How does the implementation of a Two-Tier Tobin tax affect the reproduction of key stylized facts in Agent-Based financial market Models?

Financial Transaction Taxes (FTTs) have long been proposed as instruments to reduce excessive speculation and enhance financial market stability. However, the existing literature provides mixed evidence regarding their effectiveness, with some studies highlighting stabilizing effects and others emphasizing potential reductions in liquidity and increases in volatility. In this context, Agent-Based Models (ABMs) offer a valuable framework for investigating the impact of regulatory interventions in complex financial systems characterized by heterogeneous interacting agents. This thesis examines the effects of a Two-Tier Tobin Tax within an extended version of the Franke–Westerhoff heterogeneous agent model. The model incorporates dynamic role assignment among chartists, fundamentalists, and inactive traders, as well as stochastic noise trading, allowing the endogenous generation of key financial market stylized facts such as fat-tailed return distributions, volatility clustering, and the absence of linear autocorrelation in returns. The proposed tax mechanism combines a low baseline transaction tax with a higher tax rate that is activated whenever market volatility exceeds a predefined threshold. A grid of simulation experiments is conducted by varying both the volatility threshold and the high-tax rate. The analysis focuses on the frequency of tax activation, fiscal revenues, and the preservation of fundamental statistical properties of financial returns. The results reveal a strong trade-off between market stabilization and market activity. Very low volatility thresholds lead to frequent tax activation, reducing extreme price fluctuations and volatility clustering but also risking excessive suppression of trading activity. Conversely, very high thresholds render the tax largely ineffective. Intermediate parameter configurations emerge as the most balanced solution, generating meaningful and relatively stable fiscal revenues while moderating fat tails and volatility persistence without significantly altering the underlying market dynamics. The findings suggest that the effectiveness of a Two-Tier Tobin Tax depends critically on its calibration. Rather than extreme parameter settings, selectively activated tax regimes appear to provide the most promising compromise between financial stability, revenue generation, and market efficiency. More broadly, the study demonstrates the usefulness of agent-based models as virtual laboratories for evaluating financial market regulation and exploring the unintended consequences of policy interventions.