Skip to content
All library documents

Core Topics for Bank Risk Modeling and Model Validation

Article Quant Q&A · Author: P. White
Read the original

Summary

The document outlines topics that may help someone entering a bank internship focused on evaluating or building risk models. It distinguishes market risk from credit risk and points to different regulatory and analytical foundations for each. For market risk, it names value-at-risk study, time series, and regulatory frameworks. For credit risk, it mentions Basel requirements, internal ratings models, logistic regression, and probability-of-default or scorecard models.

For model evaluation, the answers suggest monitoring measures such as AUC, KS or Gini, and population stability, alongside checks of data preparation and methodology. They also describe governance practices such as challenger models, comparison with incumbent models, and stress testing. SAS, R, and Python are listed as possible tools, with the caveat that software use varies by firm. The post is an informal orientation rather than a detailed curriculum; it does not prioritize topics or explain how to calculate or interpret the named measures.

Key ideas

  • Market risk preparation can include value-at-risk, time-series analysis, and relevant regulatory frameworks.
  • Credit risk work may involve logistic regression, probability-of-default models, and scorecards.
  • Model monitoring can use discrimination and population-stability measures.
  • Validation may cover data, methodology, challenger comparisons, and stress testing.
  • Software preferences vary across banks and teams.

Tags

Risk managementStatisticsMachine learning
Full text
# What knowledge should I have for a risk management and risk modeling job at a bank?


# What knowledge should I have for a risk management and risk modeling job at a bank?












I have a masters degree in Econometrics and a undergraduate in Finance I am going soon to start an internship at a bank where I will evaluate risk models and possibly possibly create them aswell and I would like to to know what is the common knowledge to have in this area and what I should know about mathematics and software.

Thank you in advance.

## Answer by AlRacoon (score 4)

https://quant.stackexchange.com/a/37418

I would start with reading Jorion's book on VAR. A statistical look at a range of returns seems to be the prevailing methodology of looking at risk these days. Of course the work has advanced significantly since Jorion wrote his first edition. The curriculum for the FRM presented by GARP is also a good start.

## Answer by June (score 3)

https://quant.stackexchange.com/a/37426

What Model will you be working on Market Risk or Credit Risk? As suggested above FRM1 will have part of Credit Risk.

For Market Risk

Guidelines of FRTB; Basel 3; TimeSeries;

For Credit Risk

BASEL 2 (IRB Models) BASEL 3 IRFS and CCL Logistic Regressions/Regressions

For evaluating credit Risks Logit models such as PD or Scorecard: Monitoring:

AUC, KS/Gini; Population Stability Index(PSI) Logs odd

Validation/Governance

Validating Data preparation; Validating Methodology; Creation of Challenger Model; Comparision with Champion model; Stress Testing

Software:

SAS, R, Python *My POV, every firm has its own reservation on software Usage

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.