Nominal Rate of Discount

A nominal rate of discount \(d^{(m)}\) is convertible m times per year: each sub-period is discounted by \(d^{(m)}/m\), and compounding all m of those discounts gives \((1-d^{(m)}/m)^m = v = 1-d\), the annual discount factor.

By the end you'll be able to convert freely among effective interest i, effective discount d, nominal interest \(i^{(m)}\), and nominal discount \(d^{(m)}\) for any compounding frequency m.

Predict: for the same effective rate and the same m, which is larger — \(i^{(m)}\) or \(d^{(m)}\)? Drag m and check whether the gap shrinks.

The effective annual rate here is fixed at \(i = 8\%\) (so \(d \approx 7.41\%\) and the force of interest \(\delta \approx 7.70\%\) never move). Drag m and watch the nominal interest bar fall and the nominal discount bar rise, both squeezing toward the dashed force-of-interest line — they meet only in the limit, but \(i^{(m)}\) never dips below \(d^{(m)}\) for any finite m. Hover any bar for its exact value.

1241252365

i(4) = 7.7706% · d(4) = 7.6225% · δ = 7.6961%

Check: (1 − d(4)/4)4 = 0.925926 — matches v = 1/(1+i) = 0.925926. That equality holds exactly for every m, by construction of d(4) from v.

Four names for the same growth at i = 8%, current m
i(m) (nominal interest) d(m) (nominal discount) δ (force of interest)
Formal

A nominal discount \(d^{(m)}\) takes \(d^{(m)}/m\) off the balance at the start of each of m sub-periods, so over a year those m discounts compose to the annual discount factor: \((1-d^{(m)}/m)^m = v = 1-d\). Solve either direction: \(d = 1-(1-d^{(m)}/m)^m\) (nominal → effective), or \(d^{(m)} = m\big[1-(1-d)^{1/m}\big] = m\big[1-v^{1/m}\big]\) (effective → nominal). Because effective interest i, effective discount d, nominal interest \(i^{(m)}\), and nominal discount \(d^{(m)}\) are all just different names for the same one-year growth factor \(1+i = v^{-1}\), you can convert *any* of them to *any* other by routing through \(1+i\) (or v) — never mix a nominal rate directly into an effective-only identity like \(i-d=id\); that one is for the effective pair only. The cleanest same-m shortcut links the two nominal rates directly: \(i^{(m)}/m - d^{(m)}/m = (i^{(m)}/m)(d^{(m)}/m)\), the m-thly echo of \(i-d=id\).

Applied

Discount-in-advance shows up whenever a lender deducts interest up front instead of charging it at the end — a bank that sells you a $1000 face-value note for less than $1000 today, or a T-bill quoted at a discount rate rather than a yield. Whenever you see a rate advertised as a "discount rate convertible monthly" or similar, don't plug it straight into a present-value formula built for effective i — convert it to v first, exactly as you would a nominal interest rate.

Worked example

A rate of discount of 8% convertible quarterly means \(d^{(4)} = 0.08\), so each quarter discounts by \(d^{(4)}/4 = 0.02\). Annual discount factor: \(v = (1-0.02)^4 = (0.98)^4 = 0.922368\). Effective annual discount: \(d = 1-v = \) 7.7632%. Effective annual interest: \(i = 1/v-1 = \) 8.4166%.

Your turn

Same idea, a different frequency: a nominal discount of 6% convertible monthly, \(d^{(12)} = 0.06\). Monthly discount: \(d^{(12)}/12 = 0.005\). Raising \((1-0.005)\) to the 12th power: \((0.995)^{12} = 0.941623\), so the effective annual discount is \(d = 1-0.941623 = \) 5.8377%. Now finish it: \(i = 1/v-1 = 1/0.941623 - 1 = \) ____

Reveal the answer

\(i = 1/0.941623 - 1 = 1.061996 - 1 = \) 6.1996% — a little above d, exactly as it must be. Set m = 12 on the slider above (between the 4 and 52 ticks) and compare i(12) and d(12) against these effective values.

More info — why \(i^{(m)}\) and \(d^{(m)}\) bracket δ instead of equaling it

At m = 1, \(i^{(1)} = i\) and \(d^{(1)} = d\) — the widest possible gap, since a single yearly sub-period is the least compounding you can do. Every time you double m, each nominal rate gets one step closer to the continuous limit: \(i^{(m)} \downarrow \delta\) and \(d^{(m)} \uparrow \delta\) as \(m \to \infty\), because \((1+i^{(m)}/m)^m\) and \((1-d^{(m)}/m)^m\) both converge to \(e^{\delta} = 1+i\). They never actually cross — \(i^{(m)} > \delta > d^{(m)}\) holds for every finite m — because interest-in-arrears can never be cheaper than the same growth charged continuously, and discount-in-advance can never be more expensive than it. See the Binghamton handout in Dive deeper for the full derivation.

Check your understanding

Question 1 of 4

A rate is quoted as \(d^{(2)} = 10\%\) convertible semiannually. What are the equivalent effective annual discount rate \(d\) and effective annual interest rate \(i\)?

Question 2 of 4

Holding the effective annual rate fixed, what happens to \(i^{(m)}\) and \(d^{(m)}\) as the compounding frequency \(m\) grows without bound?

Question 3 of 4

A rate is quoted as \(i^{(4)} = 8\%\) convertible quarterly (nominal interest). What is the equivalent \(d^{(4)}\) (nominal discount), also convertible quarterly?

Question 4 of 4

At a nominal discount rate \(d^{(2)} = 9\%\) convertible semiannually, what is the present value of an \$8{,}000 payment due in one year?

Recap

  • A nominal discount \(d^{(m)}\) is applied in advance, m times per year: \((1-d^{(m)}/m)^m = v = 1-d\).
  • Nominal → effective: \(d = 1-(1-d^{(m)}/m)^m\). Effective → nominal: \(d^{(m)} = m\big[1-v^{1/m}\big]\).
  • Effective interest i, effective discount d, nominal interest \(i^{(m)}\), and nominal discount \(d^{(m)}\) are four names for one growth factor \(1+i\); convert through \(1+i\) or v, never mix a nominal rate into an effective-only identity.
  • Same-m shortcut: \(i^{(m)}/m - d^{(m)}/m = (i^{(m)}/m)(d^{(m)}/m)\).
  • For any finite m, \(i^{(m)} > \delta > d^{(m)}\); both converge to the force of interest δ as \(m \to \infty\), narrowing the gap but never closing it at a finite m.

Dive deeper

Sources

  • Nominal Rates of Discount and Converting Among All Four Rates