In modelling loss given default,(LGD), we often encounter the term Margin of Conservatism. What is it in layman's terms? I am not able to find a wikipedia page on this.


If you are looking for a rigorous mathematical definition, there isn't one.

A margin of conservatism is broadly defined here to be the additional amount in model estimates relative to actual outcomes. This definition will differ depending on the model in question; where for some it may be interpreted as low threshold above some metric of accuracy, while others may require a much higher threshold. Certainly, conservatism may be considered as a back-testing objective, whose definition may be moulded to the purpose of the model being tested.

For example, when calculating LGD for economic capital, estimates might be low enough to be an accurate representation of the portfolio, yet on the side of conservatism to ensure underestimation is kept to a minimum . However, stressed LGD estimates should be inherently inflated to provide a cushion during economic downturn periods.

Institutions incorporate margins of conservatism in their risk parameter estimates to capture certain modelling deficiencies, uncertainties, or risks. The quantification of such ranges vastly among institutions and models.

  • $\begingroup$ Thanks. So in the end, it is a somewhat random number, this MOC? $\endgroup$ – Victor123 Feb 7 '15 at 15:43
  • $\begingroup$ Hypothetically, it could be. However, if you ever inherit the task of quantifying or ensuring a model has a sufficient margin of conservatism, the "additional amount" should be intuitively justifiable. As an example, if you build a simple OLS regression model, the predictions will not likely have a margin of conservatism because OLS minimizes the sum of squared residuals. In this case, you may want to apply a multiplicative or additive factor to the predictions to account for sources of model risk or uncertainty, such as quality of data used to run the regression. $\endgroup$ – dmanuge Feb 9 '15 at 15:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.