Pro Forma Financials: Adjusted Numbers and Their Limits

Pro forma means as if. Figures are restated to show what a period would have looked like under an assumption, most often that an acquisition had happened at the start of the year. Adjusted means the same period with selected items taken out. Both are legitimate, both are unaudited, and both are only usable when reconciled line by line to a statutory figure a reader can check.

Start with a household. The two operations are things people already do at a kitchen table. A household spent Rs 4,80,000 last year and wants to know whether the coming year is affordable. Two different questions get asked in that conversation and they are almost never separated. The first is what the year would have cost had the second child started school in April rather than in October. Moving the school start date changes nothing that happened and asks what a full year of the new arrangement looks like. The second is what the year cost leaving out the wedding. Leaving the wedding out changes no dates at all and lifts one line out of the total. The first question moves the calendar and the second removes an item, and a household that answers one while thinking it answered the other budgets for a year it will not have. Businesses do exactly the same two things, give them names that sound similar, and readers conflate them daily.

Three things carry into what follows. The first is that published statements are prepared under a framework, checked and signed. A figure inside them is therefore a statutory figureA number that appears in the audited financial statements themselves, prepared under the framework the business reports under. A statutory figure is one a reader can trace and an auditor has looked at.. The second is that one business can buy another part way through a year and that the statements which follow carry the bought business only from the date it arrived. The third, from earnings quality work, is that a business has views about which of its own costs are representative. The framework, the acquisition date and the business's own view of its costs together show what happens when a business publishes a number that is not in its own statements. Everything below runs on Anjani Stationers Private Limited, invented, a maker of school exercise books with a 70 per cent stake in Chitra Binding Works. One fact about it deserves saying before anything else: Anjani Stationers publishes no pro forma statement and no adjusted measure of any kind. Anjani Stationers is a private limited company that produces a statutory annual report and nothing besides. Every adjusted amount below is a labelled hypothetical rather than something Anjani Stationers reported.

What does pro forma mean, and how is adjusted a different operation?

Pro formaLatin for as a matter of form. In accounts it labels figures restated under a stated assumption, most often that a transaction happened at a different date from the one it actually happened on. figures answer a what-if. Something happened during the year, and the reader is shown what the year would have reported had that something happened earlier, usually on the first day. Nothing is taken out. No line is deleted, no cost is declared unrepresentative, and no judgement is made about whether any item belongs. A period is redrawn under a stated assumption about timing, and the assumption is the whole of the operation.

Adjusted earningsA profit figure a business publishes after removing items it says are not representative of ordinary trading. Adjusted earnings sit outside the audited statements, and each business decides for itself what comes out. answer a what-without. The period stays exactly as it was, every date is untouched, and selected items are lifted out of the reported figures because the business says they do not describe ordinary trading. Nothing is assumed about timing. Items are removed, and the removal is the whole of the operation.

Pro forma and adjusted are two different operations on two different axes, and a figure described as both has not been described at all. That is not pedantry about words. A reader who is told a figure is pro forma expects a timing assumption and will go looking for it in the note. A reader who is told a figure is adjusted expects a list of removals and will go looking for that instead. Hand either reader the wrong label and they will hunt for a disclosure that was never made and conclude, wrongly, that it is missing. The first question about any such number is which operation was performed, and where the answer is both, the next questions are which items and which assumption, separately.

When are pro forma figures a genuine service to the reader?

Pro forma has a reputation it only half deserves, so here is the honest case for it, and it is a strong one. When a business buys another part way through a year, the statements that follow describe a period that never existed as a business. Consider the two halves of that reported year. For part of the year the reported figures cover one business, and for the rest of the year they cover a larger one. The total is arithmetically correct, fully audited and completely faithful to what happened. The total also matches no business a reader can hold in mind: not the business that existed in April, and not the business that exists now.

Think of a bakery that ran alone until October and then took over the sweet shop next door. Its year to March reports six months of bread and six months of bread plus sweets. Ask the owner what the shop turns over in a year and no number in those statements answers the question. The one number that comes closest was never printed, and it is a full year of the shop as it now stands. Pro forma figures exist because the reader has a real question the statutory statements structurally cannot answer, and answering it is a service rather than a device.

The bakery gives the test for whether a pro forma presentation is doing its job. Does it answer a question the reader genuinely has and the statements genuinely cannot? A full year of the combined business after a mid-year acquisition passes that test easily. A reader working out earnings per share on the old share count is working out something already untrue, so a period redrawn after a share issue that completed after the year end but before the report was published passes as well. A pro forma presentation fails the test when a flattering assumption has been chosen over a clarifying one. To tell the two apart, read the assumption and ask whether the same question would have been worth answering had it moved the number the other way.

Which adjustments are ordinary, and which need justifying?

Adjustments are not one thing. Adjustments sit on a ladder, and where one sits on that ladder shows how much justifying it needs before it can be accepted. The ladder runs from the bottom up.

At the bottom is removing something that genuinely happened once. A godown floods and a quantity of paper is written off. The write-off is real, it is cash, it happened, and it says nothing about what next year will cost, so taking it out of a figure meant to describe ordinary trading is uncontroversial. A one-offAn item a business says arose from something that will not recur, so it does not describe ordinary trading. The word is a claim by the business rather than a defined accounting category. is the least contestable adjustment there is, and the only guard needed is to check that it really did happen once.

One rung up sits removing the transaction costs of an acquisition: the legal work, the diligence, the stamp duty on buying a business. Removing transaction costs is ordinary and widely done, and one thing about it is worth noticing. The adjustment travels in one direction only. The costs of buying come out. The earnings the purchase brought in stay firmly in. The one-way traffic is not dishonest, and it is a reason to hold the adjustment a little more loosely than the flood.

Another rung up sits share-based paymentRemuneration settled in shares or options rather than cash. Share-based payment is a genuine cost of employing people, measured and charged to profit, and the only unusual thing about it is that it never leaves the bank account., and this is where a reader has to be careful, because the argument for removing it sounds better than it is. The argument runs that it is not cash. The claim is true and it is not the point. Somebody worked, somebody was paid for working, and the payment was made in shares instead of rupees. Paying in shares transfers value from existing holders to employees just as surely as a bank transfer would. Removed, it leaves a description of a business that gets its staff free. The further up this ladder an adjustment sits the more justifying it needs, and the depreciation and amortisation charge of Rs 12,00,000 that Anjani Stationers carried is the clearest illustration of why: it is non-cash, it is entirely real, and removing it casually would describe a business whose machines never wear out.

At the top sits removing restructuring that keeps happening. A business that reports a restructuring charge in year one has had an unusual year. A business that reports one in year one, year two and year three has a cost of doing business it has chosen to call unusual three times running. There is nothing improper about the charge. The contestable part is the claim that the charge does not describe ordinary trading. Three consecutive years of it are precisely what ordinary trading looks like at that business.

Why can two businesses adjust the same way and still not be comparable?

Here is the failure that costs the most and looks the least like a failure. Two businesses both publish an adjusted profit, both say they have removed restructuring, and an analyst lines the two figures up. The word matches. The figures do not mean the same thing, and neither disclosure says so.

Three separate mechanisms produce that outcome, and they can operate together. The first is definition. Restructuring at one business means the redundancy payments and nothing else. At the other it means the redundancy payments plus the write-off of stock at the closed unit plus the cost of exiting the lease plus a reorganisation of the way deliveries are routed. Same word, four times the content. The second is period. One business removes an item in the year it was charged and the other spreads its removal across the two years the programme ran, so a single reader comparing a single year is comparing an item that is fully out against an item that is half out. The third is direction. A business that removes unusual costs and leaves unusual gains in has an adjustment policy that only ever moves the figure one way, and a policy that can only help is not a measurement rule.

Work it with numbers. The size surprises people. Take two businesses with identical statutory profit after tax of Rs 30,00,000 and identical underlying trading. The first defines restructuring narrowly and removes Rs 1,10,000, reporting adjusted profit of Rs 31,10,000. The second defines it broadly and removes Rs 1,10,000 of redundancy plus Rs 2,20,000 of stock written off at the closed unit plus Rs 2,00,000 of delivery reorganisation. The three come to Rs 5,30,000, and the second reports adjusted profit of Rs 35,30,000. Both call the line adjusted profit after removing restructuring. The gap is Rs 4,20,000, or 14.0 per cent of the statutory figure the two started from, and every rupee of it is a definition rather than a difference in trading. An adjusted figure is a private measure carrying a public-sounding name, and two businesses using the same name have agreed on the name and on nothing else.

What must always sit beside an adjusted figure?

One thing, and nothing else does the same work. A reconciliationA line by line walk from a figure in the audited statements to the figure being presented, showing every item added or removed and what each one is, so a reader can retrace the whole distance. to the nearest statutory figure, itemised, in the same place as the figure itself. Not a footnote saying the figure has been adjusted for exceptional items. A walk: this is the audited number, these are the items, this is what each one was, and this is where the walk lands.

The reconciliation earns its place by doing four things. The walk lets a reader rebuild the statutory figure, so the business is not the only party able to check its own number. The walk names each item, so some can be accepted and others refused rather than the whole adjustment being swallowed or rejected entirely. The walk shows the size of each item, so the reader can see whether a flood moved the figure or a definition did. And the same disclosure reaches every other reader. Reaching everyone is what turns a private measure into something a market can argue about.

The reconciliation is the single test of whether an adjusted figure is being offered as information or as a substitute, and a figure presented without one should be set aside rather than argued with. The words set aside are doing precise work. Arguing with an unreconciled figure means treating it as a claim that can be evaluated, and it cannot be: what came out, in what amounts, in which direction and over what period are all unknown. An unreconciled figure is not approximately right and it is not probably wrong. It is not a measurement of anything anybody can name. The response is to put it down, pick up the statutory figure, and work from that. Putting it down takes less time than arguing and it is the only response that cannot be gamed.

What would a pro forma statement look like for Anjani Stationers?

About the rupee amounts that follow. The year two figures, profit after tax of Rs 30,00,000 among them, are the same set used above, so nothing met earlier has quietly moved. Anjani Stationers publishes no adjusted measure of any kind and never has, so every adjustment amount in the walks below is a labelled hypothetical.

Start with the detail that decides the whole example. Anjani Stationers acquired its 70 per cent of Chitra Binding Works at the start of year two, not part way through it. Its consolidated figures for year two therefore already carry twelve months of Chitra Binding, and the Rs 3,22,00,000 it reports as consolidated revenue covers twelve months of both businesses. No pro forma adjustment arises for Anjani Stationers at all, and the absence is the most useful thing the example has to teach. The right answer to whether a business needs a pro forma statement is very often no, and a reader who knows why it is no has understood the mechanism better than a reader who can only build one.

To see the mechanism, take the counterfactual that did not happen. Suppose instead that Anjani Stationers had bought Chitra Binding at the half year rather than on day one. Chitra Binding Works billed Rs 60,00,000 over a full twelve months. Binding invoiced to Anjani Stationers Private Limited accounts for Rs 8,00,000 of that, and the remaining Rs 52,00,000 went to outside customers. Revenue is taken to have accrued evenly across the twelve months. The even spread is a stated assumption rather than something the business reported. Then a half year of Chitra Binding is Rs 30,00,000 of revenue, of which Rs 4,00,000 was invoiced to Anjani Stationers and falls away on consolidation. The statutory consolidated revenue under that assumption would be Rs 2,70,00,000 plus Rs 30,00,000 less Rs 4,00,000, and that comes to Rs 2,96,00,000.

Now the pro forma view of that same counterfactual. The pro forma view answers the question a reader would actually be asking, namely what the group turns over in a year. Take twelve months of Chitra Binding Works instead of six: Rs 2,70,00,000 and Rs 60,00,000 together, with the Rs 8,00,000 of internal binding stripped out, land at Rs 3,22,00,000. The bridge between the two is Rs 26,00,000, being the first half of Chitra Binding's revenue of Rs 30,00,000 less the Rs 4,00,000 of it that was internal. Work the same walk on profit, holding the assumption that no unrealised profit sits in closing stock so that eliminating the internal binding charge removes an equal amount of revenue and cost. Profit for the period would be Rs 35,00,000 under the half-year assumption and Rs 40,00,000 on the pro forma view, and the split between the owners of Anjani Stationers and the non-controlling holders of the other 30 per cent of Chitra Binding walks with it.

Read the arithmetic across each row and then down each column. Both directions have to hold before a walk is worth anything. Across: Rs 26,00,000 added to Rs 2,96,00,000 gives Rs 3,22,00,000, and Rs 5,00,000 added to Rs 35,00,000 gives Rs 40,00,000. Down: Rs 33,50,000 plus Rs 1,50,000 is Rs 35,00,000 in the statutory column, and Rs 37,00,000 plus Rs 3,00,000 is Rs 40,00,000 in the pro forma column. Every adjustment is a timing assumption about one date, and not one item has been removed from anything.

The closing move is why an example that did not happen was worth running. None of the left-hand column exists. Anjani Stationers bought Chitra Binding on day one of year two, so the Rs 3,22,00,000 it actually reports for consolidated revenue is a statutory figure, audited, sitting in the statements. The pro forma column and the real statutory figure land on exactly the same Rs 3,22,00,000, and they are not the same kind of number at all: one would have been an unaudited restatement under an assumption, the other is what the business reported. Same digits, different status, and status is what a reader is reading for.

How should an adjusted figure actually be used?

Four steps. The fourth keeps the first three honest, so do not treat it as the tidy-up.

Step one is to read the reconciliation before the figure. Physically before. The headline number is designed to be memorable and it will anchor the reading, so the walk comes first and the headline second, in that order, every time. A reader who takes the figure first spends the rest of the disclosure arguing with an anchor rather than assessing a measurement.

Step two is to recompute the statutory figure from the walk. Recomputing takes under a minute and does two things. First, it confirms the walk actually reconciles, and reconciling is not automatic. Second, it puts the audited number in the reader's hand in the same working session as the adjusted one. A walk that does not add up is a finding all by itself.

Step three is to ask, item by item, whether that adjustment would have been made had it gone the other way. Would a gain of the same nature and size have been removed as well? If unusual costs come out and unusual gains stay in, the business has a direction rather than a measurement rule. The reversal question tests the policy rather than the item, so it does more work than any list of permitted adjustments.

Step four is to carry both numbers, all the way through, and never let one of them drop. The statutory figure and the adjusted figure go side by side in the reader's own notes. Both get quoted. Both get compared across businesses. No analysis can then come to rest silently on the unaudited figure alone, and carrying both numbers is what makes the other three steps safe. This is where readers lose it in practice. Not by being fooled at the moment they read the adjustment, but by writing the adjusted figure into a spreadsheet in March, opening that spreadsheet in September, and finding a column of numbers with no memory of which sort they were.

The readings the panel produces run as follows. At the default nothing is switched on and both figures read Rs 30,00,000. Walk the slider up the ladder and the adjusted figure moves to Rs 31,20,000 after the flooded stock, Rs 32,00,000 after the acquisition costs, Rs 34,20,000 after share-based payment and Rs 37,50,000 after the third consecutive restructuring charge. The full set lifts profit by Rs 7,50,000, or 25.0 per cent, and turns statutory earnings of Rs 7.50 a share into an adjusted Rs 9.38 a share on the same 4,00,000 shares. The interesting setting is rung two. A reader who accepts only the two least contestable adjustments holds Rs 32,00,000, and the remaining Rs 5,50,000 of the total uplift comes entirely from the two rungs that owe an argument.

Who has to work with adjusted figures, and how does each of them handle one?

Three readers use these figures daily, and watching what each actually does with one is more instructive than any rule.

A lender's credit officer at a bank assessing a working capital limit treats the adjusted figure as a claim and the statutory figure as the base. The covenant in the facility agreement will be written against something specific, and which of the two it is written against is a negotiation, not a technicality. A borrower who wants the covenant tested on adjusted earnings is asking to be measured against a figure it defines itself. A credit officer who agrees to that without also fixing the definition in the agreement has handed over the measuring stick. The practical move is the one described above: rebuild the statutory number, then decide item by item which adjustments the facility will recognise, then write those into the document.

An equity analyst building a forecast uses the adjusted figure differently, and often quite properly. A forecast is about what next year will earn, so removing a flood that will not recur is not a distortion, it is the whole job. The careful analyst rebuilds both series, statutory and adjusted, for every year in the model, and keeps them in adjacent columns so a comparison never mixes the two. A category that quietly widened in year three shows up as a step in the adjusted series that the statutory series does not have, so the habit also catches definition drift.

A household does the same thing without the vocabulary, and this is where the idea becomes recognisable. Sitting down to work out whether the school fees are affordable next year, a household naturally sets aside the money spent on a wedding. There will not be another one. Setting the wedding aside is an adjustment and a sensible one. The trap arrives when the same household sets aside the hospital bill, and the car repair, and the roof work, each of which was genuinely unexpected, and concludes that it can afford fees against a figure from which every unexpected thing has been removed. Unexpected things happen every year, so a figure with all of them removed describes a year that has never once occurred. The contestable end of the ladder fails in exactly that way, written in household terms. The fix at a kitchen table is the same as the fix in a credit file: keep the number actually spent beside the number the household would like to have spent, and make the decision with both in view.

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