In this talk I will discuss a clean and tractable framework for capturing systemic risk within a dynamic structural model of solvency. The key theoretical objective is to have an economically meaningful formulation, which has a low parameter space for the dynamics, yet is rich enough to analyse the effect of common factors vis-a-vis feedback from defaults, as captured by just two distinct parameters. In terms of actually using the model, the goal is to rely on market data for making short-term predictions about systemic risk, either by looking at stock market and balance sheet data for a set of banks or by looking at the spreads for a CDS index on a set of companies. Here, I will focus on the latter, outlining an efficient numerical algorithm, highlighting some advantageous features of the model, and discussing calibration of the model parameters to iTraxx and CDX data. Part of this remains ongoing work, so the discussion of calibration will focus on preliminary observations along with a road-map for how to do out-of-sample predictions within this framework.