Saudi Arabia Is in Recession, and the Books Say Something Worse
Saudi Arabia is in recession. The IMF's growth forecast survived 24 hours before Riyadh's own data destroyed it, and the fiscal books say something worse.
The IMF published its annual assessment of the Saudi economy on 29 July, projecting growth of 1.7% for 2026 and describing early stabilisation in non-oil activity between April and June. Its validity lasted 24 hours.
On 30 July, the General Authority for Statistics released flash estimates for the second quarter. Real GDP contracted 4.8% year on year. Oil activities fell 24.7%. Seasonally adjusted, output fell 1.2% in the first quarter and 4.9% in the second, which is two consecutive quarters of contraction on the only definition that matters. Non-oil activity, the sector described as stabilising, shrank 0.5% in the quarter being described. The Ministry of Finance published its half-year budget performance report in the same window, and that is the document nobody read.
Getting to 1.7% for the year now requires roughly 4.3% growth in each of the remaining two quarters, in an economy that cannot move its principal export at volume. Population is expanding at over 2%, so real output per resident fell close to 7% year on year in the second quarter. Non-oil growth has decelerated for six consecutive quarters, from 5.4% to 5.4%, then 4.9%, 4.6%, 2.9% and 0.6%. In the seasonally adjusted quarter the only expanding line in the entire economy was government activity, contributing 0.03 of a percentage point. Year on year, government grew faster than the non-oil private sector, 0.9% against 0.6%. Strip out the state and there is nothing holding the private economy upright.
The ledger nobody read
The fiscal report is where the argument settles, because it records what happened rather than what was projected. Revenue in the first half came to 599,756m riyals, up 6.1% year on year. Spending came to 759,757m, up 15.4%. The deficit for six months is 160,002m riyals, around 42.7bn dollars, which is 96.7% of the entire deficit budgeted for the full year. Half the year has consumed almost all of the annual allowance, and the second half of 2025 ran 11% heavier on spending than the first.
The Fund projects full-year expenditure falling to 27.2% of GDP from 29.1%, and a deficit of 3.7%. Neither survives contact with the ledger. First-half spending alone is close to 16% of annual GDP and rising against an economy that is shrinking. The half-year deficit is already 3.35% of GDP. Public debt was projected to end the year at 32.1% of GDP. It stood at 1,684,993m riyals on 30 June, around 449bn dollars, or 35.3%. The projection was breached in June, with six months still to run.
There is no consolidation happening. Capital spending rose 32.4%. Military spending rose 12% and has already absorbed 52% of its annual allocation. The general items line, which carries debt cost and the emergency allocation, rose 43.8%. Education rose 5.4%, the slowest line in the budget. The state is spending more, earning marginally more, and borrowing the entire difference.
That last point is stated explicitly in the financing table. Two sources are listed. From government reserves, zero. From debt, the full 160,002m riyals. The general reserve closes the half year at 399,070m riyals, roughly 106bn dollars, untouched, while the state paid 29,099m riyals in interest over the same six months, a bill 19.6% higher than a year earlier. The reserves are not being used because using them is legible to markets in a way that issuing sukuk is not. Riyadh is buying silence with interest payments.
The revenue side removes the last defence. Non-oil revenue grew 2.4% in nominal terms, which against reported inflation of 2.2% means real non-oil revenue is flat at best. Taxes on income, profits and capital gains rose 2.4%. Other revenue fell 2.5%. Taxes on international trade fell 13%, which is Hormuz appearing directly in the receipts. Oil revenue rose 9.4% on the war premium and carried the entire increase. The central recommendation put to Riyadh is non-oil revenue mobilisation. The period under assessment contains none.
Two further numbers explain the rest. Subsidies rose 117.2% and grants rose 218.9%. The 2.2% inflation figure is not a fabrication, it is a purchase, costing over 8bn dollars in six months in price caps on fuel and food. And when oil revenue and interest are stripped out entirely, the state runs a non-oil primary deficit worth 23.3% of non-oil output. A decade into Vision 2030, the non-oil economy requires an injection of rent and borrowing equal to almost a quarter of everything it produces.
Debt as the last growth engine
Saudi Arabia raised 49.34bn dollars across 58 bond and sukuk issuances in the first half of 2026, close to half of all Gulf primary debt issuance. Outstanding Saudi debt securities passed 520bn dollars in 2025, up 21% in a year, and are projected to reach 600bn by the end of 2026, making the Kingdom the largest dollar debt and sukuk issuer in emerging markets. The 2026 borrowing plan was set at 217bn riyals, covering a 165bn deficit and 52bn of maturities. It was declared roughly 90% complete in May, before the fiscal position deteriorated further. In July the debt office redeemed 17.1bn riyals of domestic sukuk early and simultaneously issued 17.2bn in replacements running to 2041, which is maturity extension rather than repayment.
Beneath the sovereign, the same machinery runs through the corporate sector. On 27 July the national electricity provider signed a 15.8bn riyal murabaha facility, around 4.1bn dollars over seven years, arranged by seven domestic banks and unsecured. It follows a 1bn dollar syndication and a facility of up to 1.5bn earlier in the cycle. The proceeds are described as general corporate purposes, which is the standard formula for balance sheet support rather than project finance.
This is where the credit numbers acquire meaning. Private credit growth is projected to halve from 10.2% to 5.8%. That is not a demand story. Domestic deposits are being rotated out of private lending and into state-linked balance sheets, because the sovereign and its utilities are the borrowers carrying the guarantees. The private sector that was supposed to carry diversification is being crowded out by the entities diversification was supposed to replace. Direct investment has fallen from 2.6% of GDP to 1.5%, and even that is inflated by intra-group transfers and by infrastructure lease structures counted as inbound capital.
None of the corporate borrowing appears in the 35.3% debt figure, and neither does sovereign fund leverage, the national oil company’s own paper, mortgage securitisation, or the sale and leaseback contracts through which pipelines, desalination plants and power capacity have been monetised. Those structures exist so the cash arrives and the liability never shows on the sovereign line.
The riyal already has two prices
The peg holds at 3.75 for everyone with legal recourse. It does not hold for everyone else. Saudi Arabia funds roughly 70,000 students across 30 countries on a programme costing over 9bn riyals a year. Stipends are written in riyals, 6,512.5 a month for a single scholar with 5,606.25 for a spouse and 812.5 per child, and paid in the currency of the host country. Between the riyal obligation and the payment sits a conversion rate set by each cultural bureau. In July the Canadian rate moved from 2.57 riyals per Canadian dollar to 2.81747, against a market rate of 2.66. The same entitlement that produced 2,534 Canadian dollars in June produced 2,311 in July. Students across multiple countries reported deductions of at least 15% in the same month. In the eurozone, transfers fell from around 1,900 euros to 1,600, a cut of 16%, while the euro moved 1.72% against the dollar over 12 months. At 15%, the implied riyal to dollar rate is 4.41.
Under Article VIII of the Fund’s own articles, a spread of more than 2% between official rates constitutes a multiple currency practice. The assessment released on 29 July does not mention it. A state that operates one rate for its bondholders and another for its riyal obligations abroad is running a plural exchange regime. Egypt looked like this for years before it floated, and the tell was always the same, the emergence of purpose-specific rates for constituencies that cannot complain. Riyadh is not short of dollars, so this is fiscal adjustment rather than external collapse. But the direction is unambiguous. Adjustment is being pushed down into administrative tables and allowance schedules that no market prices and no auditor examines, precisely so the headline number never moves. The general education law issued by royal decree on 24 July, opening school management to private and non-profit operators, belongs to the same programme. Both are retrenchment executed on populations with no capacity to resist.
The forecast for 2027 is 5.5%. It assumes shipping normalises, oil holds, and capital returns on schedule. Even if all three arrive, the arithmetic underneath does not change. Closing a non-oil primary deficit of 23.3% requires cutting the spending that generates the non-oil growth, and every point of closure comes out of the demand the private sector runs on. Consolidation and diversification are in direct conflict, and no Gulf state has resolved that conflict without a rent windfall doing the work for it.
What the three documents describe between them is a state financing its own contraction on borrowed money, suppressing the price index with a subsidy bill it cannot sustain, crowding its banks out of private lending, and devaluing quietly against the people least able to object. The recession is the visible part. The model failing to pay for itself is the part that will still be there in 2028.