Independent financial risk tools

Transparent methods.
Traceable results.

Understand the calculation. Challenge the assumptions. Explain the difference.

A collection of financial risk models and calculation engines supporting benchmarking, model review, audit and advisory.

Explore the models Discover the project
K-Risk Engines: deposit behaviour, counterparty credit risk, market risk aggregation and planned loan prepayment modelling
BenchmarkingModel challengeAudit & advisoryResearch & education

01 / The model portfolio

Find the tool for your question.

Each model has its own repository, methodology, tests and development path. Start with its scope and evidence before using its results.

Counterparty credit riskProject snapshot

K-SACCR

Can I reproduce and explain this exposure?

Inspect SA-CCR calculations from trades to exposure at default, with EU CRR and Basel calculation regimes and a traceable calculation workflow.

Project-reported status. Version 1.0.2 is the current release: the 1.0.2 GitHub Release carries the workbook and the user manual.

Latest published release: v1.0.2.

Full status and limitations

Version 1.0.2 is the current release: the 1.0.2 GitHub Release carries the workbook and the user manual. The engine implements EU CRR and Basel CRE52, but only Basel rules are validated in v1.0.2, by the published CRE99 examples. CRR rules are implemented and tested against illustrative values, not validated (methodology).

Read the source README ↗

Last closed milestone: v1.0.2 - Clearing Coverage and CCP Completeness (2026-10-23).

Repository ↗Methodology ↗Releases ↗
IRRBB · Non-maturity depositsProject snapshot

K-IRRBB NMDs

What assumptions describe deposit behaviour?

A project for modelling customer rates, stable balances and deposit decay, with transparent assumptions and independently testable calculations.

Project-reported status. v0.2.0 – Workbook, data and Excel tests (in progress).

No published stable release.

Full status and limitations

v0.2.0 – Workbook, data and Excel tests (in progress). The v0.1.0 foundation, closed on 2026-10-09, established source layout, documentation, issue governance, CI and static source checks; v0.2.0 builds the first workbook, data import and Excel test harness. Neither delivers pricing, econometric fitting, a production workbook or certified IRRBB outputs. - Decay model (draft, not validated, milestone v0.3.0): core, public facade IRRBB_Decay, harness cases and example, reviewed on release/v0.3.0 and present here since this branch was levelled with main; built to MODEL_CONTRACTS.md and under review in #36; no Excel evidence is recorded yet. Rate and stable-amount models are not built. - Application profile: an Excel workbook will be the end-user deliverable, assembled from reviewed exported VBA source. - Production calculations live in src/core/ without direct Excel object dependencies; at present the data import's parsing, validation, summaries and fingerprint, and the draft decay calculations. - src/modules/ exposes the documented public facade (the data import and the draft decay model, listed in docs/PUBLIC_API.txt); src/workbook/ contains host glue only. - tests/ holds synthetic fixtures, expected results and the VBA regression harness; evidence from a real Excel run must be separately recorded. - No client data, confidential third-party material, vendor source code or customer-specific models are allowed in Git; the repository and its history are public.

Read the source README ↗

Last closed milestone: v0.1.0 - Repository foundation (2026-10-09).

Repository ↗Methodology ↗Roadmap ↗
Market riskProject snapshot

K-FRTB-SA

How do risk inputs become a capital requirement?

Inspect standardized market-risk aggregation across the sensitivities-based method, default risk charge and residual risk add-on. Sensitivities are inputs to the calculation.

Project-reported status. Version 0.1.0 is a prototype of the v0.1.0 — SA prototype milestone (issue #20).

No published stable release.

Full status and limitations

Version 0.1.0 is a prototype of the v0.1.0 — SA prototype milestone (issue #20). It implements the Basel standardized approach (MAR21-MAR23) in Excel VBA with parameters checked against the BIS text, and every output is reconciled with an independent Python reference. It has not yet been compiled or run in Excel, the engine platform decision is still open (scope), and no result is validated for regulatory use (validation).

Read the source README ↗

Last closed milestone: v0.0.1 — Repo setup (2026-10-18).

Repository ↗Methodology ↗Validation ↗
IRRBB · Loan prepaymentsProject snapshot

K-IRRBB Prepayment Models

How might early repayments change cash flows?

A new project for transparent loan-prepayment assumptions in Excel/VBA, supporting IRRBB and ALM benchmarking and independent model challenge.

Project-reported status. Current status: specifications only.

No published stable release.

Full status and limitations

Current status: specifications only. Repository setup (v0.1.0) closed on 2026-10-21 with accepted data, model and validation contracts; prototype setup (v0.2.0) is the open milestone. No prepayment model, VBA engine, workbook, calibration routine or validated numerical output is delivered. These are development objectives, not implemented features.

Read the source README ↗

Last closed milestone: v0.1.0 - Repo setup (2026-10-21).

Repository ↗Roadmap ↗
Project snapshot

K-SFTs

Securities financing transaction counterparty credit risk: a project under development.

Project-reported status. Status: repository setup only.

No published stable release.

Full status and limitations

Status: repository setup only. There is no calculation engine, public API, production workbook or validated method yet. Nothing here should be used to compute exposure. The regulatory basis is chosen: the EU CRR controls, with the Basel Framework as the methodological reference. The first scope is bilateral SFTs under the Financial Collateral Comprehensive Method; products, collateral and outputs are still open decisions.

Read the source README ↗

Last closed milestone: none recorded.

Repository ↗Roadmap ↗
Private project

K-CVA

Part of the risk-model portfolio. The repository is currently private; public scope, implementation status and validation evidence are not available here.

Portfolio snapshots are refreshed through reviewed updates. Status quotations describe each engine: paths, issue numbers and references to “this repository” belong to the source project. Read its full status and limitations; release availability and milestone closure do not establish model validity.

Counterparty credit risk from exposures to capital requirements: K-SACCR and K-SFTs compute the exposures, which feed the ordinary and CCP capital branches and their aggregation; CVA risk is a separate market-risk memo item
Where the counterparty credit risk calculators belong. K-SACCR computes the exposure value of derivatives and K-SFTs that of securities financing transactions; the ordinary and CCP capital branches and their aggregation follow, with CVA risk, covered by K-CVA, as a separate market-risk memo item. An architecture diagram: it does not certify implemented functionality.

02 / A common review approach

Turn a comparison into an explanation.

A matching result is useful evidence. Understanding why it matches—or why it differs—is the purpose of the review.

  1. 01

    Align the inputs

    Reconcile data, scope, conventions and assumptions. Establish that both calculations answer the same question.

  2. 02

    Trace the calculation

    Compare intermediate quantities and test sensitive assumptions. Locate the source of a difference before judging the final number.

  3. 03

    Record the evidence

    Retain the model version, inputs, reference results and explanations. Document unresolved differences and the limits of the conclusion.

Agreement with a benchmark supports a review; it does not, by itself, establish model validity or regulatory compliance.

03 / About the project

One entry point.
Independent projects.

K-Risk Engines brings together tools developed by Daniele Penza to make financial risk calculations easier to inspect, reproduce and challenge.

The collection is intended for practitioners reviewing bank models and calculations, auditors examining implementation, advisors investigating differences, and researchers and students exploring methodology.

This repository hosts the project overview and website. Each model repository owns its implementation, detailed documentation, issues, tests and releases.

Portfolio direction

Develop each model on its own evidence.

  • Explore todayUse the existing SA-CCR release and inspect the IRRBB development and FRTB-SA prototype repositories.
  • Build nextDevelop the prepayment project with an explicit modelling scope and validation plan.
  • Keep the family consistentMaintain clear naming, familiar workbook presentation and visible methodology and evidence across the collection.

This is a direction of travel, not a delivery commitment. Detailed milestones remain in each repository.

Start with the question. Follow the evidence.

Explore the model behind the result.

Browse the portfolio ↑