FA-61056 / Bond day-count conventions / Open access

Irregular first coupon accrued by quasi-coupon periods: quasi periods are accrued from their start even before issue · case 01

Accrued interest includes days before the bond existed.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

The accrual lower bound uses the quasi period start without clipping at the issue date.

VERIFIED REPAIR

Accrue from the later of the quasi period start and the issue date.

Unsuccessful approach: Clipping only the most recent quasi period misses the oldest one that contains issue.

Case contract

Inputs issue, first coupon and settlement [y,m,d], annual rate and months per period. Settlement must lie in [issue, first] else return "settlement outside first period". Quasi-coupon dates are generated back from the first coupon in steps of months (day clamped to month length) until one is on or before issue. For each quasi period [start, end), accrued days are those in [max(start, issue), min(end, settle)) and are divided by that quasi period length. Accrued = 100*rate/freq*sum, rounded to 6 decimals.

Why this case matters

Bond accrual and pricing systems depend on exact day-count arithmetic; a single-day error changes settlement cash.

1 / The failure

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json
import datetime
from fractions import Fraction
N = 1
observations = []
def solve(issue, first, settle, rate, months):
    def mlen(y, m):
        if m == 2:
            return 29 if (y % 4 == 0 and y % 100 != 0) or y % 400 == 0 else 28
        return 30 if m in (4, 6, 9, 11) else 31
    I = datetime.date(*issue)
    Fc = datetime.date(*first)
    S = datetime.date(*settle)
    if not (I <= S <= Fc):
        return 'settlement outside first period'
    freq = 12 // months
    def back(k):
        t = Fc.year * 12 + Fc.month - 1 - k * months
        y, m = t // 12, t % 12 + 1
        return datetime.date(y, m, min(Fc.day, mlen(y, m)))
    q = [Fc]
    k = 1
    while q[-1] > I:
        q.append(back(k))
        k += 1
    frac = Fraction(0)
    for j in range(len(q) - 1):
        end, start = q[j], q[j + 1]
        lo = start
        hi = min(end, S)
        if hi > lo:
            frac += Fraction((hi - lo).days, (end - start).days)
    return round(float(100 * Fraction(str(rate)) / freq * frac), 6)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression issue date clipping 1', [[2006, 8, 4], [2006, 9, 30], [2006, 8, 13], 0.05, 3], 0.122283], ['regression issue date clipping 2', [[2038, 5, 29], [2038, 7, 25], [2038, 7, 24], 0.045, 3], 0.692308], ['partial repair probe 1', [[2028, 3, 6], [2028, 4, 26], [2028, 3, 15], 0.03, 1], 0.077586], ['partial repair probe 2', [[2027, 7, 30], [2028, 9, 30], [2028, 5, 25], 0.05, 6], 4.103261], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2042, 4, 28], [2043, 1, 30], [2042, 4, 25], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2044, 7, 14], [2044, 11, 30], [2044, 7, 11], 0.02, 12], 'settlement outside first period']], [['regression issue date clipping 1', [[2044, 7, 14], [2045, 2, 28], [2044, 10, 28], 0.02, 12], 0.579235], ['regression issue date clipping 2', [[2012, 3, 21], [2012, 4, 30], [2012, 4, 2], 0.05, 3], 0.164835], ['partial repair probe 1', [[2026, 5, 8], [2026, 8, 31], [2026, 5, 21], 0.02, 3], 0.070652], ['partial repair probe 2', [[2025, 3, 10], [2026, 2, 21], [2025, 8, 17], 0.08, 6], 3.535912], ['boundary control 1', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['boundary control 2', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['normal control 1', [[2035, 5, 29], [2035, 6, 6], [2035, 5, 26], 0.0675, 6], 'settlement outside first period'], ['normal control 2', [[2039, 12, 17], [2040, 7, 20], [2039, 12, 14], 0.0675, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2006, 4, 30], [2007, 1, 26], [2006, 10, 15], 0.03, 3], 1.377359], ['regression issue date clipping 2', [[2023, 8, 31], [2023, 10, 1], [2023, 9, 15], 0.03, 1], 0.124731], ['partial repair probe 1', [[2009, 5, 24], [2010, 7, 10], [2010, 5, 8], 0.0675, 6], 6.451657], ['partial repair probe 2', [[2018, 2, 9], [2018, 5, 31], [2018, 3, 13], 0.03, 3], 0.264312], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2002, 6, 12], [2003, 1, 5], [2002, 6, 9], 0.03, 6], 'settlement outside first period'], ['normal control 2', [[2036, 11, 9], [2036, 11, 19], [2036, 11, 6], 0.02, 6], 'settlement outside first period']], [['regression issue date clipping 1', [[2034, 5, 16], [2034, 6, 30], [2034, 6, 17], 0.05, 12], 0.438356], ['regression issue date clipping 2', [[2003, 5, 14], [2003, 5, 31], [2003, 5, 14], 0.045, 3], 0.0], ['partial repair probe 1', [[2032, 5, 30], [2032, 8, 31], [2032, 6, 4], 0.05, 1], 0.068996], ['partial repair probe 2', [[2021, 12, 10], [2022, 4, 30], [2022, 1, 18], 0.08, 3], 0.847826], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2010, 9, 22], [2011, 2, 14], [2010, 9, 19], 0.05, 3], 'settlement outside first period'], ['normal control 2', [[2021, 3, 4], [2021, 5, 30], [2021, 3, 1], 0.045, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2018, 9, 26], [2019, 1, 30], [2019, 1, 26], 0.08, 3], 2.652174], ['regression issue date clipping 2', [[2001, 4, 14], [2002, 7, 31], [2002, 1, 21], 0.08, 6], 6.169349], ['partial repair probe 1', [[2020, 8, 28], [2020, 10, 22], [2020, 9, 25], 0.0675, 1], 0.509879], ['partial repair probe 2', [[2041, 11, 21], [2044, 4, 12], [2042, 9, 9], 0.02, 12], 1.6], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2031, 4, 22], [2031, 4, 28], [2031, 4, 19], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2014, 9, 23], [2015, 11, 30], [2014, 9, 20], 0.05, 12], 'settlement outside first period']]]
for label, args, expected in fixtures[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
regression issue date clipping 1	`0.597826`	`0.122283`	Failed
regression issue date clipping 2	`1.112637`	`0.692308`	Failed
partial repair probe 1	`0.155172`	`0.077586`	Failed
partial repair probe 2	`5.76087`	`4.103261`	Failed
boundary control 1	`settlement outside first period`	`settlement outside first period`	Passed
boundary control 2	`1.25`	`1.25`	Passed
normal control 1	`settlement outside first period`	`settlement outside first period`	Passed
normal control 2	`settlement outside first period`	`settlement outside first period`	Passed

SHA-256 / f81c2fe9050e665ee16dbad4ecdd3aca0e3a9c331155775c19bb4205085c7349

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json
import datetime
from fractions import Fraction
N = 1
observations = []
def solve(issue, first, settle, rate, months):
    def mlen(y, m):
        if m == 2:
            return 29 if (y % 4 == 0 and y % 100 != 0) or y % 400 == 0 else 28
        return 30 if m in (4, 6, 9, 11) else 31
    I = datetime.date(*issue)
    Fc = datetime.date(*first)
    S = datetime.date(*settle)
    if not (I <= S <= Fc):
        return 'settlement outside first period'
    freq = 12 // months
    def back(k):
        t = Fc.year * 12 + Fc.month - 1 - k * months
        y, m = t // 12, t % 12 + 1
        return datetime.date(y, m, min(Fc.day, mlen(y, m)))
    q = [Fc]
    k = 1
    while q[-1] > I:
        q.append(back(k))
        k += 1
    frac = Fraction(0)
    for j in range(len(q) - 1):
        end, start = q[j], q[j + 1]
        lo = max(start, I) if j == 0 else start
        hi = min(end, S)
        if hi > lo:
            frac += Fraction((hi - lo).days, (end - start).days)
    return round(float(100 * Fraction(str(rate)) / freq * frac), 6)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression issue date clipping 1', [[2006, 8, 4], [2006, 9, 30], [2006, 8, 13], 0.05, 3], 0.122283], ['regression issue date clipping 2', [[2038, 5, 29], [2038, 7, 25], [2038, 7, 24], 0.045, 3], 0.692308], ['partial repair probe 1', [[2028, 3, 6], [2028, 4, 26], [2028, 3, 15], 0.03, 1], 0.077586], ['partial repair probe 2', [[2027, 7, 30], [2028, 9, 30], [2028, 5, 25], 0.05, 6], 4.103261], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2042, 4, 28], [2043, 1, 30], [2042, 4, 25], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2044, 7, 14], [2044, 11, 30], [2044, 7, 11], 0.02, 12], 'settlement outside first period']], [['regression issue date clipping 1', [[2044, 7, 14], [2045, 2, 28], [2044, 10, 28], 0.02, 12], 0.579235], ['regression issue date clipping 2', [[2012, 3, 21], [2012, 4, 30], [2012, 4, 2], 0.05, 3], 0.164835], ['partial repair probe 1', [[2026, 5, 8], [2026, 8, 31], [2026, 5, 21], 0.02, 3], 0.070652], ['partial repair probe 2', [[2025, 3, 10], [2026, 2, 21], [2025, 8, 17], 0.08, 6], 3.535912], ['boundary control 1', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['boundary control 2', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['normal control 1', [[2035, 5, 29], [2035, 6, 6], [2035, 5, 26], 0.0675, 6], 'settlement outside first period'], ['normal control 2', [[2039, 12, 17], [2040, 7, 20], [2039, 12, 14], 0.0675, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2006, 4, 30], [2007, 1, 26], [2006, 10, 15], 0.03, 3], 1.377359], ['regression issue date clipping 2', [[2023, 8, 31], [2023, 10, 1], [2023, 9, 15], 0.03, 1], 0.124731], ['partial repair probe 1', [[2009, 5, 24], [2010, 7, 10], [2010, 5, 8], 0.0675, 6], 6.451657], ['partial repair probe 2', [[2018, 2, 9], [2018, 5, 31], [2018, 3, 13], 0.03, 3], 0.264312], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2002, 6, 12], [2003, 1, 5], [2002, 6, 9], 0.03, 6], 'settlement outside first period'], ['normal control 2', [[2036, 11, 9], [2036, 11, 19], [2036, 11, 6], 0.02, 6], 'settlement outside first period']], [['regression issue date clipping 1', [[2034, 5, 16], [2034, 6, 30], [2034, 6, 17], 0.05, 12], 0.438356], ['regression issue date clipping 2', [[2003, 5, 14], [2003, 5, 31], [2003, 5, 14], 0.045, 3], 0.0], ['partial repair probe 1', [[2032, 5, 30], [2032, 8, 31], [2032, 6, 4], 0.05, 1], 0.068996], ['partial repair probe 2', [[2021, 12, 10], [2022, 4, 30], [2022, 1, 18], 0.08, 3], 0.847826], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2010, 9, 22], [2011, 2, 14], [2010, 9, 19], 0.05, 3], 'settlement outside first period'], ['normal control 2', [[2021, 3, 4], [2021, 5, 30], [2021, 3, 1], 0.045, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2018, 9, 26], [2019, 1, 30], [2019, 1, 26], 0.08, 3], 2.652174], ['regression issue date clipping 2', [[2001, 4, 14], [2002, 7, 31], [2002, 1, 21], 0.08, 6], 6.169349], ['partial repair probe 1', [[2020, 8, 28], [2020, 10, 22], [2020, 9, 25], 0.0675, 1], 0.509879], ['partial repair probe 2', [[2041, 11, 21], [2044, 4, 12], [2042, 9, 9], 0.02, 12], 1.6], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2031, 4, 22], [2031, 4, 28], [2031, 4, 19], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2014, 9, 23], [2015, 11, 30], [2014, 9, 20], 0.05, 12], 'settlement outside first period']]]
for label, args, expected in fixtures[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
regression issue date clipping 1	`0.122283`	`0.122283`	Passed
regression issue date clipping 2	`0.692308`	`0.692308`	Passed
partial repair probe 1	`0.155172`	`0.077586`	Failed
partial repair probe 2	`5.76087`	`4.103261`	Failed
boundary control 1	`settlement outside first period`	`settlement outside first period`	Passed
boundary control 2	`1.25`	`1.25`	Passed
normal control 1	`settlement outside first period`	`settlement outside first period`	Passed
normal control 2	`settlement outside first period`	`settlement outside first period`	Passed

SHA-256 / 3de5036fb5aa72229c9567b994e5b8c652247417cfc17fac807efe7d39dfd58e

3 / The verified repair

Exit 0

"""Failure Map reference implementation. Python standard library only."""
import json
import datetime
from fractions import Fraction
N = 1
observations = []
def solve(issue, first, settle, rate, months):
    def mlen(y, m):
        if m == 2:
            return 29 if (y % 4 == 0 and y % 100 != 0) or y % 400 == 0 else 28
        return 30 if m in (4, 6, 9, 11) else 31
    I = datetime.date(*issue)
    Fc = datetime.date(*first)
    S = datetime.date(*settle)
    if not (I <= S <= Fc):
        return 'settlement outside first period'
    freq = 12 // months
    def back(k):
        t = Fc.year * 12 + Fc.month - 1 - k * months
        y, m = t // 12, t % 12 + 1
        return datetime.date(y, m, min(Fc.day, mlen(y, m)))
    q = [Fc]
    k = 1
    while q[-1] > I:
        q.append(back(k))
        k += 1
    frac = Fraction(0)
    for j in range(len(q) - 1):
        end, start = q[j], q[j + 1]
        lo = max(start, I)
        hi = min(end, S)
        if hi > lo:
            frac += Fraction((hi - lo).days, (end - start).days)
    return round(float(100 * Fraction(str(rate)) / freq * frac), 6)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression issue date clipping 1', [[2006, 8, 4], [2006, 9, 30], [2006, 8, 13], 0.05, 3], 0.122283], ['regression issue date clipping 2', [[2038, 5, 29], [2038, 7, 25], [2038, 7, 24], 0.045, 3], 0.692308], ['partial repair probe 1', [[2028, 3, 6], [2028, 4, 26], [2028, 3, 15], 0.03, 1], 0.077586], ['partial repair probe 2', [[2027, 7, 30], [2028, 9, 30], [2028, 5, 25], 0.05, 6], 4.103261], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2042, 4, 28], [2043, 1, 30], [2042, 4, 25], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2044, 7, 14], [2044, 11, 30], [2044, 7, 11], 0.02, 12], 'settlement outside first period']], [['regression issue date clipping 1', [[2044, 7, 14], [2045, 2, 28], [2044, 10, 28], 0.02, 12], 0.579235], ['regression issue date clipping 2', [[2012, 3, 21], [2012, 4, 30], [2012, 4, 2], 0.05, 3], 0.164835], ['partial repair probe 1', [[2026, 5, 8], [2026, 8, 31], [2026, 5, 21], 0.02, 3], 0.070652], ['partial repair probe 2', [[2025, 3, 10], [2026, 2, 21], [2025, 8, 17], 0.08, 6], 3.535912], ['boundary control 1', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['boundary control 2', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['normal control 1', [[2035, 5, 29], [2035, 6, 6], [2035, 5, 26], 0.0675, 6], 'settlement outside first period'], ['normal control 2', [[2039, 12, 17], [2040, 7, 20], [2039, 12, 14], 0.0675, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2006, 4, 30], [2007, 1, 26], [2006, 10, 15], 0.03, 3], 1.377359], ['regression issue date clipping 2', [[2023, 8, 31], [2023, 10, 1], [2023, 9, 15], 0.03, 1], 0.124731], ['partial repair probe 1', [[2009, 5, 24], [2010, 7, 10], [2010, 5, 8], 0.0675, 6], 6.451657], ['partial repair probe 2', [[2018, 2, 9], [2018, 5, 31], [2018, 3, 13], 0.03, 3], 0.264312], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2002, 6, 12], [2003, 1, 5], [2002, 6, 9], 0.03, 6], 'settlement outside first period'], ['normal control 2', [[2036, 11, 9], [2036, 11, 19], [2036, 11, 6], 0.02, 6], 'settlement outside first period']], [['regression issue date clipping 1', [[2034, 5, 16], [2034, 6, 30], [2034, 6, 17], 0.05, 12], 0.438356], ['regression issue date clipping 2', [[2003, 5, 14], [2003, 5, 31], [2003, 5, 14], 0.045, 3], 0.0], ['partial repair probe 1', [[2032, 5, 30], [2032, 8, 31], [2032, 6, 4], 0.05, 1], 0.068996], ['partial repair probe 2', [[2021, 12, 10], [2022, 4, 30], [2022, 1, 18], 0.08, 3], 0.847826], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2010, 9, 22], [2011, 2, 14], [2010, 9, 19], 0.05, 3], 'settlement outside first period'], ['normal control 2', [[2021, 3, 4], [2021, 5, 30], [2021, 3, 1], 0.045, 3], 'settlement outside first period']], [['regression issue date clipping 1', [[2018, 9, 26], [2019, 1, 30], [2019, 1, 26], 0.08, 3], 2.652174], ['regression issue date clipping 2', [[2001, 4, 14], [2002, 7, 31], [2002, 1, 21], 0.08, 6], 6.169349], ['partial repair probe 1', [[2020, 8, 28], [2020, 10, 22], [2020, 9, 25], 0.0675, 1], 0.509879], ['partial repair probe 2', [[2041, 11, 21], [2044, 4, 12], [2042, 9, 9], 0.02, 12], 1.6], ['boundary control 1', [[2024, 1, 10], [2024, 9, 15], [2024, 9, 16], 0.05, 6], 'settlement outside first period'], ['boundary control 2', [[2024, 3, 15], [2024, 9, 15], [2024, 6, 15], 0.05, 6], 1.25], ['normal control 1', [[2031, 4, 22], [2031, 4, 28], [2031, 4, 19], 0.0675, 3], 'settlement outside first period'], ['normal control 2', [[2014, 9, 23], [2015, 11, 30], [2014, 9, 20], 0.05, 12], 'settlement outside first period']]]
for label, args, expected in fixtures[N-1]:
    check(label, solve(*args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
regression issue date clipping 1	`0.122283`	`0.122283`	Passed
regression issue date clipping 2	`0.692308`	`0.692308`	Passed
partial repair probe 1	`0.077586`	`0.077586`	Passed
partial repair probe 2	`4.103261`	`4.103261`	Passed
boundary control 1	`settlement outside first period`	`settlement outside first period`	Passed
boundary control 2	`1.25`	`1.25`	Passed
normal control 1	`settlement outside first period`	`settlement outside first period`	Passed
normal control 2	`settlement outside first period`	`settlement outside first period`	Passed

SHA-256 / 0190b35df4c40fac6f2d540258ee631d13668b62f8a20db87207f8096ca61034

Verification & scope

A deterministic toy contract stated explicitly in the contract field; no claim of conformance to any published convention text. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:46:51.437318+00:00.

Case digest / 05b7dc76f5e7a45e83a25a1ae9692b0f842214c27cd8f05edc41980ffd2ec32b