Scoring Model Role - Model Development - Banking

We have fresh opening with Banking client.

Scoring/Decision Model Development

Location: Mumbai

- Look for experience in Decision models, credit risk, SAS, good exposure to regression.

- Advanced Degree (Bachelors required, Masters / PhD preferred) in Statistics, Applied Mathematics, Operations Research, Economics, MBA (Finance), or other highly quantitative discipline

- 4+ years' experience in performing quantitative analysis, statistical modeling, loss forecasting, loan loss reserve modeling, and particularly econometric modeling of consumer credit risk stress losses

- Experience with dynamics of unsecured or secured products a strong plus

- Active role in performing some analytical components of an econometric modeling-driven stress loss process (data collection, data integrity QA/QC/reconcilements, pre-processing, segmentation, variable transformation, variable selection, econometric model estimation, sensitivity testing, back testing, out-of-time testing, model documentation, and model production implementation)

- Exposure to various stress loss modeling approaches at the segment or account level preferred

- Able to communicate technical information verbally and in writing to both technical and non-technical audiences

- Proficiency in SAS/SQL/Oracle/Unix/Microsoft Word, Excel and PowerPoint

- Work as an individual contributor

Detailed JD :

The responsibility includes but not limited to the following activities:

- Obtain and conduct QA/QC on all data required for CCAR/CECL model development

- Develop segment and/or account level CCAR/CECL stress loss models

- Perform all required tests (e.g. sensitivity and back-testing)

- Validate/recalibrate all models annually to incorporate latest data. Redevelop as needed

- Deliver comprehensive model documentation

- Work closely with cross functional teams, including country/region's business stakeholders, model validation and governance teams, and model implementation team

- Prepare responses/presentations to regulatory agencies on all CCAR/CECL models built