Ruoxuan
![](/upload_mm/4/f/e/265754_fullimage_cnt-ruoxuan_500x154.jpg)
Data analytics can make an impact on safeguarding a sustainable business and improving the financial well-being of society
“My name is Ruoxuan. I work as credit risk model validator at ING. I came across ING by coincidence when I attended their two-day cross-border business course in Amsterdam and Brussels. I had studied Econometrics at Erasmus University in Rotterdam and always enjoyed ‘treasure hunting’ in data. But when it came to a career, I wasn’t sure about which industry I wanted to work in. There are many perspectives I find interesting and where it would be meaningful to apply data analytics. The ING event was great, well organised, cool people, and amazing food. I really enjoyed it, so when I was about to write my master thesis, I reached out to them, which resulted in an internship in the credit risk model validation department.
During my internship I worked on my thesis and did real internal ratings based (IRB) periodical model validation tasks. It was both challenging and exciting. Having a mentor was helpful, especially at the start when I was new to the team. He provided advice on that what wasn’t clear to me, which made my onboarding process a lot smoother.
My thesis research was about imbalanced data classification, with a focus on SMOTE, one of the most popular resampling techniques. I investigated its theoretical properties and effectiveness under different data properties for different classifiers from traditional statistical models like Logistic Regression and Bayesian Estimation to machine learning algorithms. This was done through both simulation experiments and an empirical study on the probability of default (PD) prediction, for which I worked with a dataset of Italian mortgages provided by ING. It was the first time I worked with 100% real-world raw finance data and it was a great hands-on experience understanding the portfolio and checking data quality. This was undoubtedly crucial to any further steps of modelling and analysis. I presented my progress at monthly one-hour brainstorming sessions and got practical feedback. This not only helped my thesis but also boosted my engagement with the team and prepared me well for the defence of my thesis.
After a second review of the IRB periodical validation for the Italy Mortgages PD model to grasp the essence of how PD is modelled and validated, I got the chance to perform (under guidance) a periodical validation for the Italy Mortgages EAD model.
Credit risk modelling doesn’t necessarily involve the most complicated mathematics, but it is an important area where data analytics can make an impact on safeguarding a sustainable business and improving the financial well-being of society. I am curious to keep learning how to quantitatively challenge a model. How to build models while complying with numerous legal requirements, and how to effectively use data visualisation to form an opinion on the validity of a model. Credit decision models are mainly machine learning models and I am enthusiastic about validating them because of their complexities.
Next to the actual work and my team, what really I love is ING’s international, innovative, diverse and inclusive working environment. I enjoyed the learning journey of my internship so much I asked stay. I am excited to keep contributing while learning as I start the next phase of my ING adventure!”