Quant Memo
Foundational

Day Count Conventions

Day count conventions are the fixed rules bond and loan markets use to count how many days have passed between two dates for the purpose of calculating accrued interest, and different markets use different rules that produce slightly different answers for the exact same dates.

Interest accrues continuously between coupon payments, but "how many days between two dates" is not as obvious a question as it sounds once you need to compare instruments across markets. Some conventions treat every month as having exactly 30 days regardless of the calendar; others count the actual number of calendar days; some divide by an assumed 360-day year, others by the true 365 (or 366). A day count convention is simply the fixed rule a bond, loan, or swap contract specifies upfront for doing this counting, written as a ratio like "30/360" or "Actual/365."

The reason this matters is that the same stated annual rate, applied to the same two dates, produces a slightly different dollar amount of accrued interest depending on which convention is in force — so day count conventions have to be specified explicitly in every contract and matched carefully when comparing yields across instruments that use different ones.

A worked example

A bond paying 6% annual interest, using 30/360 counting, treats a period from January 1 to March 1 as exactly 60 days (two 30-day months) and calculates accrued interest as 6%×60/360=1%6\% \times 60/360 = 1\% of face value. The same bond under Actual/365 counting would use the true 59 calendar days between those dates, giving 6%×59/3650.97%6\% \times 59/365 \approx 0.97\% — a small but real difference that matters when settling a trade or comparing yields quoted under different conventions.

A day count convention is the contractually specified rule for counting days between dates when calculating accrued interest; the same rate and dates can produce different accrued-interest amounts under 30/360 versus Actual/365 or Actual/360, so the convention must always be matched, never assumed.

Related concepts

Practice in interviews

Further reading

  • Fabozzi, Bond Markets, Analysis, and Strategies, ch. 3
ShareTwitterLinkedIn