Skills

Valuation

Valuation models and methodology across mortgages (Aareal, QuantLib), Prudent Valuation (KfW, ABN AMRO), XVA (PwC), and an independent review of a bank's mortgage valuation for an ECB on-site inspection.

Valuation is how an instrument is priced and how uncertain that price is: pricing models and their calibration, valuation adjustments (Prudent Valuation AVAs, CVA, DVA and FVA), and the controls that keep model prices and booked values consistent.

Used in: Prudent Valuation · Mortgage Valuation · Mortgage Portfolio Analytics · Market Price Uncertainty AVAs · XVA Netting Analytics View projects →

Prudent Valuation

Regulatory valuation framework implemented for ABN AMRO's rate derivatives and illiquid bonds and as a C++ extension module for KfW's full portfolio.

Regulatory valuation under CRR's Prudent Valuation framework — implemented from requirements specification through to production code for portfolios of illiquid instruments and bonds.

Used in: Prudent Valuation · Market Price Uncertainty AVAs View projects →

Counterparty Credit Risk

Counterparty credit risk modelling for PwC, including exposure simulation, netting set analysis and ISDA/CSA term optimisation across banks' derivatives portfolios.

Counterparty credit risk covers the exposure a bank faces if a trading counterparty defaults before settlement — modelled through expected exposure profiles, netting and collateral agreements, and translated into XVA charges.

Used in: XVA Netting Analytics View projects →

XVA

XVA analytics built for PwC to support banks negotiating netting sets and collateral terms with their trading counterparties.

Valuation adjustments computed on top of Monte Carlo exposure simulations to quantify counterparty credit risk and funding costs in derivatives portfolios. CVA (Credit Valuation Adjustment) captures the market value of counterparty default risk; DVA (Debit Valuation Adjustment) is the symmetric own-credit component; FVA (Funding Valuation Adjustment) reflects the cost of funding uncollateralised exposures.

Used in: XVA Netting Analytics View projects →

IRRBB

Interest Rate Risk in the Banking Book — covered in the Regnology Risk Hub customization at Accenture/TiGital and as the dedicated IRRBB implementation for Daimler Financial Services.

Interest Rate Risk in the Banking Book is a recurring focus of my work — from structure and condition contribution calculations to whole-bank ALM/FTP steering.

Used in: Integrated ALM Risk Platform · ALM & Balance-Sheet Risk View projects →

FTP

Funds Transfer Pricing — the custom FTP calculation engine built for Hamburg Commercial Bank on OneSumX for Risk, and the ALM/FTP scope of the Regnology Risk Hub rollout at Accenture/TiGital.

Funds Transfer Pricing runs through much of my banking-book work: building and extending FTP engines that price securities, trading-book deals and hybrid approaches, and wiring transfer-rate logic into whole-bank ALM/FTP steering.

Used in: Integrated ALM Risk Platform · FTP Engine · ALM & Balance-Sheet Risk · Risk-Based Loan Pricing View projects →

Market Risk

Market risk methodology at LBBW (measurement methods, approval of valuation models, a multi-curve interest-rate risk model), VaR and IRRBB in the integrated ALM platform for 27 Austrian banks, market risk management at Daimler Financial Services, and market price uncertainty for ABN AMRO.

Market risk is the risk of losses from moves in interest rates, spreads, FX and other market factors, measured with sensitivities, Value-at-Risk and stress tests: for the trading book and, as IRRBB, for the banking book.

Used in: Integrated ALM Risk Platform · Market Price Uncertainty AVAs · Market Risk Methodology · ALM & Balance-Sheet Risk View projects →

VaR

Value-at-Risk — Monte Carlo VaR on the LBBW multi-curve valuation engine, and the VaR calculations covered in the Regnology Risk Hub customization at Accenture/TiGital.

Value-at-Risk for whole-portfolio and whole-bank steering: implementing Monte Carlo VaR on top of a multi-curve valuation engine and configuring VaR calculations in risk platforms alongside ALM and FTP.

Used in: Integrated ALM Risk Platform · Market Risk Methodology · ALM & Balance-Sheet Risk View projects →

PD

PD model validation at DekaBank — backtesting PD estimates against realised default rates and testing rating distribution stability with Kullback–Leibler divergence tests in R.

Probability of Default is the likelihood that a borrower will fail to meet its obligations over a given horizon. PD model validation includes backtesting against realised defaults, discriminatory power tests and population stability analysis.

Used in: ECL Model Validation View projects →

LGD

LGD model validation at DekaBank — backtesting LGD estimates against realised losses and assessing model stability with Kullback–Leibler divergence tests in R.

Loss Given Default is the share of an exposure that is lost when a borrower defaults, net of recoveries. LGD validation covers backtesting against workout outcomes, downturn adjustment verification and loss distribution stability over time.

Used in: ECL Model Validation View projects →

ECL

Expected Credit Loss validation at DekaBank under IFRS 9, testing ECL estimators for retail and wholesale portfolios with statistical tests in R.

Expected Credit Loss is the probability-weighted estimate of credit losses over the life of a financial instrument. Validation covers the integrity of the PD, LGD and EAD inputs, the staging logic and the forward-looking macro adjustments.

Used in: ECL Model Validation View projects →

IFRS 9

IFRS 9 model validation at DekaBank, covering PD, LGD and ECL estimators for both retail and wholesale credit portfolios.

IFRS 9 governs the classification, measurement and impairment of financial instruments. The impairment model requires forward-looking expected credit loss estimates across three stages, built on PD, LGD and EAD components.

Used in: ECL Model Validation View projects →

Model Validation

Independent model validation at DekaBank under IFRS 9, running statistical tests in R to assess PD, LGD and ECL estimators for retail and wholesale portfolios.

Model validation is the independent review of risk models to assess their conceptual soundness, data integrity, discriminatory power and ongoing performance — covering backtesting, stability analysis and benchmarking against alternative approaches.

Used in: Mortgage Portfolio Analytics · ECL Model Validation · Market Risk Methodology View projects →

C++

Used to extend QuantLib for mortgage-portfolio pricing at Aareal Bank, build the Prudent Valuation module at KfW, and implement the multi-curve valuation engine at Landesbank Baden-Württemberg.

Used for performance-sensitive quant work — extending and maintaining QuantLib-based pricing engines and building custom valuation modules from scratch.

Used in: Prudent Valuation · Mortgage Valuation · XVA Netting Analytics · Market Risk Methodology View projects →

QuantLib

Extended with a custom product and valuation method for mortgage-portfolio pricing and risk steering at Aareal Bank.

Extended with custom products and valuation methods rather than used purely off the shelf — comfortable working inside its C++ internals when a project needs something the library doesn't provide out of the box.

Used in: Mortgage Valuation · XVA Netting Analytics View projects →

Python

Used to parse and validate the XML data stack feeding the QuantLib pricing engine built for Aareal Bank.

Used for data parsing, validation and integration glue around pricing engines written in other languages.

Used in: FTP Engine · Mortgage Valuation View projects →

Java

Core implementation language for the Regnology Risk Hub customization at Accenture/TiGital and the FTP calculation engine built for Hamburg Commercial Bank's OneSumX for Risk.

Used for enterprise banking-platform integration work, including customizing commercial risk platforms and building calculation modules with automated test coverage.

Used in: Integrated ALM Risk Platform · FTP Engine · Credit Risk View projects →

R

Built a custom R package to independently reproduce a bank's mortgage-portfolio valuation calculations for an ECB on-site mission.

Used to build an independent reproduction of a bank's valuation methodology from scratch, rather than just running existing scripts.

Used in: Mortgage Portfolio Analytics · ECL Model Validation View projects →

Oracle PL/SQL

Backing database for the FTP calculation engine built for Hamburg Commercial Bank's OneSumX for Risk, and the PL/SQL verification suite for KfW's Prudent Valuation C++ module.

Comfortable working directly against Oracle as the backing store for risk-calculation platforms, including schema design and query tuning, as well as writing PL/SQL scripts that check a system's output against reference cases as part of the delivery process.

Used in: FTP Engine · Prudent Valuation View projects →

SQL Server

Backing store and stored-procedure reporting layer across the Regnology Risk Hub, Wolters Kluwer credit-risk workflow, and Landesbank Baden-Württemberg multi-curve projects.

Used as both the transactional backing store and the reporting layer, including writing the stored procedures that drive downstream reports.

Used in: Integrated ALM Risk Platform · Credit Risk · Market Risk Methodology View projects →

Postgres

Storage layer for the XML-modelled data stack feeding the QuantLib pricing engine at Aareal Bank.

Used as the storage layer for a data pipeline feeding a pricing engine, with a schema driven by the XML data model it needed to serve.

Used in: Mortgage Valuation View projects →

Riskpro

The risk platform behind the last three engagements — branded OneSumX for Risk at Wolters Kluwer and Hamburg Commercial Bank, and branded Regnology Risk Hub at Regnology.

Deep hands-on experience with the Riskpro engine across its vendor rebrandings: customizing and extending it in Java for ALM, FTP, VaR and credit-risk calculations, configuring it for multi-bank platforms, and automating its batch and API workflows end to end.

Used in: Integrated ALM Risk Platform · FTP Engine · Credit Risk View projects →

Focus FIS

Platform underlying the IRRBB implementation for Daimler Financial Services, extended with VBA.

Worked directly with Focus FIS's ALM/FTP modules to implement and extend interest-rate risk calculations for a multinational banking group.

Used in: ALM & Balance-Sheet Risk View projects →

Pytest

Daily test suite detecting deal-level valuation changes after the jobchain run for Hamburg Commercial Bank's FTP engine.

Used to build automated regression suites that catch valuation changes after a batch run, rather than relying on manual spot-checks.

Used in: FTP Engine View projects →

TeamCity

CI pipeline for the end-to-end credit-risk automation workflow built for Wolters Kluwer.

Used to wire up CI pipelines for regulatory workflows, so validation runs automatically rather than depending on someone remembering to run it.

Used in: Credit Risk View projects →

Bloomberg

Market-data integration for the Prudent Valuation implementation on a portfolio of illiquid bonds at ABN-AMRO.

Familiar with pulling and validating market data through the Bloomberg Terminal and its APIs for pricing and valuation workflows.

Used in: Market Price Uncertainty AVAs · ALM & Balance-Sheet Risk View projects →

.Net

VB.Net data layer for the multi-curve valuation engine built for Landesbank Baden-Württemberg.

Used mainly for the data layer feeding valuation engines, integrating trade data with the calculation core.

Used in: Market Risk Methodology View projects →

VBA

Embedded project-specific calculations into the clients' existing tools: the Focus FIS IRRBB platform at Daimler Financial Services, ABN AMRO's Prudent Valuation tool and a pricing library at GE Capital Bank Spain.

Used to add the calculations a standard platform doesn't provide, inside the environment the client already runs, so the logic sits where the users work rather than in a separate system someone has to maintain.

Used in: Market Price Uncertainty AVAs · ALM & Balance-Sheet Risk · Risk-Based Loan Pricing View projects →