INTRATE Function (Gnumeric)

• Introduced: Recorded as implemented in the grouped Gnumeric 0.33, 0.34 and 0.35 release notes. The exact first release within that interval and the release date are not established.
• Documented support: Registered in Gnumeric 1.12.62. No retirement is established.
• Last functioning version: Unverified. Source registration is not a native application test.
• Dates and scope: Pinned official source was retained from the October 13–14, 2026 audit. This complete guide and independent calculations were checked October 15. There was no new source retrieval, native engine execution or GUI test.

INTRATE calculates (redemption-investment)/investment × B/A. It annualizes a single holding-period gain relative to the original investment. It does not solve a sequence of irregular cash flows, reinvest interim proceeds, or compound a partial-year gain into a yearly effective return. The output is a rate, so 0.1 can be displayed as 10%.

=INTRATE(settlement,maturity,investment,redemption,[basis])

investment is the amount paid at settlement; redemption is the amount received at maturity, expressed in the same currency and scale. Both dates and both amounts are required. basis is optional with default 0. There is no periodic-payment argument.

Use valid worksheet dates produced by DATE, rather than locale-ambiguous text. The registration requires four numeric/date arguments and permits one additional basis argument. The shared basis converter truncates nonnegative fractional inputs below 6; use the documented integer values explicitly instead of relying on that conversion.

Create SingleReturn with B2 =DATE(2026,1,15), B3 =DATE(2026,7,15), B4 1000, B5 1050 and B6 0. B8 =INTRATE(B2,B3,B4,B5,B6) has expected result 0.1. The holding-period gain in B9 =(B5-B4)/B4 is 0.05. With the helper’s 180/360 time fraction, B10 =B9/(180/360) agrees with B8. The 50-unit gain is received once, at maturity, in this simplified dataset.

Each row is independent and uses whole DATE values. Numeric results are source-derived reference expectations rounded to twelve decimal places where needed. Error results follow explicit callback guards, not a native error-display test.

To distinguish simple annualization from compounding, calculate =(B5/B4)^(360/180)-1 in B11. Its expected result is 0.1025, whereas INTRATE remains 0.1. The extra 0.0025 comes from a hypothetical second reinvested half-year, which INTRATE never assumes. Change redemption to 1100: the simple annualized return becomes 0.2 while the compounded illustration becomes 0.21. Label both measures rather than presenting them as interchangeable.

The basis-1 denominator is tied to the settlement year in this callback. For settlement July 1, 2023, maturity June 30, 2024, investment 1000 and redemption 1100, there are 365 actual days and the settlement-year denominator is 365, so the result is 0.1. YIELDDISC uses a different year-fraction helper and can return approximately 0.100273972603 for these same amounts and dates. This is why a same-input identity is only claimed for the controlled basis-0 fixture, not every calendar span.

These results follow the local days_monthly_basis helper, not a borrowed implementation of a similarly named financial function. For basis 4 it uses 30*month_difference + day_difference. Basis 0 starts with that expression, then subtracts 2 for a non-leap-year February settlement when maturity is outside February in the same year, or 1 in a leap year. Thus February 15 to March 15, 2026 counts as 28 under basis 0 but 30 under basis 4. The helper does not apply a universal 30th-day clamp to all month ends.

For bases 1–3 it uses elapsed date serial days. The annual denominator helper returns 360 for bases 0, 2 and 4, 365 for basis 3, and the settlement year’s 365 or 366 for basis 1. Accordingly January 15 to July 15, 2026 has 181 actual days but 180 days under the illustrated 30/360 helpers. These literal source rules explain the examples; they are not a promise of interchangeable day-count conventions across products.

Zero investment returns #NUM!, as do nonpositive counted days and an invalid basis/year denominator. The source does not require investment or redemption to be positive. A redemption of zero with investment 1000 produces -2 in the main half-year fixture; paired negative amounts can also return a rate. These are source formula outcomes, not a suitable sign convention for an ordinary purchase. Establish business input constraints before calculation.

A missing interim payment cannot be repaired by changing the calendar basis: this model accepts only beginning and ending amounts. Use a cash-flow model for multiple payments. Avoid integer-looking rate formatting that hides the distinction between 0.1 and 10.

Use actual numeric inputs for these fixtures. Blank cells, numeric text, localized formula entry, time-of-day fractions, historical date-system anomalies and floating-point extremes were not executed. Keep the original inputs and a separately calculated time fraction available when investigating a disagreement. The model verifies the stated contract and examples; it does not certify an installed release or a financial market’s required convention.

• Function-specific Gnumeric help
• Complete worksheet callback
• Official calendar-basis descriptions
• Gnumeric function catalogue

DISC, YIELDDISC, RRI.