Skip to main content
JOJonas Osman
Methodologies & deliverables

What the work looks like, without naming the clients.

Engagement details, client names, and portfolio-specific results stay confidential. This page describes the methodologies and deliverable formats behind the work — so you can judge fit before the first call.

ALM & IRRBB

EVE and NII sensitivity frameworks

Behavioural modelling of non-maturity deposits and prepayments, supervisory outlier tests, and ALCO-ready reporting — designed so the numbers survive both regulatory challenge and internal use.

Financial Risk & ALM
Credit Risk & IFRS 9

PD, LGD, EAD and ECL engines

Through-the-cycle vs point-in-time PD calibration, downturn LGD frameworks, staging rules aligned to significant-increase-in-credit-risk logic, and forward-looking macro overlays with documented scenario weighting.

Credit Risk & IFRS 9
Model Validation & Model Risk

Independent validation under SR 11-7 / TRIM / SS1/23

Conceptual soundness reviews, data-quality and representativeness testing, benchmarking and backtesting, and model-risk framework work — delivered as validation reports and remediation roadmaps.

Model Validation & Model Risk
Insurance / Actuarial & Solvency II

Reserving, technical provisions, ORSA

Chain-ladder, Bornhuetter-Ferguson and stochastic reserving; best-estimate and risk-margin technical provisions; SCR (standard formula and internal model); ORSA design and pillar-3 disclosure.

Insurance & Solvency II
Climate & Catastrophe Risk

NGFS scenarios and GPD tail calibration

Translating NGFS and bespoke climate scenarios into portfolio-level financials; catastrophe frequency-severity models with generalised-Pareto tail calibration; integration into ORSA, ICAAP, and IFRS 9 forward-looking overlays.

Climate & Catastrophe Risk
Geopolitical & War Risk

Sovereign, sanctions and supply-chain analytics

Sovereign spread and rating-transition modelling, sanctions exposure mapping, war and political-violence pricing frameworks, and supply-chain concentration analytics — translated into balance-sheet impact.

Geopolitical & War Risk
AI & Quantitative Risk Models

Explainable ML with model-risk controls

Gradient-boosting PDs and severity models, LLM-assisted risk review, and explainability layers (SHAP, PDP, surrogates) — delivered with an AI/ML model-risk framework aligned to SR 11-7 and EU AI Act expectations.

AI & Quantitative Risk Models

Typical deliverables

Formats used across engagements. Each is scoped and tailored to the specific model, portfolio, and regulatory context.

  • Technical model documentation (design, calibration, validation)
  • Independent validation opinions and remediation roadmaps
  • Board- and committee-ready summary decks
  • Regulator-facing submissions (ORSA, ICAAP, IFRS 9, IRRBB)
  • Working-model code with reproducible pipelines
  • Governance artefacts: risk appetite, KRIs, model inventory
Educational and professional information only — not financial, actuarial, legal, or investment advice. Client engagement details, results, and references are shared privately, under NDA, on request.