| Summary: | This thesis consists of three self-contained chapters. The first chapter develops a method to explore the causal transmission of a catalyst variable through two endogenous variables of interest. The method is based on the reduced-form system formed from the conditional distribution of two endogenous variables given the catalyst and combines elements from instrumental variable analysis and recursive identification of structural vector autoregressions. Conditions for uniqueness of the causal transmission are provided. The second chapter identifies and characterises periods of financial fragility over the period 1973-2016 within a simple regime-switching framework. Fragility is driven by high default risk and low profitability of intermediaries, as defined in Tsomocos (2003). Recessions occur primarily when fragility is high, but high fragility can occur without leading to recession. Credit growth is lower under high fragility, but it is not as significant a driver of regime changes when financial fragility is directly taken into account. Shocks have regime-dependent effects that are attributable primarily to switching dynamics as opposed to shock size. Finally, shocks are amplified through financial fragility, suggesting the presence of a debt-deflation channel. The third chapter assesses the linkages between the real economy and the financial intermediation sector over the period 1999-2015 within a time-varying parameter framework with stochastic volatility. Low volatility and a period of protracted economic growth initiate a feedback loop between the real economy and the financial sector, consistent with the financial instability hypothesis of Minsky (1986). Shocks that adversely affect expectations of future economic growth have the greatest impact when macro-financial linkages are strongest. Augmenting the model with macroprudential capital requirements and monetary policy adjusted for the zero lower bound, we provide evidence that the timing of policy interventions matters. Counter-cyclical interventions produce a stabilising effect when implemented early but are also capable of exacerbating debt-deflation if employed too late in the financial cycle.
|
|---|
