Ten Months After the Revenue Line Was Cut: What the EDF Pays Before Its Money Reaches Government Paper

On 1 December 2026, Pakistan's Export Development Surcharge was suspended. On 6 October 2026 — if the date on the Commerce Ministry summary is correct — the government formally recorded that the Export Development Fund needed an alternative income stream. Between those two dates sit roughly ten months in which the Fund's principal revenue line was legally gone and its replacement was not yet approved.

The remedy that finally emerged is a strange transaction. Money sitting in a government account will move to a commercial bank so that it can buy the government's own securities. The state borrows back its own funds through an intermediary and pays that intermediary a spread.

I took the file apart to see which part was still warm. In 2026 I re-watched the Oval final four times and coded all 96 balls of India's top-order collapse, mapping Mohammad Amir's 6-0-16-3 for pitch, seam position and release height. That exercise left me with a private rule: no tactical claim in print without thirty coded balls behind it. This document poses a different problem. There are no balls to code — only dates and legal instruments. So I am doing what I do when a venue changes: naming the variables that did not transfer.

The Fund and the chain of decision

The EDF is a statutory fund held in the Public Account. It is non-lapsable, meaning its balance does not lapse at year-end. It has a board and, until December 2026, it had a dedicated levy for income. Once the levy went, the Fund was left with principal and whatever return that principal could earn.

The authority chain is documented. A Prime Minister's Office working group recommended the change; the Prime Minister directed it; the Finance and Revenue Division issued SRO/2335(I)/2026 suspending the surcharge with effect from 1 December 2026. The EDF Act was then amended, with Presidential assent recorded on 26 January 2026. New board members were appointed by federal notification on 17 February 2026. The board took the matter up at its 93rd and 96th meetings and asked for facilitation to place funds in fixed-income government securities or an interest-bearing National Bank of Pakistan account. The Finance Division supported the investment but attached a statutory control: because the money would leave a government account for a commercial bank, Federal Government approval under section 23(2) of the Public Finance Management Act 2026 was required. The proposal went to the Economic Coordination Committee, which approved it, paving the way for the investment.

Nowhere in this document is a single rupee figure. No corpus size, no annual surcharge collection, no revenue forgone, no expected yield, no investment tenor, no projected income. The only numerals are dates and legal identifiers.

Every information point originates with the proponent ministry or the Finance Division. There is no opposition voice, no exporter testimony, no parliamentary committee question. In 2026, running dew models from Liverpool for a Delhi Capitals side I never saw in person, I learned that three stadiums with zero crowd let the model hear what noise usually hides. Here the logic inverts: there is no noise, so the silence itself becomes data. Silence is never neutral.

First calculation: a levy is not a yield

A surcharge is a levy on a flow — export value, every year, rising and falling with trade. Investment income is a return on principal — normally a single-digit percentage annually, on a fixed base. That asymmetry is the spine of the proposal, and the document never addresses it.

To replace levy revenue with investment income, the corpus must be a multiple of the annual revenue lost. Without the corpus size the arithmetic cannot be closed, but its direction is unambiguous. The document's own language concedes this: it does not claim the Fund's capacity will be restored, only that alternative income is needed to maintain sustainability. Restoration and sustainability are different promises, and the text chose the second.

Second calculation: the round trip and its spread

Public money moves from a Treasury account into a commercial bank, which then buys government paper. The bank takes its margin; tax treatment applies. For the consolidated public sector, the net gain is below the headline yield, and may approach or fall below zero once intermediation cost is counted. Yet the EDF's ring-fenced books will show a clean gain. Part of the "new income" is a reclassification of value, not a creation of it.

When I coded 4,300 possession sequences across 64 World Cup matches in 2026, the useful question was never how much territory a side held but what that territory cost. Same grid, different ground. ECC approval is an outcome; the question is who paid for it.

Third calculation: an empty mandate grid

Only two destinations are named — fixed-income government securities, or an interest-bearing NBP account. Nothing else. No duration limits, no concentration limits, no credit rules, no prohibited-instrument list, no benchmark. Granting a public fund investment authority without publishing a mandate is a recognised governance weakness, because without a yardstick every decision is defended by discretion and challenged only by discretion.

Fourth calculation: ten months of latency

Surcharge suspended 1 December 2026; the summary seeking alternative income dated 6 October 2026. Ten months in which the Fund's designed revenue source had been legislated away and no replacement approved. The matter appears at the board's 93rd and 96th meetings; if meetings are sequential and roughly regular, the question sat across approximately three sittings. That is an inference, not a fact — the cadence is not stated, so my confidence here is moderate.

The lesson transfers cleanly: the calendar selects the XI before the selector does. Here the calendar did its work — the levy's expiry was fixed first, the substitute approved later. When the sequence inverts, the gap in between is not anyone's error; it is a design failure.

Fifth calculation: form strengthened, substance weakened

The Fund's legal position improved. The Act was amended, assent recorded, the board reconstituted, the section 23(2) route used. The lame-duck-board objection is gone. In the same window, the Fund's substance thinned. A body built to develop exports is becoming a holder of state paper, and nothing in the document requires that the income earned be deployed on export-linked activity. That is the doorway to mission drift, and no single decision will be finger-pointable.

Sixth calculation: a date I cannot yet price

The sample here is one document, so the confidence interval gets said out loud. The summary carries 6 October 2026. If it is in fact October 2026, the proposal predates the suspension it answers, and the entire latency finding collapses — leaving a proposal that existed before the problem it claims to solve. I cannot settle this without the primary document. When the rupee figures are missing and the date is contestable, both belong inside the story as open questions rather than tidied out of it.

The contrarian angle: where the law worked

The Finance Division did not block the proposal. It supported the investment and simultaneously enforced a statutory control, identifying the section 23(2) requirement before the transfer rather than after. That is gatekeeping, not defeat, and the correct channel was used rather than bypassed.

The real failure is not legal. It is disclosure and mandate. A public fund newly permitted to earn interest should face immediate questions about its investment policy, its audit line and where the Auditor-General's report lands. The document mentions none of it. And there is a genuine upside worth keeping open: non-lapsable public funds idling in the Public Account are a real Pakistani public-finance problem. If this model spreads with a published mandate, a benchmark and a stated audit line, it becomes an institutional template; the transfer alone is only an accounting patch.

What to watch

Four things. Whether a published investment mandate appears with duration, concentration and benchmark. Whether the corpus and expected yield are ever disclosed. What happens to invested funds if the surcharge is reinstated — a question the text does not touch. And confirmation of that 6 October 2026 date, on which the whole chronology rests.

The spreadsheet had the answer three weeks before the pundits did — but only when the cells were filled. Here the sheet is empty. Until the corpus and the revenue forgone are public, this argument runs on estimates rather than tactics. The last question is the one I always ask: who paid the bill — the exporter whose levy was removed, the Fund whose income was removed, or the state that borrowed its own money back through a middleman? The next summary should answer that in numbers, not dates.