How to Analyse Related-Party Transactions

Related-party transactions are dealings between a business and the people or entities connected to it, and they are disclosed because the connection means the terms were not settled by two unconnected parties bargaining. A related-party note is read in six steps: name the parties, separate transactions from balances, measure scale against the relevant total, ask whether the terms look ordinary, check what was approved and by whom, then write down what the disclosure cannot say.

Start on a street rather than in a note. A shop near a bus stand buys its cartons from a small packaging workshop, and the workshop is run by the shopkeeper's brother. There is nothing wrong with that arrangement, and in most of the country it is how supply gets arranged at all. But a lender to that shop would still want to be told. A price agreed between two people who eat together on Sunday was not arrived at the way a price between strangers is arrived at. The brother's price might be higher than a stranger's, or lower. The point is that the usual push and pull cannot be assumed to have produced it, so the question gets asked.

The impossibility of assuming an ordinary price is the entire reason the disclosure exists, and the reason is worth fixing in place before any procedure begins. A related partyA person or an entity connected to the business closely enough that the connection could influence the terms of a dealing between them. Group companies, senior managers and the people close to them are the usual categories. is a person or an entity connected to the business closely enough that the connection could influence what the two of them agree. The note is there so a reader can see the connection, not because the connection is a problem. The disclosure requirement is built for visibility, and a reader who treats visibility as an accusation has inverted the purpose of the thing they are reading.

How to read a set of statements is already established, and so are the figures Anjani Stationers Private Limited, an invented stationery business, published for its second year. Order is what remains. Almost every misreading of a related-party note comes from doing the steps in the wrong sequence: reading the amount before the relationship, or reaching a view before the scale has been measured against anything sensible. Six steps, in this order, and a stopping rule at the end that matters as much as the steps do.

Step one: whose names are on the list, and what does the list alone tell a reader?

Open the note and read only the list first. Not one rupee yet. The list is set out by relationship rather than by size, and the usual groupings are the companies inside the same group, including subsidiaries; the key management personnelThe people with authority and responsibility for planning, directing and controlling the activities of the business, directly or indirectly. Directors and the most senior managers are the usual members., meaning the people who direct the business; the other entities those people control; and the close relations of those people. Each grouping is a different kind of closeness, and reading them as a set takes about twenty seconds.

For Anjani Stationers Private Limited the list is short. Chitra Binding Works appears as a subsidiary, 70 per cent held, bought at the start of the second year. Vaidehi Rao appears as finance controller, in the key management grouping. Two names are the whole surround. The list shows what Anjani Stationers is connected to, so the list is information before any amount is attached to it, and a reader who skips it and jumps to the figures has thrown away the only part of the note that is free.

The list alone yields one sentence. Anjani Stationers has a binding operation inside its own group as of April of the second year. In the first year it had none. Nothing about that sentence needs an amount, and it already changes what the rest of the accounts are expected to look like. A list that runs to thirty entities across four states would say something entirely different and equally free. The shape is read before the size, every time.

Step two: how is what flowed separated from what is still owed?

Take the amounts now, and split them into two columns before doing anything else with them. One column holds the transactions. Transactions are the amounts that flowed across the whole year. The other holds the outstanding balanceThe amount still owed by or to a party on the last day of the year, after everything paid during the year has been taken out. It is a position on one date, not a total for the year. figures, which are what remained owed on the last day of the year. Anjani Stationers invoiced nothing to Chitra Binding Works and was invoiced Rs 8,00,000 for binding across the second year, and Rs 1,50,000 of that was still unpaid when the year closed.

A related-party note usually carries items that are neither, so keep a third column open. Anjani Stationers has one such item: a guarantee of Rs 8,00,000 given over Chitra Binding Works' borrowing. The guarantee is disclosed and not recognised, so it sits in no expense line and no liability line anywhere in the statements. Two separate items on this note are both Rs 8,00,000, the binding invoiced and the guarantee given, and they are entirely different things that must never be added, netted or spoken of as one figure. The coincidence is the sort of thing that produces a confident sentence in somebody's summary that is simply wrong.

Why insist on the split at all? Because the same pair of numbers reads differently in each direction. A large flow that settles down to almost nothing by the year end is one situation. A small flow that leaves nearly all of itself outstanding is another. Which of the two is in front of the reader cannot be told from a single combined figure, and the note reports both precisely so that it need not be guessed.

Step three: which total is the amount measured against?

Now measure the scale. Most of the damage is done at this step, and the reflex that does it is dividing everything by revenue. Revenue is the denominator people reach for when they have not asked which total the item belongs to. The rule is simple to state and takes discipline to follow: measure a related-party amount against the total it actually belongs to, so a purchase goes against what the business spends on bought-in materials and services, a balance goes against the payables it sits inside, a loan goes against borrowings, and a guarantee goes against the borrowing it stands behind.

Work it on Anjani Stationers. The binding is a purchase, so it belongs against the cost of materials consumed. Cost of materials consumed was Rs 1,48,50,000 in the second year. Rs 8,00,000 against Rs 1,48,50,000 is 5.4 per cent. The unpaid amount is a balance, so it belongs against trade payables of Rs 22,00,000. Rs 1,50,000 against Rs 22,00,000 is 6.8 per cent. Two numbers, two sentences, and the step is done.

Watch what the habit would have produced instead. The same Rs 8,00,000 measured against revenue of Rs 2,70,00,000 is 3.0 per cent. Against total expenses of Rs 2,32,00,000 it is 3.4 per cent. Against profit before tax of Rs 38,00,000 it is 21.1 per cent. Every one of those divisions is arithmetically correct, and three of the four answer a question nobody asked. The spread from 3.0 to 21.1 is not a rounding difference, it is a seven-fold difference, and a reader who picks a denominator carelessly can make the same ordinary purchase look trivial or alarming without changing a single fact.

Four settings are worth walking through in words. At the default, Rs 8,00,000 of binding against cost of materials consumed of Rs 1,48,50,000 is 5.4 per cent and that is the relevant total. Switch the denominator to revenue and the same amount reads 3.0 per cent. The 3.0 per cent is arithmetically correct and answers a question about selling rather than about buying. Switch to profit before tax and the reading is 21.1 per cent. A reading of 21.1 per cent is the one most likely to end up in somebody's summary and the one least connected to a purchase. Switch the item to the Rs 1,50,000 still unpaid and trade payables becomes the relevant total, giving 6.8 per cent. Switch the item to the Rs 8,00,000 guarantee and the panel names none of the five. A guarantee belongs against the borrowing it stands behind, and that borrowing is Chitra Binding Works' own and appears nowhere on this face.

Step four: what can a reader actually check about the terms?

The note will usually state that the transactions were made on arm's lengthTerms equivalent to those that would apply between two parties with no connection to each other, each acting in its own interest. It describes how a price was arrived at, not whether the price is high or low. terms, meaning terms equivalent to those two unconnected parties would have agreed. Read that sentence for exactly what it is. The arm's length sentence is an assertion made by the business about its own dealings. An assertion is not evidence, and treating it as evidence is one mistake while treating it as a confession is the opposite mistake and no better.

A reader can test consistency, and three tests are available from published information. The three checks are whether the same service is bought elsewhere at a comparable rate, whether the balance settles on the same terms third parties get, and whether the volume moved sharply without an operational reason. Each of the three has an answer that can be found or explicitly cannot, and finding out which of the two applies is itself the output of the step.

Run them on Anjani Stationers. On rate, the note gives a total of Rs 8,00,000 and no rate per unit, so the honest answer is that the check cannot be completed and the question goes on the list. On settlement, Rs 1,50,000 of Rs 8,00,000 outstanding at the year end is not obviously out of line with ordinary supplier terms, though a reader would want the days to compare properly. On volume, the change is large and the reason is plain on its face: binding was done inside Anjani Stationers before, and from April of the second year it is bought from an operation the business had just acquired. An operational change with an obvious explanation is not a red flag, and the ordinary explanation is the one that goes first.

Step five: what does the approval trail record?

Now leave the amounts and read the governance sentences. Most readers skip them. Three questions: which arrangements needed approval before they were entered into, whether the audit committeeA committee of the board charged with overseeing the financial reporting process, the internal controls and the audit. Where a company has one, related-party arrangements are usually among the things it reviews. reviewed them, and whether any of them needed a resolution put to the shareholders. Write down the answers as answers, including the answer none.

The approval trail records what the business itself treated as significant, so the trail is often more informative than the amounts. Somebody inside made that judgement knowing far more about the arrangement than any outside reader ever will. An arrangement taken to shareholders was one that somebody inside decided was big enough or unusual enough to be worth the trouble. An arrangement approved routinely at board level was treated as ordinary. The difference between those two levels is free information about how the business ranked its own dealings, and collecting it costs a reader thirty seconds.

For Anjani Stationers the trail is short, and it should be recorded as short rather than as absent. A binding arrangement of Rs 8,00,000 with a subsidiary is the kind of ordinary trading a board deals with as a matter of course. Giving a guarantee commits the business to something without paying anything now, so a guarantee over a subsidiary's borrowing is the item on this note most likely to have needed a specific decision. A reader records which of the two arrangements received which treatment, and where the note does not say, the honest entry is that the note does not say. The approval trailThe record of who authorised an arrangement and at what level, from a routine board approval up to a resolution put to shareholders. It shows the level of attention a business itself applied. is a record to be copied out, not a puzzle to be solved.

Step six: what will the note never tell, however carefully it is read?

The last step is to write the limits down, in the same output as everything else, in plain words. There are three of them and they are permanent. The note cannot show whether the price was right. The note cannot show whether an alternative supplier existed and what that supplier would have charged. And the note cannot show whether the relationship influenced the decision. Influence lives in somebody's mind, and no disclosure has ever contained that.

A limit that is left implied gets quietly filled in by whoever reads the work next, so naming the three limits explicitly is a step rather than a caveat. This is the practical difference between a note that says the rate per unit is not disclosed and a note that stays silent about rates. The first hands the next reader a boundary. The second lets them assume the rate was checked and found fine. Same underlying information, completely different downstream reading.

There is a version of this step that goes wrong, and it is worth naming. Writing that the note cannot show whether the price was right, and then adding that this is concerning, has turned a limitation into an insinuation. The limit is a feature of what published accounts are, and it applies identically to a business whose related-party dealings are entirely ordinary and to one whose dealings are not. A limit that applies to everybody distinguishes nobody.

When does the reading stop?

The reading stops when the six steps are done, and the discipline of stopping is as much a part of the method as any of the steps. The sequence is finished when the parties are listed, transactions and balances are separated, scale is measured against the totals the items actually belong to, the terms question has been asked and answered as far as published information allows, the approval trail is recorded, and the limits are written down.

The output of the whole sequence is a short list of questions and a set of measured scales, and it is never a conclusion about the business or the people named in the note. On Anjani Stationers Private Limited the output is genuinely short: binding is 5.4 per cent of what the business spends on materials, the outstanding balance is 6.8 per cent of trade payables, a separate guarantee of Rs 8,00,000 is disclosed and not recognised, the rate per unit is not given and would be worth asking about, and the change in the arrangement has an operational explanation visible on the face of the accounts. Nothing in that paragraph is unusual, and saying so plainly is part of doing the work properly.

Stopping matters because the alternative is not a better analysis, it is a worse one. Once a reader passes the end of the evidence, everything they add comes from somewhere other than the accounts: a pattern they read about, a case they remember, a general feeling about businesses that trade inside their own group. Borrowed material of that kind feels like insight and behaves like noise, and it is the single most common way a competent reader of accounts turns into an unreliable one.

What must never be a step, at any point?

Three things, and each of them is common enough to be worth naming individually. The first is treating the existence of related-party transactions as a problem in itself. Existence is not a problem. Businesses trade with the businesses they are connected to constantly, and in a group structure such trading is very often why the group was built at all.

The second is inferring anything from the relationship on its own. A relationship is a fact about who is connected to whom. A relationship carries no information about price, terms or intent, and a reader who moves from the relationship straight to a view has skipped every step that would have produced evidence.

The third is treating the note's own arm's length statement as proof that the terms were not at arm's length, on the reasoning that nobody says it unless there is something to hide. The reasoning would condemn every business that complies with a disclosure requirement, and every business complies. A parent buying binding from its own subsidiary is among the most unremarkable arrangements in business, most related-party dealing is entirely benign, and the note exists so a reader can see it rather than because seeing it should worry anybody.

What does the whole sequence look like, run end to end on one set of accounts?

Here is the sequence run once, on the second year of Anjani Stationers Private Limited, with every figure taken from what the business published. The table below is a worked output rather than a template to copy. The value sits in what each row does and does not say.

Notice the missing rows. There is no row for a view, no row for a concern and no row for a score. The sequence run properly on Anjani Stationers Private Limited produces one open question and two percentages, and a reader who wanted more than that from this note wanted something the note was never built to supply.

Who reads a related-party note, and what do they actually do with it?

Three people open the same note in the same week and none of them is reading it for the same reason. The difference in purpose changes what each of them writes down.

A lender reads it for a claim it might not have. When Anjani Stationers gives a guarantee of Rs 8,00,000 over Chitra Binding Works' borrowing, a lender to Anjani Stationers wants to know that a commitment exists which pays nothing today and could require payment later. The guarantee changes nothing about whether the business is sound, and it changes what the lender models. Being modelled differently is not the same as being unsound. The lender's output is a question about the terms of the guarantee, not an opinion about why it was given.

An analyst reads it for the shape of the cost base. Binding at 5.4 per cent of cost of materials consumed tells an analyst that a real slice of the cost side now runs through an operation inside the group, so if a forecast assumes that cost behaves like an outside supplier's price, the assumption has just been named and can be tested. The useful output for an analyst is not a judgement about the arrangement but a sharper question about the forecast, and that is a genuinely different thing to write in a model.

And Vaidehi Rao, as finance controller of Anjani Stationers, reads it from the inside, for completeness rather than for interpretation. Her question is whether every related party is listed, whether the transactions and balances tie to the ledgers, and whether the guarantee is described accurately as disclosed and not recognised. A missing name in that note is the kind of error that is embarrassing precisely because the note is where a business demonstrates that it has nothing to hide.

Anjani Stationers Private Limited, Chitra Binding Works and Vaidehi Rao are invented.
Educational material. Not advice on any investment, tax, budget or market position.