Management Discussion and Analysis: Reading It Against the Numbers

Management discussion and analysis is the part of an annual report where the board's own team sets out what the year contained and why, in prose rather than in figures. The section is drafted inside the company, approved by the board, and read against the audited statements for contradictions rather than audited in its own right. Management's authorship is what makes the section valuable and what makes it partial: the writers chose every sentence in it.

Consider the note a mechanic writes at the foot of a bill. The bill above it is arithmetic anybody can check: four parts, so many hours of labour, tax, a total that adds up. The line underneath, in pen, says something the arithmetic never could. The front pads were worse than expected. The rear discs will need doing before the next monsoon. The rattle the customer complained about has gone and here is what it was. The line in pen is the most useful thing on the bill and also the only thing on it nobody verified. The mechanic who did the work wrote it, and that mechanic would quite like the customer back next year. An annual report has exactly those two halves, and management discussion and analysis is the line in pen.

A reader arrives holding a good deal already. The shape of an annual report and roughly what is bound into it are established already. So is the source hierarchy, in which a filed document outranks a slide. So is the way a reader talks themselves into a conclusion by meeting the conclusion first. And from the earnings quality work come four specific movements in the numbers of Anjani Stationers Private Limited, an invented notebook maker whose figures are published across earlier work and none of which move here. The new ground is management discussion and analysisThe narrative section of an annual report in which management sets out its own account of the year: conditions in the industry, how the business performed, what it is worried about and what it expects. Often shortened to MD and A in conversation. itself, and the discipline of reading the narrative beside the numbers rather than instead of them.

Who writes this section, and how much of it did the auditor check?

Management writes it. In a large company the finance function drafts it, the operating heads supply the parts about their own businesses, the legal and secretarial team checks the wording, and the board approves the final text before it is bound into the annual report alongside the statements. At Anjani Stationers, if it produced one, the drafting would sit with Vaidehi Rao, the finance controller, and the whole thing would run to two sheets rather than twenty. The size changes; the authorship does not.

Now the part readers get wrong. An audit opinion attaches to the financial statements and the notes behind them. The opinion does not attach to the narrative sections travelling in the same binding. The auditor does read them, and considers whether what they say is materially inconsistent with the audited figures or with what the audit team learned while doing the work. If the narrative claimed a margin that the statements contradict, that is something the auditor is expected to notice and pursue. But noticing an inconsistency is a far lighter job than forming an opinion, and no opinion is formed on the narrative at all. The narrative travels as unauditedNot covered by the auditor's opinion. The information may still be accurate; it simply has no external professional assurance attached to it, so its reliability rests on the preparer alone. material inside an audited document.

Think of a school report card. The subject marks were entered from a mark sheet, added, checked and moderated. The line at the bottom in the class teacher's handwriting, saying the child has settled well this term and could apply himself more in mathematics, was moderated by nobody. The teacher's line is very probably true, often the most useful sentence on the card, and backed in a completely different way from the marks above it. The audit opinion reaches the statements and the notes; the narrative sections sit outside it and carry no opinion of their own. A reader who assumes the whole bound document carries one level of assurance has put the boundary in the wrong place, and every judgement built on that assumption inherits the error.

What does this section ordinarily cover?

Six areas turn up again and again. Knowing what should be there is what makes an absence noticeable, so the six are worth holding before anything is opened. Conditions in the industry the business operates in. How the business itself performed, usually with its own account of its strengths, weaknesses, opportunities and threats. The risks and concerns management says it is carrying. Whether the internal control systems are adequate for the size of the operation. Financial performance discussed against operational performance, the area that connects the numbers to what actually happened in the warehouses and on the routes. And developments in human resources, meaning headcount, hiring, attrition and the state of industrial relations in most narratives.

The six areas are the ordinary contents. A listing requirementAn obligation that applies to a company because its shares are admitted to trading on a stock exchange, rather than because it is a company. A listing requirement sits on top of the obligations every registered company already carries. rather than anything in company law itself is what puts the six into a particular company's annual report, and the market regulator's own text sets out the precise list. The six above are what is ordinarily there rather than what is mandatory.

One split inside the six is worth holding on to. Three of the six areas land on lines that can be checked and three do not. Financial performance against operational performance is testable. Every claim in it is about a number printed elsewhere in the same document. The business performance discussion is partly testable. Volume and price claims show up in revenue and margin. Human resources is partly testable too. A headcount claim sits next to an employee benefit expense line. Industry conditions, risks and concerns, and the adequacy of internal control are statements about the world and about management's own judgement, and no line in the accounts confirms or denies them. Three of the six areas can be tested against the statements and three cannot, and separating them before reading is what turns reading into checking.

How to Read Management Discussion and Analysis, and what does the order do?

The method has three steps and its whole power is in the sequence. Step one, compute the movements from the statements independently and write them down. On paper, on a screen, anywhere, so long as the list exists in a fixed form before the narrative is opened. Step two, read the narrative straight through without arguing with it. Arguing halfway means finishing having read half. Step three, return to the list and mark each item: addressed and explained, addressed but in a way the numbers do not support, or not mentioned at all.

The third mark is the one the method exists for. A movement in the accounts that the narrative never names is a fact about the document, and it is a fact no sentence in the document contains. The missing movement is visible only to a reader who wrote it down first. Holding the list first is what reading againstReading a narrative with an independently prepared list of questions in hand, so the narrative is checked rather than absorbed. The opposite of reading a narrative and then going to the numbers it pointed at. means, and the activity differs entirely from reading a narrative and then going and looking at whatever it mentioned.

Why the order and not the other order? Because a narrative is written by people who chose what to put in it, and the themes they chose become the reader's themes the moment they are read. Nobody decides to accept the framing. An hour later the questions are simply all about the things the document raised, and the questions never formed leave nothing to notice. A house viewing works the same way. An agent who walks a visitor round talking about the light in the living room sends them away remembering the light. Damp, parking and water pressure written down before the bell is rung produce three answers or three silences, and the silences are worth as much as the answers. Reading the narrative first quietly decides what a reader is capable of noticing afterwards. The order is not a preference or a style of working. The order is the entire method.

The same finding arrives from the other direction in preparing for an earnings call. A management presentation that runs for forty minutes before questions open will have reorganised a listener's priorities by minute six. The discipline there is to write the questions down before the call. Two different documents, two different rooms, one identical mechanism: whoever speaks first sets the agenda, and the only defence is a list written beforehand.

Which three readings pay for the time they take?

The first reading is what is explained against what is passed over. Two movements of similar size sit in the accounts. One gets a paragraph with a cause, a quantum and a plan. The other is not named anywhere in the document. Nothing untrue has been written and no rule has been broken, but the difference in treatment is itself information, and it is information that exists in no sentence anybody wrote. A narrative that explains a favourable movement in detail and passes over an unfavourable one of the same size has said something, and what it said is not written down anywhere in it.

The second reading is what changed in the language since last year. Narratives are edited far more often than they are rewritten. Somebody opens last year's file, updates the figures, adjusts the sentences that no longer hold and leaves the rest. Carrying wording forward is not laziness but how documents get produced under a deadline, and the habit is enormously useful to a reader. Put two years side by side and the paragraphs that moved are where the year actually happened. A risk paragraph that gained two sentences, a customer concentration line that quietly lost the word temporary, an outlook sentence whose adjective softened. The side-by-side comparison takes ten minutes and almost nobody does it.

The third reading is what is claimed that the statements can check. Every narrative contains claims of two kinds. Claims about the world, beyond anything the accounts can test, and claims about the business, very often landing on a printed line. A claim about holding selling prices lands on the gross margin. A claim about tightening collections lands on the receivable book and its ageing. A claim about controlling costs lands on the expense lines beneath the gross profit. Each of those is checkable in under a minute, and the answer is not believe or disbelieve, it is consistent or inconsistent. The distinction matters. An inconsistency is a question rather than an accusation, and the honest response to one is to go and find the explanation rather than to assume there is not one.

What do the numbers at Anjani Stationers ask a narrative to explain?

Anjani Stationers Private Limited produces no management discussion and analysis at all, for a reason its private status settles below. The absence makes it the right entity to run the method on, and it forces the honest version: compute first, and see how much is in hand before any narrative exists to shape it. Every figure below is already published and none of it moves here.

Step one of the method, done properly. Six lines, five movements, and a divergenceTwo related figures that ordinarily move together moving apart instead. Receivables and revenue, or stock and sales, are the usual pairs. A divergence is a question, not a conclusion. in four of them. Revenue rose 12.5 per cent, from Rs 2,40,00,000 to Rs 2,70,00,000. Gross trade receivables rose 21.8 per cent, from Rs 78,00,000 to Rs 95,00,000, running 9.3 percentage points faster than the sales that created them. Inventory rose 47.4 per cent, from Rs 19,00,000 to Rs 28,00,000, running 34.9 points faster than revenue. The gross margin did not move at all, holding at 45.0 per cent in both years. The operating margin fell from 22.1 per cent to 15.4 per cent, a drop of 6.7 points that therefore sits entirely below the gross line. And the provision for doubtful debts went from Rs 3,00,000 to Rs 9,00,000, a tripling against a book that grew by a fifth.

All four have ordinary explanations, and those explanations are published elsewhere in this library. The existence of ordinary explanations is the point rather than a complication. The list is not an accusation and it is not a finding; it is four questions, held in hand, before anybody has had a chance to say what the year was about.

Now step two. Anjani Stationers files no such narrative, so what follows is a described one, written to show the shape of the problem rather than filed or quoted from anybody. Suppose a narrative arrived with three themes: a year of growth, with revenue up 12.5 per cent on three new school accounts; investment ahead of the season, with stock built and warehouse space committed; and input cost pressure, with paper prices firm through the year and the company working to hold its selling prices.

Step three is where the method pays. The four movements go to that narrative one at a time. Inventory is addressed, and addressed well: building stock ahead of the school year is a real reason and it fits. The margin theme is addressed but does not survive contact with the numbers. If input costs had genuinely squeezed the business the gross margin would have moved, and it did not move at all. The whole 6.7 point fall is below the gross line, so it belongs to freight, warehousing, provisioning or something else in that region, and the narrative has pointed at the wrong floor of the building. And two of the four movements, the receivable book outrunning revenue and the provision tripling, are not mentioned in the narrative anywhere. BoilerplateWording carried forward from a previous document with little or no change, general enough to be true of almost any company in almost any year. Carrying wording forward is standard practice, not a trick, and it carries no information about the year in front of the reader. would not have covered them either, because a sentence about continued focus on working capital management contains no information about a book that grew 9.3 points faster than sales.

Where does this section fall down as a source?

Three weaknesses, worth stating plainly rather than darkly. A reader who treats the section as untrustworthy will throw away the one thing in the annual report that carries intent.

The first is assurance. The narrative is not audited, and while the auditor reads it for consistency, no professional opinion attaches to any statement in it. The narrative is the most readable part of the document and the least verified part at the same time. The combination is uncomfortable, and worth remembering every time a sentence in it is persuasive.

The second is interest. The narrative is written by people who ran the business being described, whose pay, standing and next job are affected by how the year is understood. An interest of that kind does not make the narrative untrue, and naming it accuses nobody. The interest is a structural fact about the document, and the section will therefore emphasise what management believes matters and use the framing management finds natural. Every annual report in the world has this property.

The third is boilerplate, and it is by far the most common. General wording, carried forward, true of any company in any year. Challenging market conditions. Continued focus on cost efficiency. Adequate systems commensurate with the size and nature of operations. None of this is a lie. Boilerplate is what gets written when a section is required, nobody has been given time, and last year's file is open on the desk. Boilerplate is not dishonest, it is uninformative, and the correct response is to record that a question went unanswered and go and answer it from the notes rather than to read a meaning into wording that was never about this year. Reading suspicion into boilerplate is how a careful reader manufactures a finding out of nothing, and it does real damage to ordinary businesses whose narrative was written by a busy person in an afternoon.

Why does Anjani Stationers Private Limited file none of this?

Because it is private. Anjani Stationers is a private limited company whose 4,00,000 ordinary shares of Rs 10/- are held by two founding households and one outside holder who put money in some years ago. No shares trade on an exchange, so none of the obligations that attach to a listed company attach to it. The company publishes financial statements, notes, an auditor's report and a directors' reportThe board's own report accompanying the accounts, which every company registered under the Companies Act prepares. The directors' report carries a narrative account of the year alongside a set of statutory disclosures., and it files an annual return. Anjani Stationers publishes no management discussion and analysis, no investor presentation, no quarterly results, no earnings call and no shareholding pattern filed with any exchange.

The company does have the narrative section inside its directors' report, and seeing that section is genuinely useful. The instinct is the same at a much smaller scale. The board writes an account of the year: how trading went, where the money was invested, which risks the board is watching. The directors' report narrative is shorter, is not organised into the six areas, and is nobody's full-time job to write. But it is written by the same interested parties, it carries the same absence of assurance, and every one of the three readings set out here works on it exactly as well. Two years of a small company's directors' report narrative set side by side show, in the sentences that moved, where the year happened, just as they would in an annual report ten times the length.

The contrast is the whole value of a private company here: it shows where the line falls between what every registered company publishes and what only a listed one additionally files, and a reader who can see that line knows which documents will exist before opening anything. A reader who knows the line does not waste an afternoon hunting for a management narrative that was never required, and does not treat its absence as a finding about the business. The absence is a fact about the company's status. Nothing more, and nothing less.

Who reads the narrative for a living, and what do they do with it?

A lender goes to two places and almost nowhere else. The risks and concerns paragraphs, to see whether anything named there threatens the cash that services the loan, and the internal control paragraph. A lender to a mid-sized business is lending partly on the quality of the bookkeeping. Then the lender puts the narrative down and goes to the covenant lines, and the narrative has done its job if it produced one question the covenant test can answer.

An equity analyst uses the narrative to build the question list for the next call, running the same three steps in reverse gear: compute the movements, read the narrative, and every movement the narrative passed over becomes a question with a name and a number attached. An analyst who arrives at a call with four such questions is in a different conversation from one who arrived having read the narrative and nodded.

A diligence team acquiring a business pulls three years of narrative and reads them as one document. The drift between them is the cheapest signal available. A risk that appeared in year two and disappeared in year three either got solved or got quiet, and finding out which is a morning's work that occasionally changes a price.

And the same muscle works away from finance entirely. A landlord's annual letter to tenants, a school's note to parents about the fee revision, an employer's town hall about a reorganisation: each is a narrative written by an interested party about numbers the reader can partly see. Questions written down before it is read make the silences stand out as clearly as they do in an annual report.

The two paths produce different endings. At the default nothing is chosen, no movement has been tested and the panel says only that the choice decides what gets noticed. With the narrative taken first, the reader finishes holding three questions, all of them about the three themes the narrative raised, with the inventory build and the input cost story accepted as told and the receivable book and the provision never raised at all: two of the four movements do not appear on the list, and nothing in the document will ever say they are missing. With the movements taken first, the reader finishes holding five questions, with all four movements tested, one of them found to be addressed well, one found to be addressed in a way the flat 45.0 per cent gross margin does not support, and two found to be unmentioned. Same reader, same documents, same afternoon, and the difference between two tested movements and four is entirely a matter of which file was opened first.

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