Day Counts and Stub Coupons: Why Dates Change Bond Cash Flows

What is the difference between Actual/360 and 30/360?

Actual/360 uses the actual elapsed days specified by the accrual rule divided by 360. A 30/360 convention constructs a numerator using its month and day adjustments. Similar names do not imply identical algorithms: end-of-month and February treatment can differ between variants. Record the complete convention name and applicable contract language instead of treating every field containing 360 as the same calculation. The denominator alone is insufficient.

An issuer's historical note-program terms filed with the SEC provide an inspectable primary example of several named day-count alternatives and final-term selections. Those are that program's contractual definitions, not an SEC endorsement or universal bond convention. The MSRB glossary supplies additional market terminology. Neither source permits a modeler to choose whichever convention produces the most convenient reconciliation for a different instrument.

Build the calendar before calculating the fraction

Identify accrual start, accrual end, payment date and settlement date separately. A payment can move under a business-day rule without every associated accrual boundary moving identically. State which endpoint is included in the day count. A holiday calendar, roll convention and end-of-month rule may be needed to generate the schedule; a coupon frequency by itself does not identify all of these inputs.

Distinguish full coupon calculation from accrued interest at settlement. Both can involve a time fraction, but their intervals and entitlement rules need not match. A price difference attributed to yield may instead come from inconsistent dates or accrual treatment. Reconcile the scheduled amount and settlement bridge before changing the discount rate to force agreement with an unexplained target price. That adjustment would conceal the underlying specification problem.

Reperform one explicitly synthetic interval

Assume USD 100,000 principal, a 6% simple annual coupon rate and an interval from 15 January to 15 March 2027, excluding the start and including the end. The actual elapsed count is 59 days. For this mid-month interval, the stated 30/360 calculation gives 60 days. Interest is therefore USD 983.33 under Actual/360, USD 969.86 under Actual/365 Fixed and USD 1,000 under the illustrated 30/360 fraction, rounded only at display.

The amounts follow principal times rate times year fraction. They are alternative contractual illustrations, not three estimates to average and not a quote for a real bond. The same dates deliberately avoid difficult month-end adjustments, so the example does not test every 30/360 variant. Independent calendar counting and arithmetic both matter: a correct multiplication using the wrong numerator would still produce a precisely displayed but inappropriate amount.

Identify what makes a coupon a stub

A short or long first or last coupon can arise when issue or maturity does not align with the regular schedule. Calculating its amount requires the selected convention and sometimes hypothetical regular reference periods. Actual/Actual is not one universal rule either. Ask whether the contract identifies an ICMA-style reference-period calculation, an ISDA-style year treatment or another definition, and preserve that identity in the implementation.

Do not approximate an irregular coupon by dividing an annual coupon into equal calendar-month pieces unless the stated terms actually require that method. A long period may need to be split across reference periods or calendar years. Payment frequency, accrual method and discounting convention remain separate decisions. The historical SEC-filed terms illustrate why final terms and determination dates can be needed alongside the general note-program definitions.

Test boundaries where conventions diverge

A focused test set should include leap days, February month-ends, dates on the thirty-first, adjusted payment dates and irregular first and final periods. Check counts and cash amounts independently of the pricing routine. A test covering only regular mid-month dates can pass while never exercising the adjustments that distinguish competing conventions. Preserve the expected schedule and its source rather than validating one library against another undocumented default.

Useful review questions are concrete: which dates were adjusted, which were not, and why does this payment differ from the regular amount? If a system supplies an accrued-interest result without its convention, leave the comparison unresolved. An instrument identifier or a broad category such as corporate bond may help locate terms, but neither replaces the specific calculation rule governing the actual security and interval being modeled.

Keep this extension outside the regular calculator

The site's calculator assumes regular coupons, integer-year maturities and coupon-date valuation. It has no date-generation, holiday, stub or accrued-interest engine. A frequency change does not add these capabilities. These separate illustrations explain missing inputs; they do not establish supported settlement valuation.

Check library defaults, accrual boundaries and premature fraction rounding before labeling a discrepancy a pricing error. Two systems may describe different contracts. Preserve the convention, dates and rounding policy so another reviewer can identify the source of disagreement without inferring instrument-specific accuracy.

Primary references

1. SEC EDGAR: Historical issuer note-program terms, day-count definitions and final-term fields; issuer contract, not SEC guidance
2. MSRB: Glossary of Municipal Securities Terms, public full text, day-count terminology