关键词：
Agent-based modeling
Incomplete information
Latin Monetary Union
Production networks
Rational expectation
Rigidities
摘要：
Chapter 1 studies the impact of the Latin Monetary Union (1865-1927) on financial flows between European Countries. Using Machine Learning methods, we find that this monetary union fostered greater financial flows between members, especially before 1885 and except for France. Chapter 2 studies a production network model with quantity rigidities and informational frictions, where firms may be restricted in how effectively they can adjust their intermediate input quantities in response to changes in the economic environment and they need to choose their quantities under incomplete information about the realizations of shocks. The characterization results show that these two frictions lead to a reduction in aggregate output, as firms may find it optimal to rely more heavily on less volatile suppliers, even if it comes at the cost of forgoing more efficient ones. Chapter 3 builds on chapter 2 by relaxing the rational expectation assumption. It studies an agent-based model with quantity, price rigidities and decentralized labor and goods markets with search frictions. Firms make production decisions facing unquantifiable uncertainty about demand and unable to coordinate spontaneously on a Nash equilibrium. Inductive firms optimize over a set of forecasting strategies to best predict demand. Naive firms simply extrapolate from last period. The share of inductive firms is a key parameter that dramatically affects business cycle characteristics. A larger share decreases the persistence and volatility of GDP and unemployment, making the process that firms try to forecast memory-less and harder to predict. This leads to the paradox that firms trying to forecast demand individually end up making larger forecast errors in the aggregate. It also qualitatively affects the passthrough from output shocks to prices and wages.
欢迎来到三峡大学图书馆！
详细信息
