Sample engagement · computed in your browser
Market risk — VaR back-testing (FRTB)
A trading desk's daily VaR back-test measured against actual and hypothetical P&L, with the traffic-light zone and the exception count computed from the series rather than transcribed.
Mistakes we hid in this data on purpose
- — Actual and hypothetical P&L differ, so the exception count depends on which one the back-test is run against — the distinction FRTB turns on and the one a transcribed figure loses.
- — Every column is bound by hand: this pack proposes no mapping, because a book whose `hpl` and `rtpl` columns were matched by name could have them transposed with nothing noticing.
- — No claimed exception rate is on file, so the traffic-light zone is reported and the model owner's own expectation is not judged against — it was never declared.
This is made-up data, and nothing on this page is precomputed — it runs through the real checking engine right now.
The evidence, and what was declared about it
- Book
- Desk A daily back-test — synthetic
- Source
- AlphaIQ synthetic back-test book (market-backtest-dgp.json)
- Fixture
- market-backtest-desk-a.csv
- Model
- Desk A VaR Model v2.4
This sample has no pass/fail target on file — on purpose. Nobody handed us the institution’s own cutoffs, so each number below is reported but not graded, and that is exactly why the recommendation below reads the way it does.