Management Quality: The Signals That Are Actually Observable

Management quality cannot be read off a set of published accounts. The accounts carry decisions with consequences attached. One business paid Rs 21,00,000/- for control of the stage that governed its whole output rate, and on the day the shares moved that stage still ran at 100 an hour. The decision is visible, the consequence is visible, and the person who decided is not.

Three decisions, one business, one year. So who decided each of them?

Before anything gets defined, look at a record. Anjani Stationers Private Limited, an invented manufacturer, turns paper into school registers and sells them to a group of schools. In one published year it took three decisions large enough that the accounts had to say something about each, and every one of the three has an outcome written down beside it.

The first was a purchase. At the start of its second year the business put Rs 21,00,000/- into a 70 per cent stake in Chitra Binding Works Private Limited. Chitra had sat under separate control until that day, and had been sending in an invoice for binding. The second was a spending decision. Across the same year the business put Rs 24,40,000/- more into its standing baseThe costs a business carries every year whether or not it sells anything, such as rent, salaried people and the wear on machines it already has. than it had carried the year before. The third was a decision not to pay anything out. No dividend went to the holders, and the whole Rs 30,00,000/- of profit after tax stayed inside, lifting retained earningsProfit a business kept rather than paid out, added up across every year since it started. to Rs 1,02,00,000/-.

Now the question that matters. Who decided each of the three?

Answer it honestly and the answer is uncomfortable. The record names a person who decides, Anjani Kulkarni. The record names a finance controller, Vaidehi Rao, and mentions twice, in passing, that there is a board of directors. And it attaches none of those three to any one of the three decisions. Nothing says the purchase was hers, nothing says the spending was signed off by him or her or them, and nothing says who argued against the dividend. A decision is published, its consequence is published, and the decider is not, and that is the ordinary position of every reader standing outside a business.

The shape is familiar from somewhere much closer to home. Consider a household where the electricity bill halved between one quarter and the next. The bill records the change precisely: the units, the amount, the dates. The bill does not record who unplugged the second refrigerator, or whether anybody argued about it at the time, or whether the person who did it had been meaning to for two years. The evidence of the decision is complete and the evidence about the decider is missing entirely, and no amount of squinting at the bill will produce the second from the first.

So what is a reader outside actually able to observe?

Most people arrive at this question expecting a set of judgements about people. Is she any good? Does he have foresight? Are they the sort who will keep their nerve in a bad year? The word quality invites exactly those questions, and there is nothing foolish about asking them. The trouble is that the instrument to hand cannot answer any of them, and it has just been watched failing: three decisions, three published outcomes, and nought names.

So make the swap explicitly rather than drifting into it. The observable set is a decision, its date, its price where a price was paid, and what the accounts recorded afterwards. Intent, competence, foresight, character and effort are not observable at all. The second list is not missing because anybody hid it. None of those five involves an exchange between the business and anybody else, and an exchange is the only event a set of accounts is built to capture.

Follow that reasoning one step further. The next step is the load-bearing one. A set of accounts records things that moved between two parties: money out for a machine, money in from a school, an invoice raised, a liability accepted. Each has two sides and each can be checked by a stranger against somebody else's record of the same event. When a manager thinks hard for three weeks and then decides well, nothing moves between the business and anybody at all. There is nothing to write down that a stranger could verify. The silence is not a defect the accounting system is working towards fixing. The silence is the boundary of the system, and the boundary is the whole reason the observable set is a set of decisions rather than a set of people, and it is smaller than anybody expects.

A refusal that buys nothing is just a refusal, so what the swap buys is worth naming. A decision can be checked against what happened next. A decision can be argued about by somebody who knows the trade and disagrees. A decision can be handed to the next reader with the working attached, and a person who joins the file in two years then picks up the reasoning and tests it rather than inheriting a mood. An impression of a chief executive on a call does none of those three things. There is nothing under the impression to point at, so it cannot even be disagreed with usefully.

What does the record carry about a business that is Founder-Led?

The phrase has a plain gloss, and the operative half of it stands unchanged: run by the people who started the business and who still hold most of its shares. The other arrangement stands opposite, where the steering is done by managers brought in to act for holders who stand outside. The gloss is worth marking at once. The gloss holds a fact about who sits in the chair, and a fact about where the shares sit. The gloss holds not one syllable about skill.

Definitions are cheap, though. The work worth doing here is to ask what the record actually carries about such a business, and for Anjani Stationers the honest answer runs to four published sentences and then a full stop.

Anjani Kulkarni decides. She stands to lose the business itself, so on opening the statements she goes straight past the profit line to the cash in the bank and to the school fees still unsettled. She turns to the balance sheet with one question in mind. How much of everything under her control was funded out of her side rather than borrowed from somebody else's? And the house she lives in was kept out of the business altogether, so neither it nor the borrowing she took on privately to buy it appears in these accounts at all.

The stopping point here is itself the lesson. The four sentences describe an exposure and a reading habit, and neither of them is a level of skill or a track record. The sentences establish what she would lose and where she looks first. The same four sentences establish nothing about whether she looks at the right thing, or whether she has judged well before, or whether anybody has ever disagreed with her and been proved right. And nothing anywhere attaches her to a single share: the published register of membersKept inside the company itself: who is down as holding shares, and in what measure. No exchange keeps it and no exchange is sent it. for this business carries three entries, and not one of the three is a name.

In the everyday version the exposure becomes obvious in a way no diagram can beat. A person runs one stall outside one office building. The day's takings, the rent on the pitch and their own savings all live in the same tin. Nobody has to explain to that person what is at stake if the office moves, and nobody watching from across the road can say whether they are good at the job. The tin is visible. The judgement is not. The genuine differences between the two arrangements are covered under Founder-Led vs Professional Management: What Actually Changes.

And what does it carry under Professional Management?

Professional management means the business is run by people appointed to run it. Appointed managers report to somebody above them, and they did not start the thing. Apply exactly the same discipline from the other side of the arrangement: what does the record carry?

Vaidehi Rao is the finance controllerThe senior post carrying answerability for how a business keeps its books and how it reports them. Somebody is appointed into it; nobody is born into it.. She appears in the key management personnelThe handful of people whom the statements themselves identify as steering the place, gathered under that heading in the note on connected parties. grouping of this business's own related-party note. The grouping is precisely where a set of accounts points at the people who direct it. A profit movement that no change in the trading explains is one she must account for upward, so a classification in those accounts is something she has to read.

The record carries a reporting line and an obligation to explain, and it carries no measure of how well either is discharged. The second half is the half that gets skipped. Knowing that somebody must explain a movement upward establishes that the structure is there. The requirement establishes nothing about whether the explanation was any good, whether it was accepted, whether anybody pushed back, or whether the same explanation had been offered twice before.

A careless account invents, at exactly this point, the things the record does not carry. There is no tenure published for her. No appointment date. No pay. No previous employer. No performance measure of any kind, for her or for anybody, at this business or at any other business appearing in these notes. Writing a single one of those would manufacture the only evidence that mattered, and no reader downstream would ever be able to tell.

The everyday version is a shop manager who has to account for a quiet Tuesday to somebody who was not in the shop. The explaining is the job. The takings are the evidence. Whether the explanation was a good one is a thing that happened in a conversation, and conversations are not filed anywhere.

How Management Decisions Affect Business Quality: what did one purchase actually do?

One decision now gets worked from end to end, with figures. The purchase runs in four steps, and none of them is skippable. Each step is where a different reader goes wrong.

Step one: what was actually bought?

A register at this works passes through three stages in order. Cutting turns out 150 an hour, printing 125, and binding 100. Nothing can leave faster than the slowest of the three lets it out, so binding governed the pace of all three. Binding was also the one stage of the three that Anjani Stationers did not run. Binding belonged to Chitra Binding Works, which invoiced for the service. The finding stands in a single line. The business did not buy a supplier: it bought the constraintThe stage in a sequence of operations that runs slower than the others and therefore sets the pace of the whole line.. After the purchase Chitra became a subsidiaryHeld closely enough by another company for that company to decide how it runs, yet still trading under its own name, with its own people and its own books., still keeping its own accounts and its own name.

Step two: what happened on the day?

Here is the sentence this whole set of notes rests on. On the day the Chitra shares moved, binding still ran at 100 an hour, rated capacityThe output a works could turn out in a year if it ran at its slowest stage's pace for every hour available to it. was still 4,00,000 REGISTERS a year. The works put out registers that week at precisely the rate of the week before it. Not slower. Not faster. Identical.

A note on that figure before anybody writes it down, and it is not pedantry. The business produces the number 4,00,000 twice over in the same year, once as its rated capacity in REGISTERS a year and once as its ordinary SHARES in issue. Two units, two entirely unrelated divisions, one business, one figure. Any time 4,00,000 appears in a reading of this business, the unit gets named in the same breath, and a figure carried across without its unit is the cheapest error available here.

Step three: why did nothing move?

Because a share transfer and a machine are two different objects, and the ruling on that runs to eight words: the share certificate does not bind a single register. The 70 per cent gives a claim over the earnings of Chitra Binding Works, and a claim over earnings performs no binding and never will. Lifting binding above 100 an hour is a project in its own right, carrying its own bill and its own calendar, and the routes to it are equipment, hours or method. Every one of the three can be attempted and come to nothing.

Step four: so what was the decision, exactly?

Holding the stage bought the right to make the change, and it did not buy the change. Both halves of that are worth keeping. Somebody else controlled that stage, and that control put it beyond reach. Securing the standing to act on it is a genuine purchase. Standing to act and having acted are two separate states, and only one of them shows up on a shop floor. A reader who collapses the two will read a correct decision as a failed one, and will do it while getting every figure right.

The other two decisions of the year belong elsewhere, and each gets one sentence. The rise in the standing base and the movement in operating profit that followed it are worked together, and so is the decision to keep the whole year's profit rather than distribute any of it, under How Capital Allocation Shapes Long-Term Business Outcomes.

Why are the structures published somewhere else entirely?

Two records set side by side, and the gap between them does the work.

The listed side comes first, Aravalli Agro Foods, worked in full under corporate governance. The listed record carries a board of directors with its members named. The same record carries committees of that board, and a test of whether a director is independent, run on two axes rather than one. And it carries four signals a reader can genuinely read straight off a report: composition, attendance, dissent and the related-party pattern. Reading those four in order is the whole business of How to Analyse Ownership and Governance Signals.

Set Anjani Stationers beside it. The unlisted record carries three decisions, three outcomes, a person who decides and a finance controller. The same record mentions a board of directors twice and says nothing further about it either time. And against each of the three decisions, where a name would go, there is nothing at all.

On one side the structures are published and the people inside them are named, and on the other the decisions are published and nobody is named against them, and that is a difference of reporting obligation rather than a withheld disclosure. An unlisted company files no shareholding pattern with any exchange, publishes no board report and lodges no attendance record. Saying so describes a reporting regime; it does not accuse a business of anything.

A great many readers get that distinction wrong in a way that costs them, so it is worth saying plainly. The unlisted position is the ordinary situation for the overwhelming majority of companies in India. Almost every business an analyst will ever be asked about sits on the right-hand side of that pair, not the left. A reader who produces a confident verdict on management anyway has not produced a finding about the business. A reader of that kind has produced a measurement of their own confidence, and confidence measured against nothing is not evidence.

How to Assess Management Quality, when almost none of it is disclosed?

The answer stands in the first line and does not soften on the way past. Management quality cannot be assessed from a set of published accounts. Not carefully, not with a longer spreadsheet, not by weighting things more cleverly. The evidence a judgement of that kind would need is not in the document.

A refusal on its own is a shrug, though, and a shrug is worth nothing to the person who has to write something by Friday. The work that turns the refusal into a finding is to set down exactly what would have to be disclosed before anybody could. Seven rows, assembled here rather than lifted from any requirement.

1. What was decided, and when.
2. What the decision cost, priced on its own rather than folded into a larger total.
3. What happened afterwards, measured on the thing the decision was meant to move.
4. Who decided it, named against the decision.
5. What the alternatives were, and why they were set aside.
6. What the same people decided before, and how those turned out.
7. What the business would have earned had nothing been decided at all, which is the comparison every claim about management quietly assumes and almost never states.

Now land it on the case and count honestly, without rounding up out of embarrassment. For the purchase, row one is published: the start of the second year. Row two is published: Rs 21,00,000/-. Row three is published: binding still ran at 100 an hour, measured on exactly the thing the purchase was meant to move. Row four is not published. Rows five, six and seven are not published. Three of seven, and the row the reader most wants is the one that stays empty.

Here is the thing worth arguing with, and the panel above sets it out rather than asserting it. A reader who accepts that the observable set is small will still, five minutes later, ask for more disclosure as though more disclosure would fill row four. Watch what actually happens when documents arrive. With the published accounts alone the count is three. Add a full register of members, a complete and correct document in its own right, and the count is still three. The register carries entries, proportions and a statement that nothing is pledged, and it attaches no person to any entry. Add a board report of the kind a listed company produces and the count reaches four. A report of that kind shows who sat in the room and who recorded a question. Only from inside the room, on the day, does the count reach seven, and nobody reading these notes occupies that position for a business they do not run. Three, three, four, seven.

A card with four empty rows and a written reason beside each one is a finding, and a score built on the three full ones is an opinion wearing a number. The card can be handed to somebody who knows the trade. A reader of that kind disagrees with a specific row, goes and asks for row four, or explains why nobody will ever give it to them. A score cannot be argued with at that resolution, and worse, it cannot be unpicked later by a reader who inherits the file and has no idea which rows were empty when it was built.

And close where honesty requires closing. Almost nobody publishes rows four to seven. The disclosure regime does not ask for them. A business that leaves them blank has concealed precisely nothing, and this is the ordinary case rather than a warning sign.