary function. We decompose productivity growth into efficiency change (catching up) and technology change (innovation). We test the statistical significance of the estimates using recently developed bootstrap method. We also explore implications for growth of total factor productivity in the OECD and Asia economies.

This dissertation consists of three essays on the taxation of capital income. The first essay examines the effects of the interactions of home and host country tax provisions on capital investment decisions by a U.S. multinational before and after a capital-importing country switches from a conventional corporate income tax to a consumption-based business cash flow tax. The analysis considers the cases in which the U.S. deems the cash flow tax to be creditable, non-creditable or partially creditable. In addition, two methods of implementing a consumption-based business tax are considered--the R-base and the R+F-base--and the importance of the firm's choices between local debt finance and parent multinational finance is analyzed. A numerical application considers the case in which a U.S. multinational invests in Thailand. The second essay examines the incidence of a general and a classified property tax, using an n-community, two-sector, three-factor general equilibrium model in which residents are assumed to benefit from property taxation. The government distributes residential tax revenues so that a worker, a capitalist or a landlord receives benefits equal to the amount of housing tax paid, and distributes commercial property tax revenues such that either each resident receives equal benefits, or only workers receive equal benefits. Aggregate welfare depends on labor population and most of the tax burden is borne by the individuals whose resources are taxed. The third essay provides a review of the literature on the optimal taxation of capital income in a small open economy setting and of the econometric evidence that examines the effects of taxation on business location decisions. Theoretical analyses generally suggest that the optimal tax on capital income is zero, although there are some exceptions. Various explanations of why capital may not be perfectly mobile are also discussed. In the econometric review, it is argued that if the tax has a negative effect on business location measures, the small open economy assumption is to some extent a valid proposition. The essay argues that this assumption is a valid one.

During 1994 domestic and foreign investors in Mexico increased the share of TESOBONOS in their portfolios when they perceived the possibility of a future devaluation of the Mexican peso or, in other words, the abandonment of the controlled floating exchange rate regime. This work finds that both domestic and foreign investors responded to monetary policies followed by Banco de Mexico after March 1994, when adverse political events occurred, keeping their investment in Mexico in TESOBONOS rather than leaving the country. Domestic and foreign investors did not have a high expected probability of devaluation, however, they were certain that if a devaluation was going to happen the size of it would be approximately a hundred percent.