"Pakistan's debt" is usually reported as a single number and a ratio. Both are close to useless on their own. A country with 70% of GDP owed to multilateral lenders at concessional rates over thirty years is in a completely different position from a country with 70% owed to bondholders at market rates over five — and the reporting rarely distinguishes them.
This piece decomposes the stock: rupee versus foreign currency, official versus commercial, and what each category can and cannot do to you.
The short answer
Public debt was about Rs80.5 trillion at end-June 2025, roughly 70.8% of GDP. Central government debt reached about Rs79.3 trillion in January 2026.
Of that, roughly 71% is domestic — borrowed in rupees, from domestic institutions. External debt is the smaller share, about 32% of public debt and falling.
And within the external portion, the striking fact: about 82% is owed to official creditors — multilateral and bilateral — with only about 7% in international bonds. The instrument that dominates the headlines is the smallest slice of the smaller half.
The first cut: rupees versus dollars
This division matters more than any other, and for a reason that is mechanical rather than political.
Domestic debt is denominated in rupees. The government can always pay it, because it can always obtain rupees. That does not make it free — the cost of paying with newly created money is inflation, and the cost of paying with borrowed money is a larger future interest bill. But a rupee obligation cannot force a default. It can only force a choice about who bears the cost.
External debt is denominated in currencies Pakistan cannot issue. Dollars must be earned through exports and remittances, or borrowed. When they are not available on the day a payment falls due, there is no monetary workaround. This is why external debt drives crises even though it is the smaller half of the stock, and why reserves — not the debt ratio — are the number that moves markets.
| Category | Approximate scale |
|---|---|
| Central government debt (Jan 2026) | Rs79.3 trillion |
| — domestic | Rs56.0 trillion (≈71%) |
| — external | Rs23.3 trillion (≈29%) |
| Public debt (end-June 2025) | Rs80.5 trillion ≈ 70.8% of GDP |
| External debt in USD (end-June 2025) | $91.8 billion, +6% y/y |
| External share of public debt | 32%, down from 34% a year earlier |
The second cut: who actually holds the external debt
This is the decomposition that most changes the picture, and it is rarely published in the coverage that drives the search traffic.
| Creditor type | Share of external debt (Sept 2025) |
|---|---|
| Multilateral (IMF, World Bank, ADB, IsDB) | ≈56% |
| Bilateral (China, Saudi Arabia, UAE and others) | ≈26% |
| Commercial banks | ≈8% |
| International bonds (Eurobonds, Sukuk) | ≈7% |
| Other, incl. Naya Pakistan Certificates | ≈2% |
Four-fifths of Pakistan's external debt is owed to official creditors. That has three consequences the headline ratio conceals.
It is cheaper. Multilateral lending carries concessional rates and long tenors that no market lender would offer a credit in Pakistan's rating band. The effective interest cost of the external stock is therefore far below what a market-priced equivalent would be.
It is more reschedulable. Bilateral debt is renegotiated between governments. Multilateral debt is rolled through programme lending. Neither process is pleasant, and both come with conditions — but they are processes, with counterparties who have reasons to keep the country solvent.
It is more conditional. The same official concentration that provides flexibility also means a large share of Pakistan's external financing is contingent on maintaining relationships and programmes. Concessional money is not unconditional money.
Bonds are the mirror image: expensive, short, and utterly inflexible — but only about 7% of the stock.
The mechanism: why the interest bill grows faster than the debt
Pakistan's debt-service burden has risen faster than its debt stock, and the reason is on the domestic side.
Domestic debt is heavily weighted toward short tenors. When a large share of the stock reprices within a year or two, a change in the policy rate passes into the budget almost immediately. The consequence is the loop described in the budget explainer: tightening policy to fight inflation raises debt service, which widens the deficit, which enlarges the stock that will reprice at the next turn.
Lengthening the maturity profile is therefore not a technical footnote. It is the single structural change that most reduces the sensitivity of the budget to monetary policy — and it is achievable without collecting an extra rupee of tax.
There has been movement in the other direction too. In the first quarter of FY2026 domestic debt fell about 2%, to roughly Rs53.4 trillion, after the government retired about Rs1 trillion owed to the State Bank. Retiring central-bank debt is significant beyond the number: borrowing from the State Bank is the most inflationary form of deficit financing available, so reducing it improves the quality of the stock, not merely its size.
The misconception: the Eurobond is not the debt problem
Every Eurobond maturity produces a wave of coverage, because the date is public, the amount is round, and the word "default" attaches to it naturally. But bonds are about 7% of external debt, which is itself about 32% of public debt — on the order of 2% of the total.
Meeting a Eurobond is a genuine test of liquidity on a particular day. It is not a test of solvency, and it is not where the burden sits. The burden is the Rs8.1 trillion domestic interest line in the federal budget, which is larger than the entire external stock's annual service cost and is driven by domestic rates on domestic debt held by domestic banks.
The corollary matters for anyone reading the question the other way round. Because so little of the external stock is commercial, Pakistan's external position is more negotiable than the ratio implies — but because so much of the domestic stock is short-tenor, its fiscal position is more rate-sensitive than the ratio implies. The composition makes the external side safer and the internal side more fragile than the headline suggests. Coverage generally gets both backwards.
What to watch
- The Debt Sustainability Analysis, published by the Ministry of Finance, which sets out the projected path rather than the current level.
- Average time to maturity of domestic debt. Rising is good, and it matters more than the stock.
- The external creditor mix. A rising commercial share means the stock is becoming less reschedulable, even if the total is flat.
- The primary balance. Debt stabilises when the primary balance exceeds the gap between the effective interest rate and nominal growth — not when the government promises restraint.
- SBP borrowing. Any return to financing the deficit from the central bank reverses the quality improvement described above.
For what the debt costs each year, see Pakistan's budget explained. For what happens at the sharp end, see what actually happens if Pakistan defaults.