Reader question: Why can a bank loan with a 5% contractual coupon produce interest revenue at a different percentage, and how does an accounting system turn the loan’s cash flows, fees and costs into an amortised-cost schedule?
The answer is the effective interest method. Rather than treating the coupon as the complete economics of the instrument, the algorithm solves for a rate that discounts the relevant expected contractual cash flows to the instrument’s initial carrying amount. Fees, points, transaction costs, premiums and discounts that belong inside the effective-interest calculation are therefore spread through the instrument’s life instead of being treated as unrelated arithmetic.
Page role and boundary
This page explains accounting mathematics and numerical implementation. It is not personal financial advice, tax advice or an accounting opinion on a real contract. IFRS classification, derecognition, impairment and contractual interpretation can be fact-specific. The governing standard and professional judgement remain authoritative.
The core equation: find the rate that makes the cash flows fit
Suppose an instrument begins with carrying amount B₀, produces cash flows CF₁, CF₂, …, CFₙ, and has effective interest rate r. In a simple annual-period illustration, r solves:
B₀ = CF₁/(1+r) + CF₂/(1+r)² + … + CFₙ/(1+r)ⁿ.
That is an internal-rate-of-return problem. The computer is not “given” the EIR in many cases; it must solve a nonlinear equation. Spreadsheet goal-seek functions, Newton–Raphson, secant methods and bracketed root-finders can all solve the equation if they are implemented with suitable safeguards.
Once r has been found, the carrying amount evolves recursively:
Bt = Bt−1 + effective interest − contractual cash received.
For a financial asset that is not subject to one of the credit-impaired exceptions, effective interest for a period is broadly the opening gross carrying amount multiplied by the effective rate for that period. The schedule gradually amortises the initial difference created by qualifying fees, costs, premiums or discounts.
A numerical example: why a 5% coupon can have a 5.37% EIR
Consider a deliberately simplified three-year bullet loan. The bank advances $100,000. The borrower pays a $1,500 origination fee that is integral to the effective interest rate, and the bank incurs $500 of directly attributable transaction costs that qualify for inclusion. For illustration, the initial net carrying amount is therefore $99,000. The contractual coupon is 5% per year, with $5,000 received at the end of years 1 and 2 and $105,000 at the end of year 3.
The effective-rate equation is:
99,000 = 5,000/(1+r) + 5,000/(1+r)² + 105,000/(1+r)³.
The numerical solution is approximately 5.37%. That rate is above the 5% coupon because the net initial carrying amount is lower than the contractual principal that will ultimately be collected. The EIR schedule gradually recognises that difference over time.
This example is only a mechanism demonstration. Real instruments can have monthly dates, irregular periods, floating coupons, options, multiple fees, prepayments and contractual conditions that require more careful cash-flow modelling.
The first diagnostic question: which cash flows belong inside EIR?
The numerical solver is often the easy part. The harder question is scope. IFRS 9 distinguishes fees and costs that are integral to the effective interest rate from items that represent separate services or otherwise receive different accounting treatment. Transaction costs also have a defined meaning: they are incremental costs directly attributable to acquiring, issuing or disposing of the financial instrument.
This creates a useful computational rule: never begin by feeding every cash item into the root-finder. First classify the item. A model that perfectly solves the wrong set of cash flows is still wrong.
Dates matter: the hidden mathematics of day counts
Simple textbook formulas assume equally spaced periods. Banking systems rarely have that luxury. Settlement dates, leap years, month ends, business-day adjustments and irregular first or final coupons can alter the exponent or accrual fraction applied to each cash flow.
A production engine therefore needs a consistent time map: contractual date, adjusted payment date, accrual start and end, day-count convention, compounding convention and periodic-rate conversion. A one-day shift can be immaterial for a small instrument but significant across millions of records. The correct diagnostic is not “does the final balance look close?” but “can every cash flow and accrual fraction be reproduced from the contract and accounting policy?”
Floating-rate instruments are not just fixed-rate loans with changing coupons
A floating-rate instrument creates a second layer. Cash flows change as the reference rate resets. IFRS 9 contains requirements for periodic re-estimation of cash flows to reflect movements in market rates, and the treatment is not identical to an unrelated contractual modification. The engine must distinguish a scheduled reset built into the instrument from a genuine change in contractual terms.
That distinction is a good example of evidence polarity. If a payment changed because SOFR reset exactly as the original contract specified, that evidence supports the “normal variable-rate re-estimation” path. If the borrower and lender renegotiated the spread, maturity or payment profile, the modification logic may instead be relevant.
Modifications: preserve the old rate or create a new instrument?
When contractual cash flows change, the algorithm must first determine whether the change causes derecognition. If it does not, IFRS 9 can require the carrying amount to be recalculated using the modified contractual cash flows discounted at the original effective interest rate, with the resulting adjustment recognised according to the applicable requirements. For financial liabilities, the well-known “10% test” is part of assessing whether terms are substantially different, but the full derecognition analysis is broader than one number.
This gives a clean computational sequence: classify the event → decide whether the original instrument survives → determine which rate applies → rebuild the cash-flow present value → recognise the resulting difference correctly. Skipping the classification step and merely “re-solving for a new IRR” can erase the economics that the accounting rules are designed to preserve.
Credit impairment changes the base to which interest is applied
For many financial assets, interest revenue is calculated using the effective interest rate applied to the gross carrying amount. When an asset becomes credit-impaired, IFRS 9 contains an important exception: interest revenue is calculated by applying the EIR to the amortised cost — effectively the carrying amount after the loss allowance — rather than simply continuing on the gross amount. Purchased or originated credit-impaired assets use a credit-adjusted effective interest rate from initial recognition.
This is why an effective-interest engine cannot be isolated from the impairment engine. The rate, carrying amount, loss allowance and credit status interact. A change in stage or credit-impaired status can affect the interest base even when the contractual coupon has not changed.
Failure modes that make a plausible schedule wrong
- Wrong initial carrying amount. A fee is included twice, omitted, or classified as integral when it relates to a separate service.
- Wrong cash-flow sign. A root-finder can converge to nonsense if advances and receipts are signed inconsistently.
- Multiple mathematical roots. Unusual cash-flow patterns can produce more than one IRR. The effective-interest problem then needs contractual and economic interpretation, not blind acceptance of the first numerical root.
- No root in the permitted range. This may reveal incorrect cash-flow extraction, an impossible initial amount or a solver bracket that is too narrow.
- Date drift. The contract engine and accounting engine use different business-day or day-count conventions.
- Modification misclassification. A renegotiation is treated as a normal rate reset, or a normal reset is treated as a new instrument.
- Credit-status mismatch. Interest is applied to gross carrying amount when the applicable credit-impaired treatment requires amortised cost, or vice versa.
Diagnostics and falsifiers
A strong EIR engine should be able to pass a set of independent checks. First, discount its own projected cash flows using the solved rate: the result should reproduce the target carrying amount within a defined numerical tolerance. Second, roll the amortisation schedule to maturity: absent modifications, impairment effects or write-offs, the residual should reconcile to the contractual terminal amount. Third, rerun the calculation using a different root-finding method. Material disagreement is evidence of instability or multiple-root behaviour.
Confidence should be reduced if the solver is highly sensitive to starting guesses, if small date changes produce unexplained large rate changes, if cash-flow extraction cannot be reconciled to the legal contract, or if fees are classified differently by otherwise identical instruments. Those are not “IT exceptions”; they are signals that the accounting representation may be wrong.
Alternatives and comparison models
Coupon yield, simple interest, nominal annual percentage rates, yield to maturity and effective interest rate can all answer different questions. A useful model-control habit is to calculate at least one simpler comparator. If a loan with a large upfront integral fee produces an EIR almost identical to its coupon, investigate why. The comparator does not replace IFRS 9; it acts as a reasonableness check.
Connections across Bukit Timah Tutor
The arithmetic begins with the same recurrence logic used in loan amortisation and compound interest, but the accounting problem adds fee classification, carrying amount and root-finding. It connects directly to IFRS 9 expected-credit-loss modelling because credit impairment changes how interest revenue is calculated. For variable-rate mechanics, compare the operational cash-flow side with SOFR floating-rate loan calculations. The general lesson — correct mathematics can still be the wrong model — also appears in bank model validation.
Current-status note and update triggers
As of 29 August 2026, the IASB is actively considering aspects of amortised-cost measurement. In April 2026 it tentatively decided to amend IFRS 9 requirements concerning subsequent changes to the effective interest rate for certain re-estimations of contractual cash flows. A tentative Board decision is not the same as an issued amendment. This article should therefore be re-checked when the IASB completes that project, publishes an exposure draft or final amendment affecting the effective-interest method, or when related IFRS Interpretations Committee guidance changes.
Primary and high-quality references
- IFRS Foundation, IFRS 9 Financial Instruments (issued-standard text available through the IFRS Foundation; users should check the current authoritative version and licensing terms).
- IFRS Interpretations Committee, Modification or exchange of financial liabilities that do not result in derecognition.
- IFRS Foundation, IFRIC Update, March 2019, including credit-impaired interest-revenue discussion.
- IFRS Foundation, IASB Update, April 2026, for the current amortised-cost measurement project status.
Educational boundary: The examples are deliberately simplified to reveal the mathematics. Real accounting depends on the contractual facts, the applicable IFRS requirements and professional judgement.