According to theory, market concentration affects the likelihood of a financial crisis in different ways. The “concentration-stability” and the “concentrationfragility” hypotheses suggest opposing effects operating through specific channels. Using data of 160 countries for the period 1970-2007, this paper empirically tests these indirect effects of financial market structure. We set up a simultaneous system in order to jointly estimate financial stability and the relevant channel variables as endogenous variables. Our findings provide support for the assumption of channel effects in general and both the concentrationstability and the concentration-fragility hypothesis in particular. The effects are found to vary between high and low income countries.