Hybrid Instruments: Where Classification Becomes a Judgement Call

A hybrid instrument carries features of more than one kind at once: debt that converts, equity that must be redeemed, a loan whose return moves with something outside it. The classification test still applies, but it gives more than one answer, so the response is to take the instrument apart and classify each part on its own. Where the parts cannot be separated cleanly, the whole is measured at fair value.

Everything in this sequence so far has had a clean answer. A term loan is a liability. An ordinary share is equity. The business cannot avoid handing over the cash on a preference share redeemable on a fixed date, so that share is a liability whatever the certificate says. One question, asked once, and the instrument lands on one side of the balance sheet. The hybrid instrument is where that stops working, and the clean cases must be firmly in place before the hard ones are worth anything.

Anjani Stationers Private Limited, an invented stationer, has issued no hybrid instrument of any kind. Its published equity of Rs 1,42,00,000 contains ordinary share capital of Rs 40,00,000 on 4,00,000 shares and retained earnings of Rs 1,02,00,000, and nothing else. Its published liabilities of Rs 38,00,000 contain a term loan, a lease liability, deferred tax, trade payables and a contract liability, and nothing else. Its published finance cost of Rs 3,50,000 is built from cash credit interest of Rs 2,64,000, term loan interest of Rs 41,000 and lease interest of Rs 45,000, and nothing else. Every hybrid instrument below is a teaching case built beside those accounts and never inside them.

What makes an instrument hybrid, and why does one test give two answers?

Consider a receipt that could actually be held. A woman who runs a wedding hall takes Rs 2,00,000 from a customer nine months before the date. The receipt says two things at once. If the wedding is called off before a stated date, she returns the Rs 2,00,000 in cash and she has no choice about it. If the customer would rather hold the booking over to the following season, the customer may do that instead, and the hall keeper cannot refuse. She has one receipt in her drawer and two quite different promises written on it: one she cannot escape, and one she does not control the trigger for. The receipt has two limbs and they point in different directions, so the simple question, does this receipt oblige me to hand over cash, has to be answered twice.

A hybrid instrumentA financial instrument carrying features of more than one kind at the same time. The ordinary classification question returns more than one answer, and the instrument has to be taken apart before it can be reported. is that receipt written as a security. The classification question that separates a liability from equity has not changed and has not weakened: can the holder oblige the business to hand over cash or another financial asset, in a way the business cannot avoid? The change is that the instrument has more than one limb, so the question is asked of each limb and comes back with more than one answer. A hybrid instrument does not defeat the classification test. It defeats the assumption that one instrument produces one answer, and that is a different and much more manageable problem.

Four shapes cover most of what a reader meets, and it is worth seeing all four before any mechanism arrives. A convertible bond carries a repayment obligation and a right to take shares instead. A redeemable preference share carries the vocabulary of equity and the substance of a debt. A perpetual instrumentAn instrument with no maturity date, so the principal is never contractually repayable. The instrument is equity only if its periodic return can also be withheld, whatever the instrument is called. with a discretionary coupon carries the vocabulary of debt and, quite often, the substance of equity. And a loan whose return steps up or down with something outside the lending relationship carries an ordinary borrowing plus a feature that behaves like a separate bet.

The middle two rows carry the lesson that survives everything else. The redeemable preference share and the perpetual bond both look split and are not. The first is called a share and the answer is liability. The second is called a bond and the answer is equity. However hybrid they sound, neither is a hybrid in the strict sense. The one question asked of each limb returns the same answer both times. The genuine hybrids are the first and the last rows, where the limbs really do point in opposite directions and no single answer is honest.

Splitting a compound instrument or separating an embedded derivative: which problem is being solved?

Two mechanisms exist for taking an instrument apart, and the whole argument rests on keeping them apart. The two sound alike, they are described in the same paragraph of most summaries, and they answer completely different problems. Confusing them produces answers that are not merely imprecise but wrong in both directions, so each one is built separately before either is compared with the other.

The first mechanism handles a compound instrumentA single instrument the issuer has put out that contains both a liability component and an equity component of its own. The instrument is separated into those two parts when it is first recognised, and the parts are then reported on opposite sides of the balance sheet.. The problem here is presentational and it is about the issuer's own paper: one instrument the business has issued carries an obligation it cannot avoid and, at the same time, a right that would be settled by handing over a fixed number of its own shares. Part of it is a liability and part of it is equity. The response is to split the amount raised into those two components when the instrument is first recognised, put one part with the liabilities and the other part inside equity, and leave the equity part alone forever afterwards. The convertible bond example works exactly that mechanism: a hypothetical Rs 20,00,000 raise divided into a liability component of Rs 16,20,921 and an equity component of Rs 3,79,079, the two summing back to the Rs 20,00,000 received because the equity part is computed as the remainder.

The second mechanism handles an embedded derivativeA feature written inside a larger contract whose value moves with a rate, a price, an index or another outside variable. It behaves like a standalone derivative even though nobody issued it separately.. The problem here is a measurement problem and it has nothing to do with which side of the balance sheet anything sits on. Some contracts carry a feature whose value moves with an outside variable: a price, a rate, an index, the outcome of a test. The feature behaves like a standalone derivative even though nobody wrote it as one. Left buried inside its host contractThe main contract inside which an embedded feature sits. Where a feature is separated, the host is what remains and continues to be measured on its own ordinary basis. its movements would never be reported at all. So in defined circumstances it is pulled out and measured on its own, at fair value, with the movements going to profit. The host stays exactly where it was and continues on its own ordinary basis.

Splitting a compound instrument decides which side an issuer's own paper belongs on. Separating an embedded derivative decides what has to be remeasured. Conflating the two is the commonest error on this subject. Notice how little they share. One splits an amount raised; the other extracts a feature. One puts a part into equity; the other puts nothing into equity at all. One is decided once at issue and the equity part never moves again; the other creates a component that is remeasured in every reporting period. One instrument can need both, and one instrument can need neither.

The second panel carries no number for the separated feature, and the omission is deliberate. The size of a separated feature depends on a valuation of that feature. The dashed block shows that the feature exists and how it behaves, not what it is worth. A note in a real set of accounts would carry the amount and the basis on which it was arrived at, and that note is where a reader goes for it.

When is a feature left inside the contract rather than pulled out?

Most contracts carry features that move with something. A loan whose interest is set against a benchmark rate moves with that rate. A supply agreement whose price is revised each year moves with the revision. If every such feature were pulled out and measured on its own, the accounts would become a catalogue of tiny derivatives and nothing would be gained. So separation is not automatic and never was. Two conditions have to hold before a feature comes out, and both of them are easier to remember as questions than as rules.

The first asks whether the feature is closely relatedA description of an embedded feature whose economics run with those of the contract it sits inside. Where they run together the feature stays inside; where they belong to a different market or variable, it is a candidate to be separated. to the host it sits inside. Take the household version. A shopkeeper takes a loan whose interest is set against a benchmark rate, and the loan is a lending arrangement, so a feature about the cost of borrowing sits naturally inside a borrowing. The economics run together. Now take the same shopkeeper with a loan whose interest steps up if the price of paper crosses a level. A paper price has nothing to do with lending. The step up is a bet on a commodity market, tied to a loan agreement by an act of drafting rather than by economics. The first stays inside. The second is a candidate to come out.

The second question asks whether the whole instrument is already measured at fair value with the movements going to profit. If it is, the feature's movements already reach the profit figure through the whole, and separating it would achieve nothing but extra work. An embedded feature is separated only where it does not belong to the economics of its host and where the whole is not already being remeasured. Most features in most contracts fail one of those two conditions and are never separated at all.

What happens when the parts cannot be separated at all?

Sometimes the features are so entangled that measuring one of them on its own is not possible. The response then is not to guess. The whole instrument goes to fair value through profitA basis on which an item is carried at what it would fetch at each reporting date, with every change in that amount going straight into the profit figure for the period rather than being held back.. The instrument is carried at what it would fetch, remeasured at every reporting date, and every movement in that amount lands in the profit figure for the period.

The fallback sounds tidy and it is not. Fair value through the whole instrument is a fallback taken when separation fails, never a preference, and its cost is that reported profit starts moving for reasons that have nothing to do with the business's trade. Put it on Anjani Stationers' own scale to feel the size of it. The business reported profit before tax of Rs 38,00,000 in year two, earned by selling notebooks and exercise books to schools and dealers. Suppose, purely as an illustration, that it carried one instrument at fair value and that instrument moved by Rs 2,00,000 in a period. Profit before tax would read Rs 40,00,000 in one direction and Rs 36,00,000 in the other, a swing of more than five per cent either way, with the same notebooks sold to the same schools at the same prices. Nothing about the trade produced that movement, and no amount of reading the revenue line would explain it.

The fallback matters to a reader as much as to a preparer. A profit figure containing fair value movements is answering a slightly different question from a profit figure that does not, and the two cannot be compared year on year or business to business without knowing which one is in hand. The note is where that basis is disclosed.

Ind AS 109: Financial Instruments and Expected Credit Loss, so why does an impairment model appear in a guide about hybrids?

All of the above sits under one standard. The standard's name carries two quite different bodies of requirement, and a reader who has met only one of them will misjudge what the other is doing there.

So why do the two live under one name? Because both are consequences of the same question, asked from opposite ends. A financial instrument is an asset to whoever holds it and a liability or equity to whoever issued it, and one standard has to deal with both ends of the same contract. The classification and measurement requirements answer what an instrument is and how it is carried. The impairment requirements answer what happens when an instrument the business holds is not going to be collected in full. Ind AS 109 covers both the instruments a business has issued and the instruments it holds. A reader who assumes it is only about one of the two ends up surprised by the other.

Why is this judgement genuinely hard, and what does a disagreement usually mean?

Three things make it hard, and none of them is anybody's fault. The first is that the features are contractual, so they can be written in any combination a drafter can think of. A list of instruments would be endless and out of date the week after it appeared, so nobody publishes one. The second is that the requirements are principles rather than a catalogue: they set out what to ask, not which answer a particular instrument gets, and a principle applied to an unusual set of terms will produce arguments. The third follows from the first two. Two experienced people can read the same term sheet, agree on every fact in it, apply the same requirements honestly, and reach different conclusions.

The same thing happens outside accounting. Two experienced building surveyors look at the same crack in the same wall. Both have measured it, both have the same photographs, both know their work. One says it is settlement that finished years ago; the other says it is live and worth watching. Neither surveyor is lying to the other. Both are weighing the same evidence and reaching different judgements about what it means, and the honest response is for each to say what they concluded and why, so the household can see the reasoning rather than only the verdict.

A disagreement about the classification of a hybrid instrument is almost always a disagreement about the substance of an obligation, not an attempt by anybody to mislead anybody. The answer is arguable, and that is precisely why the judgement has to be disclosed. Nobody is suspected of anything. Hold that firmly, because the cynical reading is available and it is usually wrong. If classification were mechanical there would be nothing to disclose: the instrument would be stated and any competent reader would derive the treatment. The reason a business is asked to explain how it reached its conclusion is that a reasonable person might have reached a different one. Disclosure is the system's admission that judgement was needed, not its suspicion that judgement was abused.

How should a reader approach an instrument they cannot classify at a glance?

In four steps, and in this order. The order is what stops a reader arguing with a conclusion before understanding the terms it came from.

Step one, find the terms. The financial instruments note and the borrowings or share capital notes carry them, and if the instrument matters at all it will be described somewhere. Step two, ask what the business must deliver and when: an amount, on a date, to whom, and on whose trigger. Step three, ask what it can avoid: what can be withheld, deferred or declined without breaching anything. That question is the whole of the classification test and it is the one most readers skip. Step four, read what the business says it concluded and why. If the note sets out the instrument's terms but never explains how the judgement was reached, that absence is itself the finding, and it is a finding about the disclosure rather than a proof that the conclusion was wrong.

Notice what that last step does not authorise. A reader outside the business, holding a summary of the terms rather than the contract, is not well placed to overturn a conclusion reached by people who read the whole document and discussed it with their auditors. The reader's job is to understand the reasoning, to see whether it hangs together, and to notice when there is no reasoning offered at all. Understanding the reasoning is a great deal more useful than a verdict, and it is honest about what a reader can actually see.

Four instruments through the test: which response does each one need?

The whole argument now sits on one table. Four illustrative instruments, each carrying Rs 20,00,000, each run through the same question and each given the response that question produces. Anjani Stationers has issued none of them, its published equity stays at Rs 1,42,00,000 and its published liabilities stay at Rs 38,00,000 throughout.

The third and fourth columns, read together, carry the whole argument on one line each: the number of answers the test gives decides the mechanism, and the name of the instrument appears nowhere in that sentence. The convertible splits because two components genuinely exist inside one instrument. The redeemable preference share does not split, since its equity vocabulary is not an equity component. The perpetual does not split, for the mirror reason. And the loan splits nothing at all. Separating a feature is not splitting an instrument, and that distinction is the one the two mechanisms turn on.

Four readings of the panel carry the argument. The default state, repayable and with a mandatory return and nothing else, is a financial liability of Rs 20,00,000 with no mechanism needed. Turn on the conversion right and the same instrument becomes a compound one, splitting into Rs 16,20,921 of liability and Rs 3,79,079 of equity. Turn the first two clauses off instead, so nothing can ever be demanded, and the whole Rs 20,00,000 is equity however the certificate is titled. Turn on the outside link over a repayable instrument and a feature has to be separated and measured on its own, on top of whatever the rest of the instrument is. Five of the sixteen combinations carry a dashed edge on the map, and those are the ones where the honest answer is that reasonable preparers could differ.

Who reads a classification note, and what do they do with it?

Three people open the same note in the same week and none of them wants the same thing from it.

A lender reads a classification note for the cash it implies rather than the caption it produces, an analyst reads it to find out whether two businesses can be compared at all, and Vaidehi Rao reads it because she has to write the paragraph that explains the judgement. Take the lender first. A lender looking at an instrument classified as equity does not stop there. The lender asks what the instrument actually requires: is there a date on which someone can ask for money, is there a coupon that steps up so steeply that the business will realistically refinance, and does any of that fall inside the lender's own repayment period. An instrument correctly classified as equity can still carry a call on future cash, and the classification note is where a lender learns that it exists.

The analyst's use is comparability, and it is the reason the failure below happens. A difference of judgement between two businesses produces a difference in the ratios that has nothing to do with how either one trades. So before comparing two businesses on any funding measure, an analyst wants to know whether either has an instrument whose classification carries judgement. Where the note explains the reasoning, the analyst can restate one on the other's basis. Where it does not, the honest answer is that the two are not comparable on that measure and the analyst should say so rather than rank them anyway.

And Vaidehi Rao, as finance controller, has the hardest job of the three. She has to write down why. If Anjani Stationers ever issued an instrument of this kind, she would have to set out the terms, the question she asked of each limb, what she concluded and what made it arguable. The paragraph explaining why is unglamorous and it is the single most useful thing in the note. It is what lets a reader outside the business follow the reasoning instead of guessing at it. A note that says what was concluded without saying why has done the easy half.

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