Quantitative Model Validation Specialist (m/f)
Česká republika : Hlavní město Praha
About the role
Join ČSOB’s Model Validation team in Prague to apply statistics, quantitative analysis and independent testing to risk models, primarily in credit risk. You will assess model assumptions, reproduce calculations and develop independent benchmarks, with opportunities to work on other quantitative and statistical risk models.
At ČSOB, these models support credit decisioning, regulatory capital calculation and impairment measurement under IFRS 9, as well as risk assessment in other areas. They combine statistical techniques, portfolio-level estimation and machine-learning approaches.
Our team is growing. We are looking for a quantitative professional who enjoys working with data, investigating statistical relationships and testing whether model assumptions and estimates hold up against empirical evidence.
Your work will primarily focus on credit risk models, including Probability of Default (PD), Loss Given Default (LGD), Exposure at Default (EAD), application and behavioural scorecards, and Expected Credit Loss (ECL) models. Depending on team priorities and assignments, you may also work on models in other risk areas. Rather than taking model outputs at face value, you will investigate the sources of uncertainty and potential bias.
You will collaborate with specialists across ČSOB and the wider KBC Group. We welcome candidates with a solid quantitative background who are looking to move into model validation. No prior experience in risk modelling, model validation or banking is required; structured onboarding and mentoring will help you build the relevant domain expertise.
What you will do
- Perform end-to-end model validations, covering data, methodology, parameter estimation, performance, stability and implementation.
- Assess statistical assumptions, estimation methods and the robustness of model results across samples, time periods and portfolio segments.
- Reproduce model calculations and develop independent quantitative benchmarks to challenge modelling choices and identify potential weaknesses.
- Apply statistical testing, backtesting, sensitivity analysis and resampling techniques to assess model performance and estimation uncertainty.
- Investigate data quality, sample representativeness and their implications for model estimates and conclusions.
- Review statistical and machine-learning models, understanding their assumptions, limitations and key risk drivers.
- Translate quantitative results into clear, evidence-based findings and practical recommendations.
- Assess alignment with applicable regulatory requirements and internal ČSOB/KBC standards.
- Discuss your conclusions with model developers, model owners and other stakeholders, and contribute to discussions with auditors and supervisors.
- Contribute to reproducible analytical workflows, reusable validation code and the automation of repeatable quantitative tests.
What you bring
- A master’s degree or PhD in a quantitative field, such as statistics, mathematics, econometrics, physics, data science or quantitative finance.
- A solid understanding of probability, statistical inference, regression and quantitative model evaluation.
- Practical experience analysing data and implementing quantitative methods, gained through employment, research or substantial academic projects.
- The ability to understand a methodology, translate it into code and critically interpret the results.
- Hands-on experience with Python, or a strong motivation to develop your skills in it, including comfortable use of a modern development environment such as VS Code and familiarity with GitHub Copilot or similar tools to support productivity.
- Strong analytical reasoning, attention to detail and the confidence to question assumptions based on evidence.
- Clear written and spoken English, including the ability to explain technical results to different audiences. Czech is not required.
- A collaborative mindset and willingness to learn in a complex modelling and regulatory environment.
Useful, but not required
- Exposure to risk modelling, model validation or financial data analysis is welcome, but no prior experience in these areas is required.
- Familiarity with IRB models, PD/LGD/EAD estimation or IFRS 9/ECL.
- Knowledge of R, SQL or Git and version-controlled analytical workflows.
- Familiarity with machine-learning methods and their evaluation.
You do not need to meet every optional requirement. We value quantitative depth, sound reasoning and the ability to learn more than familiarity with every banking acronym or software tool.
What we offer
- Substantive quantitative work: investigate model behaviour, test alternative approaches and understand why results differ.
- Exposure to a broad range of portfolios, datasets and modelling methodologies, primarily in credit risk and potentially in other risk areas.
- Structured onboarding, mentoring and learning through practical validation assignments.
- Collaboration with an international network of quantitative specialists across KBC Group.
- Opportunities to develop both technical expertise and the ability to communicate analytical conclusions.
- A collaborative environment that values curiosity, independent thinking and constructive challenge.
Co nabízíme
