FA-61311 / Bond day-count conventions / Open access

Act/Act AFB whole-year decomposition: the leap-day search looks at the whole period instead of the stub · case 01

Stubs in periods of more than a year use 366 because of leap days in the whole years.

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

ROOT CAUSE

The 29 February search runs up to the final end date instead of the stub end.

VERIFIED REPAIR

Search for 29 February only inside (start, stub end].

Unsuccessful approach: OR-ing the correct search with the stub end year leap status still overcounts.

Case contract

Inputs start < end [y,m,d]. Count whole years n stepping back from the end date (an end of 29 February maps to 28 February in non-leap years) while the stepped date is not before start. The stub runs from start to the end stepped back n years; its denominator is 366 if a 29 February lies in (start, stub end], else 365. Return n + stub days/denominator rounded to 9 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
N = 1
observations = []
def solve(a, b):
    A = datetime.date(*a)
    B = datetime.date(*b)
    def leap(y):
        return (y % 4 == 0 and y % 100 != 0) or y % 400 == 0
    def back_years(n):
        y = B.year - n
        if B.month == 2 and B.day == 29 and not leap(y):
            return datetime.date(y, 2, 28)
        return datetime.date(y, B.month, B.day)
    n = 0
    while back_years(n + 1) >= A:
        n += 1
    stub_end = back_years(n)
    has29 = any(leap(y) and A < datetime.date(y, 2, 29) <= B for y in range(A.year, B.year + 1))
    den = 366 if has29 else 365
    return round(n + (stub_end - A).days / den, 9)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression stub leap window 1', [[2059, 3, 3], [2062, 3, 16]], 3.035616438], ['regression stub leap window 2', [[2022, 6, 12], [2030, 6, 23]], 8.030136986], ['partial repair probe 1', [[2016, 3, 31], [2019, 7, 2]], 3.254794521], ['partial repair probe 2', [[2088, 7, 31], [2088, 10, 23]], 0.230136986], ['boundary control 1', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[1991, 10, 30], [1995, 7, 29]], 3.745901639], ['normal control 2', [[2097, 11, 30], [2098, 2, 21]], 0.22739726]], [['regression stub leap window 1', [[2067, 4, 30], [2072, 9, 7]], 5.356164384], ['regression stub leap window 2', [[2053, 2, 8], [2057, 12, 7]], 4.82739726], ['partial repair probe 1', [[2068, 8, 18], [2070, 8, 23]], 2.01369863], ['partial repair probe 2', [[2040, 5, 31], [2042, 11, 14]], 2.457534247], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2017, 12, 31], [2018, 5, 3]], 0.336986301], ['normal control 2', [[2052, 2, 28], [2059, 11, 6]], 7.68852459]], [['regression stub leap window 1', [[2030, 1, 31], [2034, 12, 4]], 4.84109589], ['regression stub leap window 2', [[1993, 2, 28], [2000, 11, 9]], 7.695890411], ['partial repair probe 1', [[2068, 3, 23], [2069, 12, 20]], 1.745205479], ['partial repair probe 2', [[2076, 7, 3], [2077, 8, 10]], 1.104109589], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2036, 9, 30], [2038, 9, 29]], 1.997260274], ['normal control 2', [[2019, 7, 18], [2024, 2, 29]], 4.617486339]], [['regression stub leap window 1', [[2004, 5, 26], [2008, 2, 29]], 3.761643836], ['regression stub leap window 2', [[2010, 12, 22], [2012, 2, 29]], 1.18630137], ['partial repair probe 1', [[2084, 8, 12], [2087, 8, 15]], 3.008219178], ['partial repair probe 2', [[2092, 4, 30], [2092, 10, 27]], 0.493150685], ['boundary control 1', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2011, 3, 11], [2012, 2, 29]], 0.969945355], ['normal control 2', [[2012, 8, 30], [2015, 5, 30]], 2.747945205]], [['regression stub leap window 1', [[2001, 5, 31], [2009, 1, 23]], 7.649315068], ['regression stub leap window 2', [[2015, 2, 28], [2021, 8, 1]], 6.421917808], ['partial repair probe 1', [[2056, 2, 29], [2059, 9, 9]], 3.528767123], ['partial repair probe 2', [[2068, 3, 28], [2071, 10, 31]], 3.594520548], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[2046, 3, 23], [2046, 7, 28]], 0.347945205], ['normal control 2', [[2080, 6, 2], [2082, 2, 16]], 1.709589041]]]
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 stub leap window 1	`3.035519126`	`3.035616438`	Failed
regression stub leap window 2	`8.030054645`	`8.030136986`	Failed
partial repair probe 1	`3.254794521`	`3.254794521`	Passed
partial repair probe 2	`0.230136986`	`0.230136986`	Passed
boundary control 1	`4.99726776`	`4.99726776`	Passed
boundary control 2	`1.0`	`1.0`	Passed
normal control 1	`3.745901639`	`3.745901639`	Passed
normal control 2	`0.22739726`	`0.22739726`	Passed

SHA-256 / 734b982bced0397fbddf0df5463f9981abbfcbe47be014f5348757da17775b41

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json
import datetime
N = 1
observations = []
def solve(a, b):
    A = datetime.date(*a)
    B = datetime.date(*b)
    def leap(y):
        return (y % 4 == 0 and y % 100 != 0) or y % 400 == 0
    def back_years(n):
        y = B.year - n
        if B.month == 2 and B.day == 29 and not leap(y):
            return datetime.date(y, 2, 28)
        return datetime.date(y, B.month, B.day)
    n = 0
    while back_years(n + 1) >= A:
        n += 1
    stub_end = back_years(n)
    has29 = any(leap(y) and A < datetime.date(y, 2, 29) <= stub_end for y in range(A.year, stub_end.year + 1))
    den = 366 if has29 or leap(stub_end.year) else 365
    return round(n + (stub_end - A).days / den, 9)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression stub leap window 1', [[2059, 3, 3], [2062, 3, 16]], 3.035616438], ['regression stub leap window 2', [[2022, 6, 12], [2030, 6, 23]], 8.030136986], ['partial repair probe 1', [[2016, 3, 31], [2019, 7, 2]], 3.254794521], ['partial repair probe 2', [[2088, 7, 31], [2088, 10, 23]], 0.230136986], ['boundary control 1', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[1991, 10, 30], [1995, 7, 29]], 3.745901639], ['normal control 2', [[2097, 11, 30], [2098, 2, 21]], 0.22739726]], [['regression stub leap window 1', [[2067, 4, 30], [2072, 9, 7]], 5.356164384], ['regression stub leap window 2', [[2053, 2, 8], [2057, 12, 7]], 4.82739726], ['partial repair probe 1', [[2068, 8, 18], [2070, 8, 23]], 2.01369863], ['partial repair probe 2', [[2040, 5, 31], [2042, 11, 14]], 2.457534247], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2017, 12, 31], [2018, 5, 3]], 0.336986301], ['normal control 2', [[2052, 2, 28], [2059, 11, 6]], 7.68852459]], [['regression stub leap window 1', [[2030, 1, 31], [2034, 12, 4]], 4.84109589], ['regression stub leap window 2', [[1993, 2, 28], [2000, 11, 9]], 7.695890411], ['partial repair probe 1', [[2068, 3, 23], [2069, 12, 20]], 1.745205479], ['partial repair probe 2', [[2076, 7, 3], [2077, 8, 10]], 1.104109589], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2036, 9, 30], [2038, 9, 29]], 1.997260274], ['normal control 2', [[2019, 7, 18], [2024, 2, 29]], 4.617486339]], [['regression stub leap window 1', [[2004, 5, 26], [2008, 2, 29]], 3.761643836], ['regression stub leap window 2', [[2010, 12, 22], [2012, 2, 29]], 1.18630137], ['partial repair probe 1', [[2084, 8, 12], [2087, 8, 15]], 3.008219178], ['partial repair probe 2', [[2092, 4, 30], [2092, 10, 27]], 0.493150685], ['boundary control 1', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2011, 3, 11], [2012, 2, 29]], 0.969945355], ['normal control 2', [[2012, 8, 30], [2015, 5, 30]], 2.747945205]], [['regression stub leap window 1', [[2001, 5, 31], [2009, 1, 23]], 7.649315068], ['regression stub leap window 2', [[2015, 2, 28], [2021, 8, 1]], 6.421917808], ['partial repair probe 1', [[2056, 2, 29], [2059, 9, 9]], 3.528767123], ['partial repair probe 2', [[2068, 3, 28], [2071, 10, 31]], 3.594520548], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[2046, 3, 23], [2046, 7, 28]], 0.347945205], ['normal control 2', [[2080, 6, 2], [2082, 2, 16]], 1.709589041]]]
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 stub leap window 1	`3.035616438`	`3.035616438`	Passed
regression stub leap window 2	`8.030136986`	`8.030136986`	Passed
partial repair probe 1	`3.254098361`	`3.254794521`	Failed
partial repair probe 2	`0.229508197`	`0.230136986`	Failed
boundary control 1	`4.99726776`	`4.99726776`	Passed
boundary control 2	`1.0`	`1.0`	Passed
normal control 1	`3.745901639`	`3.745901639`	Passed
normal control 2	`0.22739726`	`0.22739726`	Passed

SHA-256 / ee19ab500d8ff0d65fa2ca3fdafffd4e2e0bab291e7fb2ebc89f028695e3c8d1

3 / The verified repair

Exit 0

"""Failure Map reference implementation. Python standard library only."""
import json
import datetime
N = 1
observations = []
def solve(a, b):
    A = datetime.date(*a)
    B = datetime.date(*b)
    def leap(y):
        return (y % 4 == 0 and y % 100 != 0) or y % 400 == 0
    def back_years(n):
        y = B.year - n
        if B.month == 2 and B.day == 29 and not leap(y):
            return datetime.date(y, 2, 28)
        return datetime.date(y, B.month, B.day)
    n = 0
    while back_years(n + 1) >= A:
        n += 1
    stub_end = back_years(n)
    has29 = any(leap(y) and A < datetime.date(y, 2, 29) <= stub_end for y in range(A.year, stub_end.year + 1))
    den = 366 if has29 else 365
    return round(n + (stub_end - A).days / den, 9)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[['regression stub leap window 1', [[2059, 3, 3], [2062, 3, 16]], 3.035616438], ['regression stub leap window 2', [[2022, 6, 12], [2030, 6, 23]], 8.030136986], ['partial repair probe 1', [[2016, 3, 31], [2019, 7, 2]], 3.254794521], ['partial repair probe 2', [[2088, 7, 31], [2088, 10, 23]], 0.230136986], ['boundary control 1', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[1991, 10, 30], [1995, 7, 29]], 3.745901639], ['normal control 2', [[2097, 11, 30], [2098, 2, 21]], 0.22739726]], [['regression stub leap window 1', [[2067, 4, 30], [2072, 9, 7]], 5.356164384], ['regression stub leap window 2', [[2053, 2, 8], [2057, 12, 7]], 4.82739726], ['partial repair probe 1', [[2068, 8, 18], [2070, 8, 23]], 2.01369863], ['partial repair probe 2', [[2040, 5, 31], [2042, 11, 14]], 2.457534247], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2017, 12, 31], [2018, 5, 3]], 0.336986301], ['normal control 2', [[2052, 2, 28], [2059, 11, 6]], 7.68852459]], [['regression stub leap window 1', [[2030, 1, 31], [2034, 12, 4]], 4.84109589], ['regression stub leap window 2', [[1993, 2, 28], [2000, 11, 9]], 7.695890411], ['partial repair probe 1', [[2068, 3, 23], [2069, 12, 20]], 1.745205479], ['partial repair probe 2', [[2076, 7, 3], [2077, 8, 10]], 1.104109589], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2036, 9, 30], [2038, 9, 29]], 1.997260274], ['normal control 2', [[2019, 7, 18], [2024, 2, 29]], 4.617486339]], [['regression stub leap window 1', [[2004, 5, 26], [2008, 2, 29]], 3.761643836], ['regression stub leap window 2', [[2010, 12, 22], [2012, 2, 29]], 1.18630137], ['partial repair probe 1', [[2084, 8, 12], [2087, 8, 15]], 3.008219178], ['partial repair probe 2', [[2092, 4, 30], [2092, 10, 27]], 0.493150685], ['boundary control 1', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['boundary control 2', [[2019, 3, 1], [2024, 2, 29]], 4.99726776], ['normal control 1', [[2011, 3, 11], [2012, 2, 29]], 0.969945355], ['normal control 2', [[2012, 8, 30], [2015, 5, 30]], 2.747945205]], [['regression stub leap window 1', [[2001, 5, 31], [2009, 1, 23]], 7.649315068], ['regression stub leap window 2', [[2015, 2, 28], [2021, 8, 1]], 6.421917808], ['partial repair probe 1', [[2056, 2, 29], [2059, 9, 9]], 3.528767123], ['partial repair probe 2', [[2068, 3, 28], [2071, 10, 31]], 3.594520548], ['boundary control 1', [[2024, 1, 1], [2025, 1, 1]], 1.0], ['boundary control 2', [[2023, 2, 28], [2024, 2, 29]], 1.0], ['normal control 1', [[2046, 3, 23], [2046, 7, 28]], 0.347945205], ['normal control 2', [[2080, 6, 2], [2082, 2, 16]], 1.709589041]]]
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 stub leap window 1	`3.035616438`	`3.035616438`	Passed
regression stub leap window 2	`8.030136986`	`8.030136986`	Passed
partial repair probe 1	`3.254794521`	`3.254794521`	Passed
partial repair probe 2	`0.230136986`	`0.230136986`	Passed
boundary control 1	`4.99726776`	`4.99726776`	Passed
boundary control 2	`1.0`	`1.0`	Passed
normal control 1	`3.745901639`	`3.745901639`	Passed
normal control 2	`0.22739726`	`0.22739726`	Passed

SHA-256 / 26de7292414bf5fb5b2c4d0ad11e595a34e858457d12eccb6301694b96364d2e

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 2026-09-29T14:46:54.176905+00:00.

Case digest / 585fd63352f1a2ebf89d81f285cb97319f6b9609c2ae32e1ec0c0bfd49bf758a