Fin Maverick
Foundations VocabularyAccounting & ReportingEconomics & MacroQuant Methods & ProgrammingBusiness & Company AnalysisCorporate Finance & ValuationBehavioural Finance
Banking & Market InfrastructureFixed Income & RatesDerivatives & Structured ProductsPublic EquitiesTransactions & DealsPortfolio ConstructionFunds & AMCs
Private Markets & AlternativesRisk, Treasury & ControlAI & Digital FinanceStochastic Calculus & PricingWealth & Personal FinanceIndian Markets & RegulationProfessional Practice
CalculatorComparison
Frameworks
Explore Bootcamps
Equity ResearchPortfolio ManagementMutual Fund MasteryFinancial LiteracyInvestment Banking Analyst
Private Equity AnalystHedge Funds AnalystBreaking Into VCBreaking Into QuantsAI For Finance
Financial Analyst ProgramRisk Management ProgramPrivate Wealth ManagementDebt Capital MarketsDerivatives Foundation
Explore Internships
Equity Research InternMutual Fund Intern
Portfolio Management InternFinancial Literacy Intern
Explore Micro Courses

Equity Research6

Writing an Investment ThesisBuilding a Discounted Cash FlowReading an Annual Report FastReading a Sector Before a CompanySpotting Quality of Earnings Red FlagsBuilding a Revenue Forecast From Drivers

Portfolio Management3

Rebalancing: When, Why and What It CostsStrategic and Tactical Asset AllocationMeasuring Risk in a Portfolio

Mutual Fund Mastery3

Comparing Funds Without Being FooledHow a NAV Is Struck and Which Day You GetReading a Fund Factsheet Properly

Derivatives Unlocked4

Hedging a Real ExposureThe Greeks, PracticallyFutures, the Basis and What Moves ItReading an Option Payoff

AI For Finance2

Retrieval and Grounding for FinanceDocument Extraction in Finance

Breaking Into Quants4

Backtesting a StrategyHypothesis TestingCleaning Financial DataRegression for Finance

Breaking Into VC3

Sizing a MarketReading a Term Sheet as a FounderHow a Venture Round Actually Works

Financial Analyst Program4

Common Size and Trend AnalysisReading a Cash Flow StatementRatio Analysis That Says SomethingBuilding a Working Capital Schedule

Risk Management Program2

Credit Exposure and How It Is ReducedValue at Risk and What It Hides

Investment Banking Analyst3

Precedent Transactions and Why They DifferReading a Term Sheet StructurallyBuilding a Comparable Companies Table

Private Wealth Management3

Tax Aware Portfolio DecisionsBuilding a Client Risk ProfileGoal Based Planning Arithmetic

Debt Capital Markets3

Analysing an Issuer's CreditDuration and What It Does Not Tell YouBond Pricing and Yield Mechanics

Private Equity Analyst2

Fund Waterfalls and CarryThe LBO in Structure

Hedge Funds Analyst2

Short Selling MechanicsLong Short Mechanics
Courses
Explore Career Roadmaps
Investment Banking AnalystEquity Research AnalystVC AnalystPrivate Equity AnalystHedge Funds Analyst
Quant AnalystAI For FinanceFinancial Analyst ProgramPrivate Wealth ManagementDebt Capital Markets
Risk Management ProgramDerivatives FoundationPortfolio ManagementMutual Fund Mastery
PartnershipsShowdown
Log inSign up
Derivatives, Hedging & Structured Products
1Derivative Fundamentals
DerivativesLong PositionMark to MarketThe UnderlyingThe Derivative ContractHow Derivatives Transfer Financial…
2Forwards and Futures
The Futures ContractLong and Short PositionsThe Spot PriceThe Forward ContractSpot Price vs Forward PriceThe Futures PriceForward and Futures PositionForward vs FuturesHow to Read Futures Margin and Mark-to-MarketHow Futures Margin and Mark-to-Market WorkDeliveryRolloverOpen InterestOpen-Interest ChangeBasis vs Basis RiskHedge Ratio vs Hedge Effectiveness
3Options
OptionsThe Call OptionThe Strike PriceThe Put OptionOption DeltaOption Buyer and Option WriterCollar and Protective PutCall and Put OptionsHow to Map What…How to Take an…Exercise Price and Strike PriceOption Price DriversThe Expiration DateIntrinsic Value and Time Value
4Option Strategies and Payoffs
Option SpreadsOption PayoffVertical and Calendar SpreadsHow to Map an Option PayoffMaximum GainThe Iron CondorThe Covered CallMaximum LossStraddle and Strangle
5Volatility and the Greeks
The Implied Volatility SurfaceThe Option GreeksHow an Option Payoff…What an Implied Volatility…How Delta, Gamma, Theta…How Option Volatility Surfaces…Delta HedgingTime DecayHistorical VolatilityImplied Volatility vs Historical Volatility
6Swaps and Rate Derivatives
The Interest Rate SwapSwap Rate and Forward RateThe SwapThe Currency SwapInterest Rate Swap and Currency SwapThe Payment DateThe Reset DateThe Swap CurveThe Swap Payment CalculatorHow to Map a…Cross-Currency BasisDay Count ConventionsDerivative and UnderlyingExchange Traded and Over the CounterFixed Leg and Floating LegHow to Read a Derivative ContractHow to Map a Derivative ExposureHow to Read Derivatives Market DataHow to Map Derivative…How to Write a Derivative Research NoteHow to Run a…How to Maintain a Derivatives Decision Log
7Hedging Application
The HedgeHedge RatioHedge or SpeculationFraming a Hedge ObjectiveExposureOffsetBasis RiskHedge Risk or Counterparty RiskThe Hedged Item
8Structured Products
What a Structured Product IsStructured Product and Mutual FundHow to Take a…Participation RatePrincipal Protection and Capital Guarantee
9Clearing, Margin and Settlement
The Settlement PriceThe Three MarginsInitial, Variation and Clearing MarginPhysical and Cash SettlementHow a Position Moves…Market SurveillanceCounterparty RiskNettingNetting and SettlementPosition LimitsPosition Limits and MarginMarket ManipulationHow Corporate Actions Can…
10Derivatives Discipline and Cases
Derivative ResearchOpen Interest DataPost-Mortem and Performance Marketing,…Market Observation and Trade SignalScenario Analysis and ForecastReading Derivatives Data When…What a Derivatives Post-Mortem…

The Swap Payment Calculator: What One Period Settles At

The calculator below settles one period of a swap and nothing else. Given a notional, the rate on the fixed leg terms, the floating reading for that period and a day count fraction, it computes each leg in full, subtracts one from the other, and reports the size of the difference and which side it travels towards. One period. The worth of the whole arrangement goes unstated.

Work it out

One period, settled from the documents at hand

The calculator opens on the worked case set out below. Change any field and the day count fraction, both legs, the netting, the reconciliation and the drawing recompute on the keystroke. Nothing is stored and nothing is fetched.

Four settings worth reaching for
The build-upWhat it doesRupees
Notional, worked on by both legsstart hereRs 10,00,00,00,000/-
Day count fraction, 365 days over 365multiplies both legs1.000000
Fixed leg, notional times 7.2000 per cent times 1.000000Chitrakoot Cements owesRs 72,00,00,000/-
Floating leg, notional times 6.0000 per cent times 1.000000Saranga Capital owesRs 60,00,00,000/-
Net difference, one periodChitrakoot Cements pays Saranga CapitalRs 12,00,00,000/-
Route one, the two legs: Rs 72,00,00,000/- less Rs 60,00,00,000/- leaves Rs 12,00,00,000/-. Route two, the gap: Rs 10,00,00,00,000/- times 1.2000 percentage points times 1.000000 gives Rs 12,00,00,000/-. The two roads agree to the rupee. And adding back: the floating leg of Rs 60,00,00,000/- plus the difference of Rs 12,00,00,000/- returns the fixed leg of Rs 72,00,00,000/-.
Movement since the last change: nothing yet, this is the opening state.
Day count fraction
1.000000
Fixed leg, gross
Rs 72.00 crore
Floating leg, gross
Rs 60.00 crore
Size of the difference
Rs 12.00 crore
Share of the notional
1.2000 per cent
Which side hands it over
Chitrakoot Cements Limited pays
Which side takes it in
Saranga Capital Limited receives
THE TWO LEGS AND THE DIFFERENCE, ON ONE SCALE All three bars share one scale, set by the wider of the two legs. The difference runs right of the nil line when it travels towards Saranga Capital, and left of it when it travels towards Chitrakoot Cements. nil THE FIXED LEG gross, before netting Rs 72.00 crore gross THE FLOATING LEG gross, before netting Rs 60.00 crore gross THE DIFFERENCE the only thing that moves Rs 12.00 crore towards Saranga Capital The nil line is where the two legs come out equal and nothing moves on the payment date.
With a notional of Rs 1,000 crore, the fixed leg terms at 7.2000 per cent a year and the fixing for the period at 6.0000 per cent a year over a day count fraction of 1.000000, the fixed leg comes to Rs 72.00 crore gross and the floating leg to Rs 60.00 crore gross. Chitrakoot Cements Limited hands Saranga Capital Limited Rs 12.00 crore of net difference for this period.
Printed beside the answer, not beneath it
  • The instrument cannot say what the whole arrangement is worth today.
  • The settlement of the second period is beyond it, for want of a reading.
  • The rate that would have made the exchange an even one at the outset is beyond it too.
  • Whether an arrangement of this shape suits anybody is not arithmetic at all.
Educational illustration. The fixing is not drawn from any published series. One period only, and every figure is the one typed in rather than one read off a screen. The three day count bases are offered so that whichever the agreement names can be applied; the instrument does not know which one a given agreement names. No compounding within the period, no rounding convention beyond the whole rupee, no tax and no charge of any kind. Nothing is stored between visits.

Underneath it sits a product of three things and one subtraction. A leg is what a rate produces when it meets a figure and a stretch of time. The settlement is the gap between the two legs. Because both legs are worked on the same figure and the same fraction, the whole thing folds down to one figure meeting one distance.

One agreement sits behind every rupee here, and it carries two signatures. Chitrakoot Cements Limited committed itself to a rate somebody typed into the document and nobody has touched since, 7.20 per cent a year. Back the other way travels whatever the benchmark read for the period. Saranga Capital Limited signed that same bargain held up to a mirror. Rs 1,000 crore is the notional both obligations are worked on, and not one paisa of it travels anywhere.

One number is missing here. Nobody can put a worth on the whole arrangement today. A worth of that kind needs a list running reset by reset out to the end, saying where the benchmark is expected to sit at each one, and no such list exists. The arithmetic would run and produce a figure carrying no mark separating what was computed from what was guessed. So the tool halts at the end of the first period, announces that it has halted, and prints its reason where the second answer would sit.

What does this calculator work out, and what does it flatly refuse to?

The refusals come before the features. A tool that leaves its limits to be discovered by accident has already misled the reader once, at the moment of greatest trust. Greatest trust arrives the instant a number appears on screen.

The calculator works out four things. The fixed leg in full for one period, the floating leg in full for the same period, the difference between those two, and which of the two sides that difference travels towards. Every item on that list is a multiplication or a subtraction that could be done on the back of an envelope.

Four other things are declined deliberately. The first is what the whole agreement would fetch today, a valuationPutting one figure on what a whole agreement is worth on a given day.. The second is the settlement of the next period. The third is the rate that would have made the exchange even at the outset, a par rateThe rate that would leave both sides owing each other the same worth at the outset, so neither pays the other to start.. The fourth is whether signing something of this shape is sensible. Sensible is not arithmetic and never was.

The four are refusals, not features somebody has not got round to building yet. A gap in an interface invites someone to find the missing number elsewhere and paste it in. A refusal that explains itself states why the number would be wrong wherever it was found.

THE WHOLE TERRITORY, FENCED ON BOTH SIDES IT WORKS OUT The fixed leg in full, for one period The floating leg in full, same period The difference between the two legs Which of the two sides it moves towards IT WILL NOT WORK OUT What the arrangement is worth today What the second period settles at The rate that would make it even Whether anybody should be in one Both panels were drawn on purpose. The right one is not a list of features still to come.
The right panel stands on its own without the left: four things this calculator declines to work out, each declined for a stated reason rather than left as a gap somebody forgot to fill.

What are the four inputs, and where is each one found?

Four fields, four different places to go. Knowing what a notional means is no help when the screen wants one typed in and the analyst is staring at an agreement running to forty printed sides. The useful question is not what the word means but which document carries the number.

The notional is read off the front of the agreement, usually inside the first schedule. The notional is the figure both legs are computed on. Here it reads Rs 1,000 crore, or Rs 10,00,00,00,000/- written out.

The rate on the fixed leg terms is read off those terms in the same agreement. The rate arrives with a period attached to it in the document. A rate detached from its period has stopped being a rate, so the period is copied across along with the number. Here it reads 7.20 per cent a year.

The floating reading is read off the reset record for the period being computed, not off a screen showing what the benchmark says this morning. The reading was captured on that period’s reset date and carries that date with it. A record with no date is not the record required. Here it reads 6.00 per cent a year.

The day count fraction is the odd one out. There is nowhere to read it off. Apply the counting method named in the agreement to the actual dates of the period and what comes back is a decimal. Here it is 1.0000, one whole year. The instrument at the top opens that field out into the two things it is made of, the count of days and the basis they are divided by, with the reset frequency naming which count to expect.

The discipline running through all four is that every input comes off a document or a record, never out of somebody’s head. An input somebody remembered cannot be checked afterwards, and a settlement with no audit trailThe chain of documents that lets somebody months later show where a figure came from. behind it is believed only until somebody asks where it came from.

EVERY INPUT, AND THE PLACE IT IS READ OFF THE AGREEMENT AND THE RESET RECORD NOTIONAL, SCHEDULE 1 Rs 1,000 crore FIXED LEG TERMS 7.20 per cent a year RESET RECORD, PERIOD ONE 6.00 per cent a year DAY COUNT FRACTION not printed. It has to be worked out. Read off the front of the agreement. It is the one figure both legs are computed on. Read off the fixed leg terms of the same document, and it carries its period with it. Read off the reset record for this period, not off any screen today. It carries its own date. Nowhere to read this one off. Apply the counting method to the actual dates of the period. Three of the four inputs are copied straight off a document. The fourth has to be worked out, and it is the one the failure below is about.
Put a finger on each of the four fields in turn and name the document it comes from: three of them are transcription, and the fourth, the day count fraction, is printed nowhere and has to be derived from the period dates.
Try it out

One period of this swap is about to be settled. Of the four inputs, which one is printed as a number nowhere at all, neither in the agreement nor in the reset record?

How does the arithmetic actually run, in words rather than notation?

Slowly, and on paper if preferred. Not a step of it needs more than a hand and a pen. The tool is faster than the analyst and no cleverer.

Multiply the notional by the rate on the fixed leg terms, then multiply what comes back by the day count fraction. Rs 1,000 crore at 7.20 per cent a year is Rs 72.00 crore, and a fraction of 1.0000 changes nothing, so the fixed leg lands on Rs 72.00 crore gross, or Rs 72,00,00,000/- written out.

Now the same with the floating reading. The same Rs 1,000 crore at 6.00 per cent a year over the same fraction leaves the floating leg at Rs 60.00 crore gross, or Rs 60,00,00,000/-. A rate does not accrueTo build up gradually across a stretch of time, rather than arriving all at once on one date. anything on its own: it needs a figure and a stretch of time.

Take one leg from the other and Rs 12.00 crore of net difference is what remains, or Rs 12,00,00,000/- written out. The rate typed into the agreement on the Chitrakoot Cements side came out above the reading the benchmark gave, so Chitrakoot Cements is the side handing it over.

There is a shorter way. Both legs sit on the same notional and the same fraction, so those two lift out and the whole thing collapses to the notional meeting the distance between the two rates. The distance is 1.20 percentage points, or 120 basis points, and against Rs 1,000 crore it brings the same Rs 12.00 crore by a shorter road.

The tool prints the long form anyway, and it does that on purpose. Somebody who has only ever seen the gap route has never watched two legs get computed, so when one of them is wrong they have nothing on screen that could look wrong. The objection is not about accuracy. The shortcut is exact. The shortcut is just silent.

TWO ROADS TO ONE ANSWER THE LONG FORM Notional times 7.20 per cent times 1.0000 Fixed leg, Rs 72.00 crore gross Notional times 6.00 per cent times 1.0000 Floating leg, Rs 60.00 crore gross Rs 72.00 crore less Rs 60.00 crore Net difference, Rs 12.00 crore Four figures written, two of them legs Both legs stay on screen THE SHORTCUT 7.20 per cent less 6.00 per cent A gap of 1.20 percentage points Notional times 1.20 percentage points Net difference, Rs 12.00 crore Two figures written, and no legs The legs are implied Exact to the rupee, and silent about the legs A wrong leg leaves no trace Rs 12.00 crore of net difference One road shows the two legs. The other shows the gap and swallows the legs.
Follow each panel top to bottom and both arrive at Rs 12.00 crore of net difference, but only the left one puts Rs 72.00 crore and Rs 60.00 crore on screen where a wrong leg would be visible.
Risk Management Program Bootcamp — Fin Maverick

Why does one number come out of it rather than two?

Because the two obligations mature together: one date, one currency, the same pair of signatures. When two obligations meet like that there is no sense in each side sending the other a transfer that crosses in the middle. The two obligations are set against each other, and one transfer discharges both.

Picture two neighbours who share an autorickshaw to work. One hands over a flat Rs 60/- a day whatever happens; the other pays whatever the meter says. At the end of the week they do not exchange two bundles of notes. Whoever is behind covers the shortfall, once.

Both instruments here show all three figures at once, side by side. A calculator that printed only the answer would be teaching the shortcut and hiding the mechanism, and a reader who has never seen the two legs beside the settlement has no way of noticing when one of them has gone wrong.

The worry is not hypothetical. Rs 12.00 crore is what comes out of Rs 72.00 crore less Rs 60.00 crore, and equally what comes out of Rs 82.00 crore less Rs 70.00 crore, and one number on a screen could not tell those apart.

TWO LEGS IN, ONE TRANSFER OUT the floating level Rs 72.00 crore Rs 60.00 crore Rs 12.00 crore less gives what Chitrakoot Cements owes what Saranga Capital owes what actually moves on the day The lime slice on the first bar is exactly as tall as the third bar: it is the same figure. One date, one currency, one pair of signatures: a single transfer discharges both.
Lay a finger across the dashed line and the lime slice above it on the first bar is the whole of the third bar: Rs 72.00 crore gross less Rs 60.00 crore gross leaves Rs 12.00 crore of net difference to travel once.
Try it out

The long form and the shortcut land on the same Rs 12.00 crore. So why does the tool bother printing both?

The worked instance, every figure of it

Here is the whole computation in one place. Every figure below is also what the instrument at the top opens on. A worked example and the calculator beside it drifting apart is the fault that destroys confidence in both at the same moment.

What goes inWhere it is read offThe figure
NotionalFront of the agreement, first scheduleRs 10,00,00,00,000/-
Rate on the fixed leg termsFixed leg terms, same agreement7.20 per cent a year
Floating reading, first periodReset record, taken on that reset date6.00 per cent a year
Day count fractionCounting method applied to the period dates1.0000
Fixed leg, grossNotional times 7.20 per cent times 1.0000Rs 72,00,00,000/-
Floating leg, grossNotional times 6.00 per cent times 1.0000Rs 60,00,00,000/-
Net difference, first periodChitrakoot Cements pays Saranga CapitalRs 12,00,00,000/-

Two sentences worth carrying away. The Rs 12.00 crore measured against the Rs 1,000 crore it came from is 1.2 per cent, a shade over one hundredth. And that Rs 1,000 crore does not move at all: the settlement is a separate amount the size of one hundredth of it, not a slice taken out of it.

Try it out

Have a go before reading on. In the tool below, the rate on the fixed leg terms is about to be dragged from 7.20 per cent a year down to 5.00 per cent a year, with the floating reading pinned at 6.00 per cent a year. Where does the settlement go?

Play with it

One control, and the answer crossing through nil

Drag the rate on the fixed leg terms. The floating reading stays at 6.00 per cent a year. A reading for a period that has already reset is a fact rather than a setting. The notional stays at Rs 1,000 crore. The day count fraction stays at 1.0000. Watch which bar refuses to move.

low end, 5.00 per cent a yearset at 7.20 per cent a yearhigh end, 9.00 per cent a year
ONE CONTROL, THREE BARS THAT MOVE AND ONE THAT DOES NOT Upper strip drawn at Rs 1,000 crore across the full width. Lower strip drawn ten times larger. THE NOTIONAL never moves at any setting Rs 1,000 crore THE SETTLEMENT same scale as the bar above Rs 12.00 crore of net difference 1.20 per cent of the bar above THE TWO LEGS AND THE DIFFERENCE, TEN TIMES LARGER THE FIXED LEG gross Rs 72.00 crore gross THE FLOATING LEG gross, and it never moves Rs 60.00 crore gross THE SETTLEMENT signed, with nil marked nil Rs 12.00 crore towards Saranga Capital Rs 12.00 crore CHITRAKOOT CEMENTS LIMITED the fixed payer SARANGA CAPITAL LIMITED the floating payer
The notional, held still
Rs 1,000 crore
Fixed leg, gross
Rs 72.00 crore
Floating leg, gross
Rs 60.00 crore
Size of the difference
Rs 12.00 crore
Share of the notional
1.20 per cent
Which way it moves
towards Saranga Capital
With the fixed leg terms set at 7.20 per cent a year and the reading for the period standing at 6.00 per cent a year, one leg comes to Rs 72.00 crore gross and the other to Rs 60.00 crore gross. Chitrakoot Cements Limited hands Saranga Capital Limited Rs 12.00 crore of net difference for the first period.
Educational illustration. The benchmark reading is not drawn from any published series. One period only. The floating reading is held at 6.00 per cent a year because it belongs to a period that has already reset. The day count fraction of 1.0000 is supplied rather than worked out. The two ends of the control are declared for teaching and say nothing about where any rate has been or will go. Worth today, the par rate and any later period all lie outside what one period of arithmetic can reach.

The two ends of that control were picked so the crossing would sit well inside the range, and no published figure stands behind either. At 5.00 per cent a year the gap opens to 1.00 percentage point the other way, so Rs 10.00 crore travels towards Chitrakoot Cements. At 9.00 per cent a year it is 3.00 percentage points, putting Rs 30.00 crore on its way to Saranga Capital.

Derivatives Foundation Bootcamp — Fin Maverick

Why is the biggest input never the answer?

Type Rs 1,000 crore into the top field and Rs 12.00 crore comes out of the bottom. The gap between those two figures is where most of the confusion about swaps lives.

The notional enters as an input and leaves as nothing at all. Nobody has lent it, nobody has borrowed it, and on the payment date it does not move by so much as a rupee. A separate amount moves instead, the difference between two rates applied to the notional, and here that amount is the size of 1.2 per cent of it.

Most calculators would print those two figures as two lines of text and leave the comparison to the reader. The figure below and the drawing in the simulation put the answer on the same scale as the notional instead. Two numbers in a column invite the assumption that they are comparable. A picture does not. The answer bar is barely visible beside the notional bar. Barely visible is the correct impression rather than a drawing fault.

There is a household version that lands immediately. A neighbour mentions a home loan of Rs 40,00,000/-. Another mentions that the monthly instalment went up by Rs 1,200/-. Both are about the same loan and only one is about money leaving anybody’s hands this month. The notional is the first kind of statement; the settlement is the second.

THE INPUT AND THE ANSWER, DRAWN ON ONE SCALE THE NOTIONAL an input, never an output Rs 1,000 crore THE FIXED LEG gross, before netting Rs 72.00 crore gross THE FLOATING LEG gross, before netting Rs 60.00 crore gross THE NET DIFFERENCE the only thing that moves Rs 12.00 crore of net difference the size of 1.2 per cent of the bar at the top THE SAME NET, MAGNIFIED twenty times larger, so it can be seen at all The fourth bar renders about six pixels long here, and that is the correct impression rather than a drawing fault. None of the bar at the top travels at all, on any date.
The fourth of the four bars takes some finding: it is 1.2 per cent of the size of the top bar, and the top bar itself does not move by a rupee on the payment date.
Try it out

The notional bar in the tool above does not change height at any setting of the control. Which reading of that stillness is right?

Why does the tool stop dead after one period?

Because it has run out of facts, and it says so rather than inventing more.

Carrying the computation into a second period needs a floating reading belonging to that second period. Not a guess, not the first period’s reading pushed forward, not something interpolated out of neighbouring numbers. A reading, captured on the reset date that goes with that period, and dated where anybody can check it.

No such reading sits behind this guide. There is one for the first period, 6.00 per cent a year, and nothing after it. A tool that quietly filled in a default would produce a fabricated answer wearing a confident face. A blank field looks like something still to be done; a number looks like something that has been done.

The interface therefore draws the second period as a row that exists and cannot be completed, rather than declining to offer one. A control that silently stops says nothing about why. A row with its reason printed inside it states what has to be obtained, and what shape it has to arrive in.

THE ROW THIS TOOL CANNOT FILL IN PERIOD FLOATING READING WHAT IT SETTLES AT WHY Period one the first period 6.00 per cent a year Rs 12.00 crore of net difference the reading exists and carries its date Period two the one after it no reading cannot be filled no reading for this period exists here WHAT WOULD FILL IT: a reading of the benchmark belonging to that second period, captured on its own reset date and dated. No default stands in for one. The empty row is drawn rather than left out, because a missing row teaches nobody anything.
Look at the second row rather than the first: it is drawn, labelled and left unfillable on purpose, because the one thing that completes it is a reading of the benchmark belonging to that second period, captured on the reset date that goes with it.
Try it out

One thing has to arrive before the tool will settle the second period as well. Which?

Reading an Option Payoff — free micro-course from Fin Maverick

How is the output read, including the day it comes back nil?

An answer out of this tool has three parts, and a figure quoted without all three cannot be acted on or checked. An amount. A direction. And the period it belongs to.

Take the direction first. The direction is the part people assume they already know. The labels on the two sides do not settle it. Chitrakoot Cements is called the fixed payer, and Chitrakoot Cements receives whenever the floating reading has come in above the rate on its own fixed leg terms.

The name describes which leg it is on the hook for, not which way money travels in any particular period. With 5.00 typed into the fixed rate field at the top, the direction sentence turns round while both labels stay where they were.

Then the middle case, hidden by most calculators. When the two rates come out equal the answer is nil, nothing moves on the payment date, and both sides are fully discharged. A nil answer is a settlement, not a failure of the tool. Both instruments here word that state in a sentence rather than clearing the field. An empty field looks like something broke.

And the period. The same Rs 12.00 crore means opposite things depending on which side it travels towards, and it means nothing at all until the stretch of time it settles is stated.

THE SETTLEMENT AGAINST THE RATE ON THE FIXED LEG TERMS Vertical, Rs crore of net difference. Horizontal, the rate on the fixed leg terms in per cent a year. +30 +20 +10 nil minus 10 5.00 6.00 7.00 8.00 9.00 the difference goes towards Chitrakoot Cements the difference goes towards Saranga Capital Both legs come to the same figure at 6.00 per cent a year, so the settlement is nil. The default sits at 7.20 per cent a year, where Rs 12.00 crore travels towards Saranga.
Trace the line from left to right and it passes clean through nil at 6.00 per cent a year, the reading at which neither side is due a transfer from the other.
Try it out

The tool returns nil. Has something gone wrong?

In practice

Who fills in each of the four fields, and what they each check

Picture the tool in working life by field rather than by job title. Each of the four fields belongs to somebody different, and the errors cluster at the joins.

Whoever holds the agreement fills the first two fields, the only two that never change during the life of the arrangement. The check on them is not arithmetic. Open the document and confirm the digits. Confirming digits sounds too obvious to write down, until a spreadsheet turns up where somebody typed Rs 100 crore.

Whoever keeps the reset record fills the third, and cares above all that the reading carries its capture date. A reading without its date is not evidence of anything, and it becomes a serious problem when somebody months later tries to reconcileTo check a figure against a second figure reached by an independent route, so a mistake in either shows itself. the settlement against a counterparty’s own working.

Whoever knows the period dates fills the fourth. The fourth field has no natural owner, and the missing owner is why it goes wrong. The fraction is not printed in the agreement as a number. The fraction is derived each period from the counting method and the calendar, and derived work with no obvious owner quietly gets skipped.

And a lender or an analyst looking at all four from outside asks a fifth question the tool cannot answer: how the arrangement moves the party’s cash flow when the benchmark shifts, across a run of periods rather than this one. A settlement figure is evidence towards an answer rather than the answer itself.

The error that gets made, and what it costs

Somebody settles a quarter length period. The notional goes in correctly. The rate on the fixed leg terms goes in correctly. The right floating reading comes off the reset record for that period. And the day count fraction is left at 1.0000. The value was already sitting in the field from last time, and among the four it is the one that never shows up as a headline figure anywhere in the agreement.

The answer comes back four times too large. Rs 12.00 crore of net difference, where the true figure for a quarter length period is Rs 3.00 crore. The fixed leg should have been Rs 18.00 crore rather than Rs 72.00 crore, and the floating leg Rs 15.00 crore rather than Rs 60.00 crore.

Watch why nobody catches it. The figure is the right order of magnitudeRoughly how many powers of ten a figure carries. Two figures of the same order look equally believable at a glance., so it does not look absurd. The figure came from three inputs that are all correct, so anybody retracing the working finds three fields that check out. And the reviewer looks at the rates.

The cost is a payment instruction or a forecast line that is wrong by a multiple rather than by a rounding, sitting on a figure nobody is ever going to derive a second time.

The instrument at the top carries one defence against the error. The instrument knows what a period of the chosen reset frequency counts to under the chosen basis, and when the days field disagrees it says so above the answer and prints the multiple the entry is out by.

THE SAME THREE INPUTS, ONE FIELD APART AS THE FIELD WAS LEFT NOTIONAL Rs 1,000 crore RATE ON THE FIXED LEG TERMS 7.20 per cent a year READING FOR THE PERIOD 6.00 per cent a year DAY COUNT FRACTION 1.0000 WHAT COMES OUT Rs 12.00 crore of net difference AS THE DATES REQUIRED NOTIONAL Rs 1,000 crore RATE ON THE FIXED LEG TERMS 7.20 per cent a year READING FOR THE PERIOD 6.00 per cent a year DAY COUNT FRACTION 0.2500 WHAT COMES OUT Rs 3.00 crore of net difference FOUR TIMES TOO LARGE. Three of the four inputs were right and the fourth was left where it sat. The answer is out by a multiple rather than a rounding.
Without their bottom rows the two panels are identical: the only difference is 1.0000 against 0.2500, and it turns Rs 3.00 crore of net difference into Rs 12.00 crore.
The output is an amount, a direction and a period. See what nil says.

How is the error made to appear on screen?

The failure is worth producing as well as reading. A mistake watched appearing is harder to make than one only read about. Three moves in the instrument at the top, and none of them touches a rate.

Set the day count basis to thirty day months over 360 and the reset frequency to Quarterly. The days field fills itself with 90, the fraction reads 0.250000, and the answer settles at Rs 3.00 crore of net difference towards Saranga Capital. Rs 3.00 crore is correct for a quarter length period.

Now type 360 into the days field. Typing 360 is what leaving last period’s whole year value in place amounts to. The fraction jumps back to 1.000000 and the answer to Rs 12.00 crore. A panel appears above the build-up saying the days field disagrees with the reset frequency, and naming the multiple: 4.00 times the figure those dates support.

Nothing else on the screen changes. The notional is right, both rates are right, and the two roads still agree to the rupee. The two roads agree on whatever fraction they are handed. An answer four times too large arrives looking exactly as reasonable as the right one.

The panel can only compare two things it has been given. The panel knows what a period of the chosen frequency counts to under the chosen basis, and it knows what was typed. The period dates are beyond it, so the fraction can never be confirmed right. The one thing the panel can report is that the fraction and the frequency describe different periods.

Try it out

In the instrument at the top, with the basis set to thirty day months over 360 and the reset frequency set to Quarterly, 360 is typed into the days field. Which settlement has just been built?

What is this tool not, and who sets the conditions it stays silent on?

The calculator is not a valuation. Nobody is offering anything, so the figure is not a quote anybody would honour. And no such arrangement was ever struck between two real parties, so the figure is not a record of something that happened.

The figure is also not what goes into anything that has to be filed. A computed settlement and a figure required to be reported are two different objects, and the second is set by an authority rather than by arithmetic.

India

What is routed rather than written

One row, one authority, nothing written inside the box. Reporting on a privately agreed arrangement, in whatever shape and by whatever date, is territory the Reserve Bank of India occupies at rbi.org.in.

Reporting on a privately agreed arrangement, to whom and by whenSet by the Reserve Bank of India, rbi.org.in. Confirm it there.

The box stays empty because a stated requirement would be wrong the first morning it was revised. The same holds for whatever collateralAssets one side lodges with the other so an obligation has something standing behind it. either side lodges, and for whether an arrangement qualifies for hedging treatmentAccounting conditions that decide how an arrangement shows up in published statements.. Exchange traded contracts sit with the Securities and Exchange Board of India (SEBI) instead, at sebi.gov.in.

Try it out

A settlement figure comes out of this tool. Is that the figure that goes into a regulatory report?

What does a computed number not license?

Working out what a period settles at is arithmetic. Arithmetic is not a reason to do anything, and the ease with which a tool produces a figure has a way of feeling like an argument for using the thing the figure came from.

Whether an agreement of this shape suits any particular party is a question for advice. Arithmetic can say what a period settles at. Arithmetic cannot say that entering the arrangement was worth doing.

Consider what somebody would need before that question could be answered. The obligations that party already carries. The movement in its cash across a run of periods when a benchmark shifts. Over how long. And its own plan for a period where the answer came back on the other side, the period nobody plans for. A calculator holds none of that and cannot go and get it.

One more thing worth saying plainly. A period that once went against a party is not a failing on anyone's part. A settlement travels one way or the other according to where a benchmark happened to land, and nobody made a mistake by not knowing in advance where it would.

Try it out

Name the three parts of this tool's answer.

Five things are missing here, each for its own reason. Two are covered under their own subjects. One is handed back as a field to be filled in. One cannot be computed without facts that nobody holds yet. And one belongs to somebody with the authority to set it.

Three facts sit behind the floating reading of 6.00 per cent a year: where it came from, the date it was captured on, and the record it was written into. All three are set out under benchmark fixings. By the time the reading reaches this tool it is already fixed.

Turning a stretch of calendar into a decimal is a craft with conventions of its own, set out under day count conventions. Day count conventions are exactly why the count of days and the basis are fields to be filled in rather than something derived out of sight.

Which morning the Rs 12.00 crore of net difference actually leaves one account and lands in the other is a calendar question, covered separately. Size and direction are what this tool supplies. On the day, it says nothing.

Worth today, and the settlement of the second period, are both blank. Both need a stated expectation for the benchmark at every future reset, and no such expectation stands behind these figures.

Anything that must be told to an authority about an arrangement struck privately between two parties is the Reserve Bank of India's territory, rbi.org.in. The row for it is drawn above with its box left empty.

Breaking Into Quants Bootcamp — Fin Maverick

Sources, and what each one is here for

SourceSiteWhat it covers hereConfirmed
Reserve Bank of Indiarbi.org.inReporting on privately agreed arrangements: what must be told to an authority, in what shape and by when.28 August 2026
Securities and Exchange Board of Indiasebi.gov.inExchange traded contracts and what a dealer lodges against them.28 August 2026
Bank for International Settlementsbis.orgCross border statistics on privately agreed arrangements, each series published with the date it was compiled.28 August 2026
The two instruments in this guideno siteThe inputs and the settlement they produce, computed entirely from figures typed in rather than from any published series.28 August 2026

Chitrakoot Cements Limited and Saranga Capital Limited are invented.
Educational material. Not advice on any investment, tax, budget or market position.

← PreviousNext →
Fin Maverick Micro CoursesExplore Micro Courses
Fin Maverick BootcampsExplore Bootcamps
Fin Maverick

Finance education that ends in a job, not a certificate that gathers dust. Built for young India.

LEARN
CalculatorsFrameworksComparisonsCareersShowdown
RESOURCES
All CoursesMicro CoursesBootcampsInternships
COMPANY
AboutJob openingPartnership
LEGAL
Privacy PolicyTerms & ConditionsContent LicenseReturn & Refund Policy
© 2026 FIN MAVERICK / BUILT FOR INDIA.DO FINANCE, DO NOT JUST READ ABOUT IT.