Topics in Economic Dynamics and Time Series

研究将在三个不同的主题上完成：首先，该项目开发了一种新的方法来表征弱相关时间序列中的非线性。采用非线性主成分分解来表示连续时间马尔可夫过程的演化过程，并总结出相关结构。这些分解有助于我们更好地理解不同地区相对价格的演变，以及由交易成本或运输成本引起的非线性。本工作借鉴了三个文献：功能主成分分析；马尔可夫过程的二次建模；以及实际汇率的非线性随机模型。还考虑了强依赖或非平稳的非线性模型。其次，研究者继续研究私人代理人和政策制定者面临模型错误规范的经济体。该项目探索并实现了非线性随机均衡模型的求解方法。它研究的货币模型中，宏观经济政策制定者对错误规范的担忧不同于私人代理人的担忧。它建立在对错误规范和过滤的时间序列计量经济学文献的见解基础上，以调查私人代理人决策中鲁棒性的替代公式。最后，研究鲁棒性、自适应学习和基于信息的惯性模型之间的联系。第三，该项目考察了人力资本和证券投资不确定性下动态模型的可观察意义。对时间序列和资产收益横截面的影响进行了推导和检验。本文还考虑了异质性在劳动收入随机结构中的作用。最后，研究了金融市场结构对人力资本影子估值的影响。

Research will be completed on three different topics:First, the project develops a new approach to characterizing nonlinearities in weakly dependent time series. Nonlinear principal component decompositions are used to represent the evolution of continuous-time Markov processes and to summarize the correlation structure. These decompositions help us understand better the evolution of relative prices across locations and the nonlinearities induced by transaction or transport costs. This work draws on three literatures: functional principal component analysis; quadratic-form modeling of Markov processes; and nonlinear, stochastic modeling of real exchange rates. Nonlinear models that are strongly dependent or nonstationary are also considered.Second, the investigator continues his study of economies in which private agents and policy-makers confront model misspecification. The project explores and implements solution methods for nonlinear, stochastic equilibrium models. It studies monetary models in which the misspecification concerns of macroeconomic policy makers differ from those of private agents. It builds from insights in the time-series econometrics literature on misspecification and filtering to investigate alternative formulations of robustness in the decision-making of private agents. Finally, the project studies connections between robustness, adaptive learning and information- based models of inertia.Third, the project examines the observable implications of dynamic models under uncertainty with human capital and portfolio investment. Implications for time series and cross sections of asset returns are deduced and tested. The role of heterogeneity in the stochastic structure of labor income also is considered. Finally, the project investigates the role of financial market structures on the shadow valuation of human capital.