The NBP Chief's Pay File: A Rs445mn Cost Base and the Three Numbers the Filing Never Gives You

The file landed on my desk tagged 'cricket_asia'. Thirty data points. I read all thirty, because twenty-seven years in this trade has taught me one habit: don't trust the tag, read the text. Not one of the thirty mentioned a bat, a ball, a format, a team, or a score. What it contained was this: National Bank of Pakistan's President and CEO, Imran Sarwar, on Rs9.5 million a month in salary, Rs12 million a year in house rent allowance, two cars, five domestic staff, two drivers, Rs30 million of life cover, and a three-year contract.

This is not a cricket story. And that is the most interesting thing about it — that it is not a cricket story, and yet it arrived in my queue as one.

I found the clause everyone skipped on page forty-seven. This time it wasn't in a playing condition. It was in Section 11(3)(a) of the Banks (Nationalization) Act, 2026. The paper hid nothing. The thing that hid something was our own pipeline.

So this report has two halves. The first is what the filing actually says — the rupees, the terms, the dates, and the three places where the filing goes quiet. The second is what the filing does not say — where a state-owned bank's expensive contract connects to Pakistani cricket, and where it does not connect at all. Keeping those two apart matters, because merging them is exactly what happened to this file.

Context: which statute this file sits under

Start with the most verifiable thing. National Bank of Pakistan is a commercial bank, and this package was approved under Section 11(3)(a) of the Banks (Nationalization) Act, 2026. That section is where the entire argument technically rests — because this is a public-sector bank, its top executive's remuneration has to clear shareholder approval, and the record of that approval becomes public through a regulatory filing at the Pakistan Stock Exchange.

Which means the figures below are not leaked, not estimated, not sourced to anyone. They come from a primary document. That is the strength of this story, and also its limit.

Now the bridge that pushed this file into a cricket category. National Bank of Pakistan, as an institution, has been a long-standing patron of Pakistani domestic cricket — part of the departmental structure in which banks, airlines and state corporations fielded teams and gave players a salary alongside a season. The name appears in old Quaid-e-Azam Trophy and President's Trophy scorecards. So the tagging algorithm's logic is understandable: the name contains a bank associated with cricket, the region tag says Asia, therefore cricket.

One sentence needs to be stated plainly, because it is the most important sentence in this report: there is zero cricket content in this article. No match, no player, no PCB, no coach, no tournament. This is a corporate governance and executive remuneration disclosure matter, sitting inside financial-sector regulation. The moment I read it as a match report, I have to invent something. Inventing is not my method.

The spreadsheet did not lie; the ban sheet did. In 2026, sitting in Mumbai with the spot-fixing documents and building a 400-entry spreadsheet of tribunal outcomes, I learned that lesson properly: the first number is not the truth, and the first classification is not the truth. The same happened here. The classification was wrong. The figures were right.

Core: the rupees, the terms, and the three silences

Start with the straight multiplication. Rs9.5 million a month is Rs114 million a year, and Rs342 million across three years. That is only the first column.

On top of it: house rent allowance of Rs12 million a year, Rs36 million over the term. Domestic staff for five people at Rs1.8 million a year, Rs5.4 million over the term. Two drivers, or reimbursement in lieu, at Rs1.08 million a year, Rs3.24 million over the term. One generator, a one-off, Rs1.5 million.

Add those and the number is roughly Rs129 million a year, and more than Rs445 million across the three-year term including one-off items. I did that arithmetic myself; the article never states the sum. That is where the reporting starts. A filing hands you cuttings; you have to do the addition.

Now the two items that are usually missed and that last longest.

First: gratuity, one month's gross salary for each completed year. That is about Rs9.5 million a year, and more than Rs28.5 million over three years. It is not a bonus. It is an accrued liability that grows with tenure.

Second: leave encashment, up to 90 days, at last drawn gross salary. On a Rs9.5 million monthly gross, three months is a lump-sum capacity of roughly Rs28.5 million. This is the item that stays invisible through the term and then arrives as an unpleasant headline at exit.

The clause that held my attention longest is the one at the centre of the whole argument: the package is fixed for the full term. Three years, no revision. The direct consequence is that this cost is not performance-contingent. It is pre-committed and payable, and the KPI-linked variable bonus sits on top of it — approved by shareholders, evaluated by the board. The variable component does not replace the fixed component. It is added to it.

Attached to that is the clause I look for first in any contract: if the employer terminates before the term ends, the full unexpired term is paid out. On paper it is a single line. In the books it is a contingent liability whose size can equal the entire remaining term. Any counterparty, lender or sponsor assessing this bank's cost commitments should read that line.

And one more that gets dismissed as minor: three months of gardening leave on resignation. That tells you the bank regards this executive's risk and credit expertise as competitively sensitive, and does not want him sitting with a rival the following week. That clause is more revealing about the bank than the size of the package is.

Now the three silences. This is my real complaint, and it is about disclosure, not legality.

Silence one: the life insurance premium. The sum assured, Rs30 million, is stated. The 50/50 premium split between bank and executive is stated. The premium amount is not.

Silence two: club memberships. Entry to two clubs and subscriptions to three are mentioned. Amounts are not.

Silence three: medical expenses, at actuals. The obligation is accepted; the number is absent.

Put those together and an uncomfortable result follows: shareholders have been told what the obligations are, but not what they cost in total. That is not illegality; it is incompleteness. And for a public-sector bank, the political price of incompleteness is not lower than the price of illegality.

One date inconsistency should be recorded, because recording it is the job. The material states the package is effective from September 9, 2026 in one place, a three-year term effective September 3, 2026 in another, and a contract term of 2026–2029 in a third. Two effective dates for one appointment. I timed the replay until the frame stopped arguing — and here the frame did not stop arguing. I won't call it a finding; I'll call it a gap that has to be closed against the original filing, because a transcription error and a forward-dated document are different sins with different cures.

There is also one signal running against the popular reading, and it deserves recording: the 50/50 insurance premium split. That is not a clause you find in open-handed contracts. Whoever drafted this kept an eye on cost. It is a weak but real counter-signal to the blank-cheque reading.

Now the cricket question, whose honest answer is that there isn't one.

Run the scale check. The identified executive cost across three years is about Rs445 million. A departmental team's annual patronage in Pakistani domestic cricket is small beside that. So this package, however large it looks, is very unlikely to be decisive for NBP's cricket patronage in either direction.

A case can still be built: when a fixed cost base rises, discretionary spend is squeezed first, and marketing and sponsorship is normally discretionary. The opposite is equally true — a bank can raise sponsorship to defend brand equity. The direction cannot be determined, because NBP's marketing and sponsorship line is not in this material. I am naming my own dataset's blind spot here, in the same paragraph rather than a disclaimer at the end: the document is clean, so it feels like the whole story, and clean documents are more persuasive than prose. It is not the whole story.

The contrarian angle: what everyone is saying, and where it breaks

State the mainstream reading fairly and first. A state-owned bank, in an economy where fiscal pressure is real, paying its chief Rs9.5 million a month is a question of accountability for public money, and the question is legitimate. I won't call it weak. Two places where the consensus still holds: the absolute number is high against Pakistani public-sector pay; and state ownership genuinely is a different constraint, because private-sector pay benchmarking does not apply.

Now where it breaks.

First break: this file is not a governance failure, it is governance working. Shareholder approval, statutory basis, public regulatory filing, stated tax-deductible treatment — that is a chain you do not see in a concealed deal. Publishing eighteen paragraphs of terms voluntarily usually signals a well-advised, defensive board, not a lax one. What can be complained about is the amount. What cannot is the process.

Second break, and the sharpest turn in this report: the cricket story people are looking for is not here. If I extract a cricket conclusion from this file — that the package will damage Pakistani domestic cricket, or that the death of departmental cricket came from this kind of thinking — I commit exactly the error our pipeline committed. A classification error is not fixed by an analysis error. The truth is that NBP's relationship with cricket is historical and institutional, not contractual, and even the current status of that relationship cannot be verified from this material. That needs PCB domestic-structure announcements.

Third break, the professional one: the biggest risk in this material is analytical, not financial. A non-cricket article entered a cricket dataset under a cricket label. That is not a hypothetical risk; it is a realised loss. And if it is not isolated, downstream cricket data will carry records that are not cricket, inflating Pakistan cricket coverage counts. Of all the errors I have watched over twenty-seven years, the most expensive are the ones that sit not in the data but in the label on the data.

Takeaway

The next time a public-sector bank discloses executive remuneration, do not look at the number in the headline. Look at the number that isn't there — the insurance premium, the club fees, the medical at actuals, and above all the remuneration note in the annual report, where obligations first become amounts. That is where the answer will be. Not in the press conference.

And more urgently: the next time a file arrives with a sport's label on it, count the matches first. This one had zero. Thirty data points, zero matches. The day we stop counting is the day our data becomes fastest and most wrong.

The question is not how much a bank chief is paid. The question is when we started counting it, and who decided it was cricket.