Part of wholesale risk modelling team, developing and maintaining IFRS9 Models, Rating Model and stress testing models

Developed credit risk models for climate risk Stress testing for HSBCDeveloped Early warning Signal system using Django Python and Nuxt Js (Vue 2) frameworkPart of development team for development of IFRS 9 system for banks in UK. Tool is developed using Python Django frameworkReviewed models for changes required in FRTB StandardizedApproach and Internal Model Approach.

Conducted IFRS9 model validation for SAMA (Saudi Central bank ) in AL Rajhi Bank, Saudi ArabiaDeveloped Credit risk model for CBUAE stress test exercise for banks in UAEDeveloped stress testing models, worked on IRB,ICAAP & ILAAP for banks in middle east.Developed VBA based RAROC tool for Hatton National Bank, SrilankaDeveloped credit risk models for REIT funds in Middle east.Developed & Validated IFRS9 models for clients in middle eastDeveloped Convolutional neural network based model which would detect tables in Financial reports and then perform financial analysis for rating models. Model was developed using Python, Luminoth, Tensor flow and Front end was developed using Flask framework with Jinja TemplatePerformed Functional Analyst role in implementing Validation tool for clients in middle east.

Lead developer in interest rate derivatives system which would help insurance companies to perform valuation of interest rate derivatives and check hedge effectiveness test as mandated by IRDA and RBI. This tool in first phase was developed using VBA and in second phase it is being developed as web application using Flask – Python web framework, Jinja templates, Bootstrap 4 UI, JavaScript, SQLite embedded DB. Also, as part of tool created C# libraries for valuation of plain vanilla interest derivatives using one factor Libor market model and Monte Carlo simulationDeveloped credit risk models for telecommunication client in Middle East where PIT PD is developed using Markov Chains and LGD model is developed using Support vector, Beta regression and Jacob – Frye method.Audited IFRS9 models for client in middle east

Developed IFRS9 credit risk models for Banks in MENA region where algorithms like Decision tree, Support vector machines, Beta Regression, Logistic Regression were used.Validated wholesale credit risk model and market risk models for reporting of CCAR, DFAST whichincluded validation  of  applicability  of  model,  recreating  results  on  R  and  Python  and  validating regression models in C#.Validated expected credit loss models (PD, LGD & EAD) covering exposures of greater than USD 500 billion based on methodologies such as linear regression, cox-proportional hazard and logistic regression.Validated VaR model for available for sale and held to maturity securities.Validation of regression model developed using C++ library like MLPACK etc.Conducted model performance review by performing VaR back testing;Assisted banks in developing time series model with conditional heteroskedasticity and equity factor depending upon economic conditions.

Assisted banks in monitoring risk limits for all trading activities using quantitative measures, including Value at Risk (VaR) and stress tests;

Assisted in  development  of  derivative pricing model  for  traders  which  involves  components  like  default  probability, projected average coupon, WALA, CPR, VaR analysis which would be used for pricing CMBS, RMBS and ABS;Assisted Banks in automating wholesale funding and generating strategy report which reduced time from 2 hours to 30 minutes;Responsible for  migration  of  the  portfolio  analysis  system  into  a  new  consolidated    This major project included participating in the vendor demo presentation and selection process, sending them sample data, taking steps to improve the process through new platform, creating templates and testing the results/calculations;Responsible for supporting the NY/London distressed asset trading desk to help them analyze / price bid pools, do weekly position reconciliation and create exit portfolio;Credit and risk analysis of the portfolio which involved studying fine level of details of each loan and at the same time efficiently analyzing large volumes of data under demanding time constraints (bid timings). The role involved functions like: data scrubbing, calculations, data analysis, error analysis, generating reports in multiple formats, running the data through licensed models like CAS (Collateral Analysis System), Standard & Poor Levels and generating risk reports.

Developed High frequency trading Alert system for Black Bull Markets New Zealand to capture forex trades which has less run time using Python and My SQL. This system would generate alert where any trader crosses trading and time limit

Assisted in underwriting loan based on income and credit reports which would help in computing LTV and DTI ratio;Assisted in generating funding reports for wholesale and warehouse loans.