• agent-based model
• complexity theory
• computational economics
• finance
• heterogeneous expectations
• heuristics
• macro-finance
• macroeconomics
• quantitative finance

This thesis extends the Keynes meets Schumpeter (KS) model by introducing a financial market for government bonds. The financial market is shaped through an Agent-based Computational Finance (ACF) model in which agents are provided with empirically-based heuristics to set expectations. In the financial market, agents take an endogenous macroeconomic signal as fundamental, while the market price of the bond enters into the balance sheets of banks via fair value accounting. Our focus is studying how various trading heuristics impact both price volatility and the macroeconomy. Our model is highly stylized and we find that different specifications of heuristics - various market sentiments - lead to different price volatilities, deviations from the fundamental and outliers. Moreover, prices are mostly endogenous formed by the interaction of agents, and bubbles are a self-fulfilling phenomenon originating from the herd effect in extrapolative heuristics. Also, simulation results are in line with less-is-more. Finally, the introduction of the financial market impacts the macroeconomy by increasing the volatility of banks' balance sheets and, depending on the heuristics employed, reduces credit supply and thus GDP.