In finance systemic risk often comes about as the unintended consequences of the attempts of individual institutions to reduce risk. I will present two models that illustrate this point. One is a model of leveraged value investing funds, and the other is a network model of leveraged funds (such as banks) who are connected by holding common assets. Both of these models demonstrate how systemic risk comes about through nonlinear amplification mechanisms that emerge as an unintended side-effect of otherwise reasonable individual behaviors -- indeed such behaviors are often intended to reduce risk. This problem is exacerbated in evolutionary settings where individual agents are all changing their behavior, and are thus unable to extrapolate the behavior of others as they evolve. I will argue that systemic risk in finance is often driven by disruptions in market ecology, which are related to deviations from market efficiency, and thus are not well addressed by traditional economic theory, which assumes such inefficiencies away.