Borrowing Against Growth:
Bangladesh Budget FY2026-27

Node 1 · the trigger Chapter 02 / Revenue

The Arithmetic of Ambition

A revenue target is only as real as the monthly run-rate behind it. Bangladesh's headline number is ambitious; its run-rate is the story the headline hides.

The National Board of Revenue (NBR) target of $49.1bn (BDT 6,040bn) implies roughly $4.1bn a month. The recent pace is closer to $2.5bn, a gap of about 40%, on a tax-to-GDP ratio stuck near 6.8% for years. The FY26 shortfall was already past $8bn. The base is narrow and exemption-heavy: revenue forgone reached $8.8bn in FY23 alone, around 2.4% of GDP.

This is the trigger node. A miss does not stay an accounting problem. It transmits straight into bank borrowing, treasury-bill yields, payment arrears and development-spending cuts. The table shows what each scale of miss adds to the financing gap.

Node 2 · the financing structure Chapter 03 / Financing

The Only Lender Left

A country with many lenders can absorb a bad year. One with a single domestic lender cannot. Bangladesh has been narrowing toward one.

Of the $19.8bn deficit, banks supply $9.1bn (BDT 1,120bn). National savings instruments, once the retail channel, supply only $1.2bn (BDT 150bn). Net sales of savings certificates have turned negative, about −$45m over Jul-Feb FY26 after −$497m in FY25.

That leaves the state and private borrowers competing for the same damaged balance sheet. Interest already absorbs one in five taka of revenue, before a single development project is funded. The way out is to widen the base. The appetite exists: sovereign sukuk was 11.5× oversubscribed, remittances run near $35bn, and there are 80m-plus mobile-money accounts. The missing piece is distribution: instruments have to reach households, institutions and diaspora money more easily.

Node 3 · the transmission · hidden in plain sight Chapter 04 / Crowding-out

The Crowding-Out Engine

Crowding-out is a textbook idea most executives file under theory. In Bangladesh it is not theory. It is already visible in the credit numbers. This is where the budget fights itself.

A bank-financed deficit drawn from a stressed banking system leaves a visible mark on credit. Private-sector credit growth hit 4.72% in March 2026 and was 4.75% in April, down from 10.13% in mid-2024. Lending rates sit near 16-17%. When banks fund more sovereign debt, the private sector faces tighter allocation and a higher rate for working capital. Since passage, Bangladesh Bank has capped banks' weighted-average lending-deposit spread at 4 percentage points to pull borrowing costs down. But once the spread is fixed, a bank can no longer price the extra risk of a private borrower, so risk-free government securities — no provisioning, a minimal capital charge — become the easier way to earn the same capped spread. For a system that is already capital-negative and rebuilding capital while absorbing $9.1bn of new state borrowing, a measure meant to help borrowers can quietly deepen the flight into sovereign debt it was designed to counter.

The loop is not cheap. The budget earmarks about $10.4bn (BDT 1.28tn) for interest payments in FY27, more than the $9.1bn (BDT 1.12tn) it plans to borrow from the banking system. Debt service alone now costs more than the new bank financing that feeds it, so every revenue miss does more than widen the deficit: it compounds the interest bill that crowds out everything else.

The numbers are not abstract. Private investment is projected at only 21.3% of GDP in FY27, against a long-period level around 23-24%. Public investment is expected to rise from a revised 10.8% of GDP in FY26 to 13.1% in FY27. That is a big challenge when multilateral bodies forecast around 4.3-4.7% growth and FY25's final growth was just 3.49%, well below the official 6.5% target.

The remedy is to treat one sentence as a hard constraint rather than a footnote: deficit financing must not crowd out private credit. That means capping net bank borrowing, accelerating non-bank instruments, and sequencing bank recapitalisation with accountability, so the lender can lend again.

Node 4 · is the growth real Chapter 05 / Capacity

Is the Growth Real?

Reserves and remittances strengthen the national balance sheet, the way a strong cash position boosts a weak operating business. They buy time; they do not build factories.

The external buffers are genuinely strong: remittances at a record $35bn, gross reserves recovered to about $34.8bn. But buffers are not productive capacity. Development-budget execution is at a 15-year low, and public investment remains constrained by implementation. External stability leans on workers abroad and on ready-made garment (RMG) concentration, not on diversified capacity, exactly as graduation from Least Developed Country (LDC) status approaches.

If those buffers mask weak capacity, the budget's "resilience" is borrowed time rather than structural strength. The honest test is to judge foreign investment as a delivery chain: land, power, customs, VAT refunds, currency repatriation, logistics and standards. Incentives matter only if the chain works.

Node 5 · the external & political test Chapter 06 / Survival

Shock & Reform Survival

A budget is written in calm and judged in shocks. This one is also the first of a newly elected government after an interim stabilisation period, so it is judged politically as well as financially.

It was written under IMF discipline but will be tested against inflation of 9.42% (May 2026), a market-driven exchange rate near BDT 123 / USD, oil and shipping risk, and an LDC graduation timetable now conditionally moving to a United Nations Committee for Development Policy (CDP)-recommended deferral to 24 November 2029, pending UN General Assembly approval. The government's own targets, 6.5% growth and 7.5% inflation, sit well outside IMF and ADB views (4.3-4.7% growth, 8.5-9.0% inflation).

The real variable is continuity. Reform continuity sustains investor confidence; reversal pushes subsidies and revenue pressure back through the system. The roughly $1.6bn (BDT 200bn) of off-budget power-sector arrears is the warning. The government is also spending more than $3.3bn (BDT 400bn) recapitalising banks, alongside risk-based supervision and governance reform. That can buy stability only if it buys reform: protect poor households, rationalise subsidies transparently, attach accountability to recapitalisation and deepen currency hedging.

The exit · widen the base Chapter 07 / The way out

The Escape Route

The exit is not a slogan. It is a funding shift. Bangladesh has to widen who buys public debt before bank borrowing absorbs more of the credit channel.

The encouraging part is that demand already exists. The sovereign sukuk was 11.5× oversubscribed; remittances run near $35bn; there are 80m-plus mobile-money accounts; and the diaspora-bond ceiling was removed in December 2024. The weak point is not appetite. It is product supply, distribution and trust.

The FY27 budget moves in that direction through a settlement roadmap toward T+0, faster repatriation through non-resident investor accounts, a first municipal-bond framework, removal of dividend double-taxation, and a shift of withholding to advance tax. Passage confirms the budget year has begun; the real test is whether these rules become investable instruments that households, institutions and diaspora investors can actually buy.

So the real question is not whether the budget is big. It is whether Bangladesh can widen its investor base fast enough to stop financing tomorrow's growth by starving today's.

Closeout · enacted from 1 July Chapter 08 / Post-passage read

The Loop Closes

The Finance Bill was passed on 29 June 2026 and the national budget on 30 June. The budget is no longer a proposal. The circuit described here is now the execution baseline.

The passage-stage amendments gave visible relief: the tax-free individual income threshold was raised to BDT 4 lakh (about $3,250) for FY2026-27 and FY2027-28, mandatory Taxpayer Identification Number certificates for most bank-account openings were dropped, the TIN requirement for partition deeds and property mutations was scrapped, the criticised real-estate route for unaccounted money was withdrawn, and the proposed withdrawal of the zero-coupon bond tax exemption was dropped, leaving that income tax-free for individual investors. Withdrawing the unaccounted-money route is a governance win. The others reduce compliance pressure at the exact point where the budget needs a larger, more formal revenue base.

That is the closeout. Parliament changed the politics around the budget more than the financing arithmetic inside it. The $76.3bn (BDT 9,380bn) spending plan still depends on a sharp revenue lift, planned bank borrowing of $9.1bn (BDT 1,120bn), and an interest bill of $10.4bn (BDT 1,275bn). The original risk case is now the base case to monitor.