Interest Ledger Calculator

A payoff quoted without its date and its basis is already wrong

The large figure is what is owed on the date named beside it, and it is the sum of exactly two things: the principal still outstanding and the interest that has accrued but not been paid. Underneath it sits the per diemThe interest one further day adds at the current balance — the number that tells you what a delayed settlement costs. — what one further day adds at the current balance. A settlement figure without a per diem goes stale the moment it is quoted, because the amount owed on Tuesday is not the amount owed on Friday.

That is what the per diem is for: outstanding principal times the daily rate under the active convention. It is the number that makes a settlement negotiable, because it turns “we will pay next week” into a rupee amount rather than a promise, and it is what stops a payoff figure from quietly going stale — a total quoted for one date is wrong on any other, and the per diem says by how much. On the ledger the tool opens with, that is ₹450 for every day the money stays out.

The line of small text under the total is the method disclosure, and it is also the control that changes it. It names the compounding, the rate, the day-count conventionThe agreed rule for how many days a period contains and how long a year is — the single most common reason two honest people compute two different interest totals., the allocation rule, whether arrears themselves earn interest, and how rounding is handled. Nothing about how the number was produced is hidden behind a tooltip, because the person you are showing it to will ask.

Compounding is the entry on that line worth checking first. The default is simple, and nothing compounds unless you ask for it, because a note quietly compounding is how a modest rate becomes an unpayable one. If your agreement has monthly, quarterly or annual rests, switch Compounding to match — the setting is then named on the screen, in the statement and in the workbook, so nobody can be surprised by it later.

The Reconciled mark is not decoration. Eight arithmetic identities are checked on every recalculation — that the segment days add up to the term, that the interest column adds up to the total, that principal advanced less principal repaid equals the balance, and so on. If any of them failed, the tool would say so instead of showing you a number.

A repayment larger than the whole balance is not an error either, and it is not quietly absorbed: the borrower goes into credit and the tool says so, because the lender owes the excess back. No interest is charged on a negative balance unless you deliberately switch that on, since charging the lender interest on money they are holding is a separate agreement, not an arithmetic default.

What it cannot tell you is whether the figure is collectable. It computes; it does not adjudicate, and it will faithfully produce a total under whatever basis you set, including one that no court in your jurisdiction would enforce. Interest-rate ceilings, the treatment of penal charges, and the statutory power to reopen an unfair bargain all vary by place and by the status of the lender, and none of that is modelled here.

A loan is a list of dated events, not three inputs and a formula

What the engine holds is a sorted list of dated events, and it cuts the timeline at each one, producing a set of accrual segments in which the balance and the rate are both constant. Within a segment the arithmetic is a single line:

Repayments therefore go in on whatever date they actually happened. A private loan is repaid in lumps whenever the borrower has cash, so a payment on the 3rd, another on the 17th and a rate change on the 29th all land where they belong, and the balance that stood between them is the balance the interest is charged on. You can add an adjustment, or a second advance after the balance has already gone to zero, without breaking the arithmetic.

Events on the same date are applied in a fixed order — advances, then the accrual cut, then repayments, then rate changes — so money advanced and repaid on the same day behaves sensibly and a rate change governs only the time in front of it. Inserting a backdated row does not patch the balance forward; the whole ledger is recomputed from the origin, which is the only way a late-discovered receipt produces the same answer as one entered in order.

Money is never a floating-point number here. Every balance, rate and day-fraction is carried as an exact fraction and rounded exactly once, at display. That is why the segment column adds up to the total to the paisa rather than to within a rounding error, and it is what makes the reconciliation checks meaningful rather than decorative.

It is also why a hand-kept bahi-khata and this tool can differ by a few rupees without either being wrong. A book rounds each entry to the whole rupee as it is written down; software carries the full fraction and rounds once at the end. Switch Rounding to “Rounded each segment” and the tool keeps the book’s habit, and shows you exactly what that habit costs against carrying full precision.

₹15,00,000 lent, one part-payment, a ₹11,00,575 settlement

A lender advances ₹15,00,000 on 29-Nov-2025 at 1.35% per month, simple, for 6 months — maturing 29-May-2026. On 26-Feb-2026 the borrower returns ₹5,00,000 of principal. Under 30/360 (US) the loan splits into two segments:

Principal outstanding at maturity is ₹10,00,000 and the interest is ₹1,00,575, so the settlement is ₹11,00,575. Each further day adds ₹450. Note the divergence the tool reports quietly underneath: that term is 181 actual calendar days but 180 under 30/360 (US). Both are correct inside their own system, and the tool refuses to pick one silently.

The same 6 months is 180 days under one convention and 181 under another

A day-count convention is the agreed answer to two questions: how many days are in this period, and how many days are in a year. It sounds like bookkeeping trivia and it is the single most common reason two honest people compute two different totals. Here is the same worked example above, recomputed under each convention:

30/360 (US) makes every month 30 days and every year 360. Dates on the 31st are treated as the 30th, and an end-of-February date is treated as the 30th so that February is not short-changed. It is the default here because a rate quoted per month implies months of equal length. 30E/360 is the Eurobond variant: any 31st becomes a 30th unconditionally, with no February rule, so a February period is genuinely shorter than under the US rule.

The Actual family counts the real calendar. Actual/365 divides by a fixed 365 whatever the year. Actual/360 divides by 360 — the money-market basis, which charges about 1.39% more than Actual/365 over the same period, because 365 ÷ 360 is 1.0139 and the days do not shrink to match. Actual/Actual splits the accrual at each 1 January and divides each part by the true length of its own year, so a leap year is genuinely 366 days long.

So when a lender produces a different total from the same numbers, the convention is almost always the reason, and the fix is agreement rather than arithmetic. 29-Nov-2025 to 29-May-2026 is 180 days if every month is 30 days and 181 actual calendar days. The table above prices that difference for each convention, and a single day at the closing balance here is ₹450; over a longer loan the gap grows into real money. None of the five is wrong — they are different agreements, and setting both sides to the same one here is what makes two totals converge.

None of that helps if the two of you disagree about the events themselves. The arithmetic is only as good as the dates and amounts in the ledger, and a disputed receipt is a disputed receipt whatever the segment table says. Where the disagreement is about the basis rather than the facts, the table above is the useful screen: it shows what each competing reading is worth in money, which is usually a faster route to a settlement than arguing about principle.

When a ₹5,000 payment moves only ₹2,000 of principal

When a payment arrives, something has to decide what it pays off. The United States RuleThe standard treatment for simple-interest notes: a payment clears the interest accrued to that date first, and only what is left reduces principal. — the default here, and the standard treatment for simple-interest notes — applies the payment first to the interest accrued up to that date, and only the remainder reduces principal. If the payment is smaller than the accrued interest, the principal does not move at all and the shortfall carries as arrearsInterest that has fallen due and was not paid — carried as a separate balance rather than added to principal.. On the conventional treatment those arrears do not themselves earn interest, which is precisely what stops a simple-interest note from quietly becoming a compounding one.

The Merchant’s RuleThe single date every amount in a settlement is valued at, so sums paid on different dates can be compared. does no running allocation at all. Interest runs on the full advance to a single focal date, each payment separately earns interest from its own date to that date, and the two sides are netted. Take a small ledger: ₹1,00,000 lent on 01-Jan-2026 at 1% a month for six months, with a single ₹5,000 payment three months in, when ₹3,000 of interest has accrued.

• US Rule. ₹3,000 of the payment clears the accrued interest and only ₹2,000 comes off the principal. Total payable at maturity: ₹1,00,940.
• Merchant’s Rule. Interest runs on the whole ₹1,00,000 for the full term, and the payment earns its own interest credit for the remaining half. Total payable: ₹1,00,850.

The gap is ₹90 — exactly the interest that the ₹3,000 sent to interest under the US Rule would have saved had it reduced the principal instead. That is the whole difference between the two rules, in one number. A third option, principal-first, is what a payer means when they hand over money and say it is for the principal; under simple interest with no arrears it produces the same total as the Merchant’s Rule, from the opposite direction.

Which rule applies is a question about what was agreed, not about which one is correct. In informal lending the payer often hands the money over and says it is for the principal — tag that row “Principal only” and the tool applies it that way regardless of the global setting, which is exactly what the worked ledger above does. Leave the row on “Auto” and the US Rule governs: a payment smaller than the interest accrued to its date moves no principal at all, and the shortfall sits as arrears until something clears it.

The Merchant’s Rule is here because a counterparty may be using it, not because it is recommended. It is the older netting method, and under simple interest with no arrears it gives exactly the same total as applying payments to principal first. It comes out lower than the US Rule whenever a payment failed to cover its own interest — which is the thing to check before agreeing to switch to it, and the reason to run both before conceding either.

₹20,925 saved, and the two ways the tool proves it

Interest on a reducing balance is rent on money, charged by the day. Return part of the principal and you stop paying rent on that part from that day forward — nothing more mysterious than that. The saving is therefore the amount returned, multiplied by the days it would otherwise have stayed outstanding, multiplied by the daily rate.

In the worked example above the borrower returned ₹5,00,000 with 93 days of the term left, and the daily rent on that slice was ₹225. So: ₹5,00,000 unpaid for 93 days at ₹225/day = ₹20,925.

The tool does not take that on trust. It computes the saving the hard way first — by deleting the repayment and re-running the entire ledger to the same end date, which gives ₹1,21,500 of interest instead of ₹1,00,575, a difference of ₹20,925 — and only then checks it against the one-line formula. A saving figure quoted without its proof is an assertion; quoted with one, it is arithmetic the other party can redo on the back of an envelope.

So when the panel shows two numbers there, they are not two answers to one question. The re-run is the authoritative figure; the one-line formula is the independent check, and when the two agree the saving stops being a claim and becomes something you can redo on paper. Where a later payment would re-allocate once the first is removed, the two methods legitimately diverge, and the tool labels the direct formula an estimate rather than pretending otherwise.

Send the workbook when they do not believe you

A settlement travels through four different channels and they want four different things, so the exports are genuinely different artifacts rather than one file under four names.

None of the four costs you any privacy. The whole ledger is computed in the page, and the workbook, the statement and the CSVs are all built in the browser too — nothing is uploaded to produce them. The share link carries the figures inside the URL rather than on a server, which is why the tool warns you before you copy it: anyone holding that link can read your numbers.

One limitation belongs to the workbook rather than to the engine. Under Actual/Actual the tool splits a segment that crosses 1 January and divides each part by its own year length; the exported formulas use the year of the segment start throughout, so a segment spanning a year end can differ there by a few rupees. Every other convention is exact in the file. Adding a brand-new transaction to the workbook, or moving one past another, changes where the segments fall — regenerate it from the tool rather than editing around it.

Eight words to settle before you settle the loan

Accrual

Interest earned by the passage of time, whether or not it has been billed or paid.

Reducing balance

Interest charged on what is still outstanding rather than on the original sum, so a repayment lowers every future charge.

Broken period

A stretch shorter than a whole month or quarter, which is what most segments in a real private loan are.

Rest

The point at which unpaid interest is folded into the principal and starts earning interest itself. Monthly rests are the regulated norm for banks; a private loan has none unless the parties agree one.

Arrears

Interest that fell due and was not paid, tracked separately so the principal is left alone.

Focal date

The single date every amount is valued at, so sums paid on different dates can be compared.

Per diem

What one further day adds at the balance as it stands.

Day-count convention

The agreed rule for how long a period and a year are — the setting that most often makes two honest calculations disagree.

Sources and methodology

The allocation default comes from the decision the United States Rule is named after; the two 30/360 variants follow the documented NASD and European methods; the contrast with regulated lending comes from the RBI direction requiring monthly rests on advances; and the caution about enforceability comes from the statute that lets a court reopen a usurious bargain. Links open in a new tab.

• Story v. Livingston, 38 U.S. (13 Pet.) 359 (1839) — the decision the United States Rule comes from, directing that receipts be applied first to interest already accrued and only then to the discharge of principal, which is exactly the default allocation used hereUnited States Reports, Library of Congress · 1839 · verified 2026-09-01
• DAYS360 function — documents the two 30/360 day-count methods this tool implements, including the US (NASD) rule that rolls a month-end date and the European rule that simply treats any 31st as a 30thMicrosoft · 2026 · verified 2026-09-01
• Master Direction — Reserve Bank of India (Interest Rate on Advances) Directions, 2016 — section 4 requires regulated lenders to charge interest on advances at monthly rests, which is the contrast this tool makes visible: an informal loan carries no such default and stays simple unless the parties agree otherwiseReserve Bank of India · updated 1 October 2025 · verified 2026-09-01
• The Usurious Loans Act, 1918 (as modified to 3 December 2018) — gives courts power to reopen a transaction where the interest is excessive and the dealing substantially unfair, which is why this tool computes a figure rather than asserting that a figure is enforceableGovernment of India · 1918 · verified 2026-09-01

Related calculators

Tools that share the same interest arithmetic from other angles:

More in Finance, or browse all calculators.