source verification/studies/calibration-v3/05-cal-den-2-registration.md
sha256 84de9b5b9176a10229db0895e0b843ed3645a350de1dcf2fabe17b2003540658
bytes 4,909
RESEARCH · NOT THE CORE STANDARD

Research extension. This page is a study, not a measurement. The core standard is CEBE, Senior Claims %, CEBE mNAV and Claims Grade, each published with its formula and date. Studies here test priors against outcomes and can be retired; retired studies stay on the site with their correction notes. Nothing on this page enters a tracker figure.

Cal-den-2, registration, sealed text

2026-09-03, re-issued 2026-09-04 with section 2 filled from the Con-den-1 findings of 2026-08-31 (task #C-126, registration committed at f9923f5 before the read) and section 3 amended by ruling of the same date to add the conversion-eligibility field; re-issued 2026-09-05 to correct the field's motivating example on the MSTR anchor's filing read, on Bobby's word. All changes precede sealing. No section other than 2 and 3 changed. Hashes with Cal-num-4 and the ladder when their gates return. Delivery and moneyness definitions inherited from Con-den-1 unchanged. Lifetime horizon. Scoring on resolved instruments only, the Cal-den-1 lesson. Sealed resolution set of all outstanding roster convertibles at seal. Wilson 90 percent intervals per class. A failed class prior retires; refits are Con-den-2.

1. Priority

Registers alongside Cal-num-4, not after, and seals under the same external hash. The red team's sharpest program-level objection is that the priors were published for observed conversions and nothing tests conversions. This study is that test, and its absence was the gap.

2. Findings carried in

The fitted delivery rates by issuance class, with counts, and the band minimum, fill from the Con-den-1 findings file on origin (docs/con-den-1-return, #C-169). They are recorded here as the priors under test and are not refit inside this study.

Priors under test, delivery by moneyness class at issuance, as published in Con-den-1 section 2. Rates rest on 26 delivery observations across 72 instruments; 11 of the 26 are inferred from a converted status rather than read from a filing. Intervals are the Wilson 95 percent bounds as published; this study's own scoring uses Wilson 90 percent per section 1 and the two are not the same object.

class at issuanceinstrumentsdelivery coveragedelivered anyprior rateWilson 95 percent
ITM14551.0000.5655 to 1.0000
Near12520.4000.1176 to 0.7693
Deep OTM241570.4670.2481 to 0.6988
Unbanded22111.0000.2065 to 1.0000

The registered band cut is 30 percent premium with the ITM boundary at M = 1.00. Con-den-1 section 5 shows the ordering of these priors is not robust to the cut, which is recorded here so that a failed class prior is read as a result about the classifier as registered and not about convertibles generally.

Band minimum per class: not stated in the Con-den-1 findings. It inherits from the Cal-den-1 registration unchanged, per section 1, and is not restated here.

No other section of this registration depends on the values above.

Amended 2026-09-17: the Con-den-1 findings and tables are published at cebetracker.io/research/con-den-1/ at sealing, and this registration cites that path.

3. Resolution and scoring

Conversion eligibility, added by ruling 2026-09-04. Each instrument in the resolution set carries one field beside its moneyness class, convertible during the observation window, with values yes, no, or undisclosed, sourced from the indenture or governing terms with accession. A class prior is scored only on instruments for which the value is yes. Instruments carrying no or undisclosed are reported in the class table with their count but do not enter the rate. The reason is a contractual gate the moneyness class cannot see. Con-den-1 section 6 first stated it using three Strategy convertibles said to be closed to conversion until 2028-03-15; the MSTR anchor's read of ten covering filings on 2026-09-05 found that premise wrong for four series (the 2028, 2030A, 2031 and 2032 were convertible at holder option during quarters of 2025 in the issuer's own words, and holders declined), and right for two (the 2029 and 2030B, closed end to end). The corrected reading is recorded here so the field is not credited with a case it does not have. The principle stands on the two: a study that scores delivery by moneyness alone would read a closed conversion right as holder behavior, and a contingent right counts as open only for the periods the issuer reports its condition satisfied. Field shape and sourcing rules are R-0904 F-4. Lanes populate the field by courier before any class reaches its band minimum.

An instrument resolves on termination, by conversion, redemption, repurchase, or maturity, and its resolution is recorded by its anchor lane as a standing item, one line to the study folder with the accession. A class scores when it reaches its band minimum of resolved instruments. The interval rule and the retirement test for a class prior are inherited from the Cal-den-1 registration unchanged.

4. Seal and owners

External seal per G-1. Registration, copilot, on the findings read. Ratification, Bobby. Resolution tracking, each anchor lane. Scoring, a study desk, per class, on reaching the band minimum. Page, Code.

More studies, and the construction they test, are listed on the research page. The core standard's construction is documented on the methodology page.