Transparent methods. Traceable results.
🌐 Visit the website · 📚 Project wiki
A single entry point to independent financial risk models and calculation engines for benchmarking, model challenge, audit, advisory and education.
| Project | Purpose | Project-reported status | Last closed milestone |
|---|---|---|---|
| K-SACCR | SA-CCR calculation and reconciliation | 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 caveats | v1.0.2 - Clearing Coverage and CCP Completeness (2026-10-23) |
| K-IRRBB NMDs | Non-maturity deposit behavioural modelling | v0.2.0 – Workbook, data and Excel tests (in progress). No published stable release. Full status and caveats | v0.1.0 - Repository foundation (2026-10-09) |
| K-FRTB-SA | Standardized market-risk aggregation | Version 0.1.0 is a prototype of the v0.1.0 — SA prototype milestone (issue #20). No published stable release. Full status and caveats | v0.0.1 — Repo setup (2026-10-18) |
| K-IRRBB Prepayment Models | Behavioural loan prepayment modelling | Current status: specifications only. No published stable release. Full status and caveats | v0.1.0 - Repo setup (2026-10-21) |
| K-SFTs | Counterparty credit risk for securities financing transactions | Status: repository setup only. No published stable release. Full status and caveats | None recorded |
K-CVA is also part of the portfolio. Its repository is currently private; public scope, implementation status and validation evidence are not available here.
The table and full evidence report are generated snapshots, not live project dashboards. Pinned source links identify the evidence used; consult the source repositories for changes since the last refresh.
Each repository owns its implementation, methodology, tests, issues and releases. Consult it for current scope, evidence and licensing. This hub contains presentation and portfolio documentation, not calculation code.
The static landing page is index.html; images and styles live in assets/.
Serving it requires no build step or application dependencies. Repository checks
and automation use Python, curl and GitHub Actions as documented below.
For GitHub Pages, select Settings → Pages → Deploy from a branch → main → / (root). The live website is danielep71.github.io/K-Risk-Engines.
Run python3 .github/scripts/check-website.py from the repository root.
Add --external to check external URLs with curl.
The dependency-free check covers explicit HTML tag nesting, document title,
language and doctype, duplicate IDs, image alt attributes, local assets,
CSS url() references and internal anchors. It also checks links in every
tracked Markdown file: inline, reference-style, autolinks and HTML href/src,
outside fenced and inline code. Relative targets must exist inside the
repository, and fragments must match a heading (GitHub's anchor rules) or an
explicit <a id>. It is a focused structural check, not a full HTML standards
or accessibility validator. Its own tests run with
python3 -m unittest discover -s .github/scripts -p 'test_*.py'.
GitHub Actions runs local checks on every pull request and push to main
or a release/** branch. External links are checked on pushes, weekly and
on manual runs; redirects are accepted. Broken links fail the check. Network
errors, HTTP 403/429 and server errors are reported as warnings because
availability could not be verified. External page fragments are not validated;
external URLs from Markdown files follow the same rules.
A daily workflow proposes source-attributed status, release and last-closed-milestone updates in a reusable review PR. See operation and review rules.
Daily repository analytics are stored on traffic-history.
See setup and data definitions for the required analytics
environment secret, schedules, badges and alerts. This tracks repository traffic,
not GitHub Pages visitors. Clone counts include this repository's own CI
checkouts and maintainer clones, so they do not measure outside interest.
See portfolio governance for ownership, project status definitions, change checks and issue handling. Report vulnerabilities privately under the security policy. For hub defects and improvements, use the issue chooser. The maintainer assigns a milestone and confirms the priority during triage. Engine calculation and methodology reports belong in their own repositories.
The hub's code and documentation are licensed under MPL-2.0.
The portfolio artwork (assets/K-Risk-Engines-social-preview.png) is excluded;
no separate reuse licence is granted for it. See licensing and attribution.
Linked engines retain their own licences.
- Align data, scope, conventions and assumptions.
- Compare intermediate calculations and investigate differences.
- Record versions, reference results, explanations and unresolved limitations.
Agreement with a benchmark supports a review; it does not by itself establish model validity or regulatory compliance.
Develop the existing projects on their own validation evidence, develop the new prepayment project, and maintain consistent presentation and documentation across the collection. Detailed milestones remain in each model repository.
👤 An independent project by Daniele Penza. Individual project licences apply to their respective contents.