Credit Risk: What Are We Actually Measuring?

Credit risk modeling for financial institutions focuses on measuring potential losses from lending activities. The core of this measurement involves estimating three key parameters: the Probability of Default (PD), the Loss Given Default (LGD), and the Exposure at Default (EAD). These components are multiplied together to calculate the expected loss on an individual loan or an entire portfolio. Banks use these risk estimates not only for internal management but also to determine the amount of capital they must hold to absorb unexpected losses, a process governed by regulatory standards like the Basel framework. The development of these models is a data-driven task, requiring statistical analysis of historical data to predict future outcomes.