COUPDAYS in WPS Spreadsheet: Measure the Whole Coupon Period

The official WPS COUPDAYS lesson documents this function for WPS Office for Windows Personal. The source was checked on October 19, 2026. It does not establish the first release, the last supported release, or an exact installed build. Availability in Mac, Linux, web, mobile and other editions remains unverified.

Examples below distinguish documented behavior from independently calculated controls. No results were captured from a new WPS execution for this guide. No native screenshot or cross-client compatibility test is claimed.

=COUPDAYS(settlement,maturity,frequency,[basis])

The documented basis values are 0 (US/NASD 30/360), 1 (actual/actual), 2 (actual/360), 3 (actual/365), and 4 (European 30/360). Set this explicitly when comparing workbooks. The common examples use basis 1 so the calendar-day arithmetic is transparent. A basis is a financial convention, not a date-display format.

Use DATE rather than ambiguous date text. The official lessons say arguments are truncated to integers; this guide deliberately uses whole-day dates and integer frequency/basis values. Fractional-input validation order and settlement exactly on a coupon date still need target-build testing.

Enter labels in A1:A4 and these values in B1:B4:

For the independent model, assume a regular semiannual schedule anchored on maturity, with coupons every May 15 and November 15. Around settlement, the previous coupon is November 15, 2025 and the next is May 15, 2026. These deliberately non-month-end dates avoid assuming any special end-of-month roll convention.

=COUPDAYS(B1,B2,B3,B4)

The expected 181 days is the actual calendar difference from November 15, 2025 to May 15, 2026. Settlement lies inside that period; it does not shorten the period to the remaining 25 days.

These expectations are calculated from the stated calendar model, not from native WPS execution. The model anchors regular coupons on maturity and steps backward by 12/frequency months. It counts actual date differences without adding an extra day for both endpoints.

Control from the official lesson

WPS also publishes settlement January 25, 2011, maturity November 15, 2011, frequency 2 and basis 1 with result 181. The surrounding regular period is November 15, 2010 to May 15, 2011. This is a documented example, separately from the original 2026 dataset above.

Keep settlement, maturity and basis unchanged, but evaluate the three permitted frequencies:

These are calendar-model controls, not captured WPS outputs. They show why annual days divided by frequency is not a general replacement for actual/actual period length. Quarterly periods do not all contain 365/4 actual days.

For the semiannual input, the elapsed fraction is =COUPDAYBS(B1,B2,B3,B4)/COUPDAYS(B1,B2,B3,B4), expected 156/181, approximately 0.8618784530. Keeping the denominator in its own cell makes it easier to spot an accidentally annual period in a semiannual calculation.

A regular coupon calculation does not reconstruct an irregular first or final coupon from issue-date information: there is no issue-date argument here. Nor does the function model business-day holidays, negotiated payment-date adjustments or a changing coupon schedule through the four arguments. Reconcile the assumed schedule with the security’s terms before using the result in a financial model.

Keep basis-1 calendar checks separate from 30/360 checks. Do not treat ordinary date subtraction as a universal substitute for a financial day-count convention. Month ends, leap days, settlement exactly on a coupon date and fractional arguments are useful additional native test cases, but no result for those untested boundaries is asserted here.

Use this function for a period denominator. COUPDAYBS supplies elapsed days, while COUPDAYSNC supplies remaining days. The three values answer different questions even though their arguments look the same.

• Official WPS COUPDAYS lesson (Chinese; Windows Personal documentation).

The documentation establishes the syntax and stated behavior. Original examples were checked independently under the assumptions stated alongside them; target-build execution remains a separate check. Version boundaries have not been inferred from another spreadsheet product.