Andrey Gusev Email & Phone Number

Berlin, De

Berlin, Germany

Berlin, Germany

- Initiated and developed from scratch a behavioral PD model for the German market, which showed an increase in discriminatory ability by 2 times compared to current models- Reviewed current scoring costs and suggest a new scoring logic, which decreased scoring costs by ~40% and increased the profitability of the whole portfolio by ~7-10%- Reviewed, validated, and made amendments in current scoring models and cut-offs for German and Sweden markets

Moscow, Moscow City, Russia

• Russian Top-10 bank, Developed IRB PD models validation methodology for SME and Large corporates segments. Methodology includes quantitative and qualitative tests for scorecards, PD estimation models and Shadow based models. • Kazakhstan state organization. Developed the Antifraud model for the SME segment, this model is used to evaluate the counterparty for the whole SME market of Kazakhstan.• Russian Top-20 bank, Developed a strategy for entering new regions, taking into account the… Show more • Russian Top-10 bank, Developed IRB PD models validation methodology for SME and Large corporates segments. Methodology includes quantitative and qualitative tests for scorecards, PD estimation models and Shadow based models. • Kazakhstan state organization. Developed the Antifraud model for the SME segment, this model is used to evaluate the counterparty for the whole SME market of Kazakhstan.• Russian Top-20 bank, Developed a strategy for entering new regions, taking into account the uncertainty of the level of risk. Developed recommendations for improving the credit pipeline.• Russian Top-60 bank, Developed a PD scorecard for the SME segment.• Russian Top-80 bank, Validated the application scorecard for the car loans segment, segmentation analysis of the incoming application flow by risk level. Also developed the behavioral PD model for this segment. Show less

Москва, Россия

• Developed scoring model for Russian SME sector based on web data;• Russian Top-5 bank, IFRS 9 ECL optimization project. Developed behavior PD models for retail portfolio, changed LGD estimation methodology, changed Staging methodology. As a result, ECL was reduced over than 2 billion rubles. Also led 2 junior analysts.• Russian Top-150 bank, Developed the application PD scorecard for the mortgage loans segment

Moscow, Russia

- Developed Application model for mortgage loans- Developed LGD model for corporate loans- Developed PD model based on internal application score and external scores- Validated external developer’s antifraud model for car loans applications- Automated Expected Credit Losses (ECL) assessment for subsidiary bank with PD and LGD models implementations- Automated credit risk reports