Off-Balance-Sheet Financing: Where Obligations Hide

Off-balance-sheet financing describes real obligations that never enter the liability total, either because the recognition rules keep them out or because they are promises about the future rather than present debts. Most of it is routine and fully disclosed a few headings further on in the notes: commitments, guarantees given, and claims that are possible rather than probable. Reading it means rebuilding, not suspecting.

Here is what sits underneath that. A liability total is not a list of everything a business will ever have to pay. A liability total is a list of what qualified, and qualifying takes three things at once: an obligation that already exists, an outflow that is probable, and an amount that can be measured reliably. Plenty of perfectly real promises fail one of those three tests. The rules do not throw them away when that happens. The rules send them to the notes, where the amount is stated in full alongside the reason it is not in the total.

The word hide needs taking apart before anything else. The obligations are hidden from the total, and they are not hidden from the reader. In an ordinary set of accounts every one of them is printed, in figures, in the same audited document, a few headings behind the statement itself. The reason obligations get missed is almost never that somebody buried them, and almost always that the reader stopped at the total.

Reading it well comes down to four skills: naming the test an item failed, telling a commitment from a guarantee from a contingent liability, finding all three in a report without being told where they are, and rebuilding the obligations of Anjani Stationers, an invented stationery business, from the reported Rs 38,00,000 to Rs 56,80,000 while saying exactly what that rebuilt figure assumes.

What does off-balance-sheet actually mean?

Off-balance-sheet means outside one total, and nothing more than that. The balance sheet reports what a business holds and what it owes at a single date, and the liability side of it carries a total. An item is off-balance-sheet when it does not enter that total. An off-balance-sheet item can still be described in the same report, in figures, with a heading of its own. Whenever somebody says an obligation is off the books, two different things are being confused. One is recognitionPutting an amount into the numbered statements themselves, so that it enters the totals and changes what the statement adds up to., putting an amount into the statement so it enters the totals. The other is disclosureStating a fact in the report without putting it into the numbers on the statement, most often as a note behind the statements., stating the fact somewhere in the report without putting the amount into the totals.

Off-balance-sheet means not recognised, not undisclosed, and the whole subject rests on that single distinction. A household makes the same distinction without thinking about it. Asked what it owes, it would name the home loan and the balance on the card. The household would probably not name the two years left on the rental agreement, or the signature given as guarantor on a cousin's vehicle loan last winter. Neither of those is a secret, and either would be described to anyone who asked. Neither shows up when today's debts are added up, and a business faces exactly the same distinction with a rulebook attached to it.

Why do the rules let a real obligation sit outside the total?

Because the total would stop meaning anything if it did not. Recognition is a gate with three tests, and an item has to pass all three. Is the obligation present, meaning it already exists rather than arising from something the business has not done yet? Is an outflow probable, meaning more likely than not? Can the amount be measured reliably? Open that gate wider and the liability total starts to fill with things that might happen, and a reader can no longer tell what the business actually owes from what somebody thought it might one day owe. Keep the gate where it is and the total stays a hard number, with everything that fails a test standing just outside it in writing.

An obligation sits outside the total because it fails a named test, and naming the test it failed is how a reader decides whether to add it back. The three items Anjani Stationers discloses have something in common. Every one of them is measurable. Rs 10,80,000, Rs 8,00,000 and Rs 2,40,000 are exact figures, not estimates, so measurement is never the reason any of them is out. The business has not yet used the warehouse years it has agreed to pay for, so the commitment fails the present test. Chitra Binding is expected to pay its own borrowing and the claim is being contested, so the guarantee and the disputed claim both fail the probable test. Different tests, different reasons, and the reason is what tells a reader what the amount means.

What kinds are there, and which of them are routine?

Three kinds cover almost everything a reader meets, and the honest answer to the second half of the question is that all three are routine. A commitmentAn amount a business has agreed to pay for something it has not yet received or used, such as goods ordered for later delivery or space taken for future years. is an amount agreed for something not yet received or used. A guaranteeA promise to pay somebody else's debt if that other party does not pay it. The promise is real from the day it is given, even if it is never called on. is a promise to pay somebody else's debt if that other party does not. A contingent liabilityA possible obligation whose existence or amount depends on something that has not happened yet, stated in the notes rather than added into the liability total. is a possible obligation whose existence or amount depends on something that has not yet happened, most often a dispute. Ordinary businesses carry all three without anything being wrong.

None of these three kinds is evidence of anything by itself, and treating the presence of a guarantee or a contingency as a warning sign is a misreading of what the notes are for. A vegetable seller who takes a stall on a three-year agreement has a commitment. A father who signs behind his daughter's education loan has given a guarantee. A tailor arguing with a supplier over a delivery that arrived torn has a contingent liability. None of the three is doing anything clever. Each of them is a normal fact of trading that the accounting rules have decided belongs in words and figures beside the statement rather than inside its totals. The size of each one, and the event that would turn it into a payment, are the things that matter, and both are stated in the notes.

Where in the report are these obligations found?

In the notesThe numbered explanations behind the statements that break down and add to the figures on them. The notes are part of the audited report, not an appendix to it., under two headings that are worth memorising because they are almost always called the same thing: commitments, and contingent liabilities. Now here is the part that catches people. Every recognised line on the balance sheet carries a note number beside it, so a reader can start from a figure and walk to the explanation behind it. The whole point of the commitments note and the contingent liabilities note is that there is no figure, so neither has a figure on the statement to hang off. Nothing on the balance sheet points at them.

No line leads to these two notes, so the only way to find them is to open the notes and read the headings. The missing line is why they get missed, and it has nothing to do with concealment. Reading a report by starting at a total and clicking through the references is efficient and it works everywhere else. Clicking through simply cannot work here. The habit that catches these amounts is dull and reliable: after finishing with the statement, go to the notes and read the headings from first to last, and stop at the two that describe things that are not in any total.

How is a fuller picture rebuilt, and what does that change?

By taking the reported total and adding back what the notes disclose, one item at a time, saying out loud what each addition assumes. Adding back is the whole method. Nothing in the accounts is wrong, so the rebuilt figure is not a correction. The rebuilt figure is a reader's own working, built for a purpose the statement was never designed to serve: getting a sense of the full weight of what a business has taken on, rather than only the part that qualified for the total.

Adding the warehouse commitment of Rs 10,80,000 and the guarantee of Rs 8,00,000 to the reported Rs 38,00,000 gives rebuilt obligations of Rs 56,80,000, 49.5 per cent more than the total the reader started from. That changes a comparison a reader might have been about to make. Against equity of Rs 1,42,00,000, the reported liabilities are 0.27 times. The rebuilt figure is exactly 0.40 times. Neither number is the truth and neither is a lie: the first answers what does this business owe today, and the second answers what has this business taken on. Adding the Rs 2,40,000 claim would assume that a claim the business is contesting turns into a payment, and the accounts say that is not probable, so the claim is left out of the rebuild. A stress test may add it anyway, provided that assumption is stated.

The readings the calculator produces are these. The warehouse commitment alone gives Rs 48,80,000, or 0.34 times equity. The guarantee alone gives Rs 46,00,000, or 0.32 times. Both together, the rebuild to start from, give Rs 56,80,000, or exactly 0.40 times. All three give Rs 59,20,000, or 0.42 times, and that last figure carries an assumption the accounts explicitly contradict. Every switch changes the number and none of them changes the accounts. A reader's working figure differs from a restatement in exactly that way.

Why can two businesses with the same position report different liabilities?

Because the recognition gate responds to the shape of an arrangement, not only to its weight, and two businesses can take on comparable weight in two different shapes. Take two invented notebook makers of the same size. Kanchan Copybooks has bought its warehouse and borrowed Rs 10,80,000 to do it. Palash Paper Works has taken an identical warehouse on a three year arrangement short enough to fall within an exemption from capitalisationPutting an arrangement on the balance sheet as an asset with a matching liability, rather than leaving it to be described in the notes and charged as it is used., at the same Rs 3,60,000 a year. Everything else about the two businesses is identical, and each reports Rs 27,00,000 of other liabilities.

Kanchan Copybooks reports Rs 37,80,000 of liabilities and Palash Paper Works reports Rs 27,00,000, a difference of Rs 10,80,000 that is entirely a matter of shape, and both figures are correct. Read the notes and the two converge at once: Palash Paper Works discloses Rs 10,80,000 of commitment, and the rebuilt figures for both are Rs 37,80,000. There is a fairness point buried in this and it is worth saying plainly. Kanchan Copybooks looks more indebted than its neighbour for having chosen the shape that puts the amount on the statement. Kanchan Copybooks also carries the warehouse itself as an asset, and Palash Paper Works does not. The presentation differs. Neither business did anything clever, and neither did anything wrong.

What sits outside Anjani Stationers' Rs 38,00,000?

Three items, no more, set out in full below with the reason each one sits where it does. The last column is doing the real work: it is the reason, not the amount, that settles what to do with each figure.

The three amounts outside the total come to Rs 21,20,000, more than half the size of the reported liability total itself, and not one rupee of it is anywhere other than the notes. That proportion is the reason this subject matters at all. A reader who stops at Rs 38,00,000 has read a figure that is correct and has missed an amount worth 55.8 per cent of it, sitting two headings away in the same document. Nobody moved it. Nobody obscured it. The reader simply did not read on.

How does a lender actually use these two notes?

Outside the classroom, commitments and contingencies are not an idea people admire. Commitments and contingencies are two headings a credit officer turns to on purpose, in a fixed order, before the file goes any further. Anjani Kulkarni asking for a larger facility would find her report read like this, and none of it starts from suspicion. The reading starts from the ordinary professional habit of not letting a total be the last thing read.

A lender reads the commitments note for what the business has already agreed to, the contingent liabilities note for what could be called on, and then rebuilds a working figure that sits beside the audited one rather than replacing it. Notice what happens to the guarantee in that reading. The Rs 8,00,000 given for Chitra Binding is not a debt, and a lender does not treat it as one. A lender treats the guarantee as a size: the largest amount Anjani Stationers could be asked for if the business it holds a stake in does not pay. A lender is entitled to ask exactly that before adding to its own exposure. And notice what a good lender does not do. A good lender does not mark the file down for having disclosed anything. The disclosure is the good behaviour.

What does the rebuilt figure still not say?

Three things, and each of them is something a reader assumes a bigger number has settled. A commitment spread over three years and a guarantee that may never be called are being added together as though they were the same kind of thing, so the rebuilt figure does not say when any of these amounts falls due. The rebuilt figure does not say whether the business could meet them. Whether it could is a different question, answered from different statements. And the rebuilt figure does not claim to be more correct than the reported one. Neither figure is more correct, and each answers a different question.

A rebuilt figure is only as good as the assumptions stated beside it. Quoted without them, it looks like a measurement and behaves like an opinion, and that is worse than the reported total it replaced. A fourth silence is worth naming. The notes state what has been disclosed. The notes cannot state what nobody thought to disclose, and no amount of rebuilding produces an item that is not there. The silence is not a reason to distrust the notes. The silence is the reason the audit exists, and the reason a reader treats a rebuilt figure as a question to ask rather than an answer to file.

Anjani Stationers Private Limited, Chitra Binding Works Private Limited, Anjani Kulkarni, Meera Rao, the Sunrise Public School group, Kanchan Copybooks and Palash Paper Works are invented.
Educational material. Not advice on any investment, tax, budget or market position.