Government uses infrastructure and procurement as industrial policy. Taxes pay the base, but domestic bonds, Eurobonds, development banks and bilateral lenders finance the pace — with transparency and foreign-currency risk as the price.
The largest single balance sheet
Serbia’s state is not larger than every private investor or buyer combined. It is, however, the largest single organiser of capital expenditure and purchasing. General-government investment exceeded 7.3 per cent of GDP in 2024, one of the highest rates in Europe. The 2026 budget allocated RSD740bn to capital investment, equivalent to 6.7 per cent of projected GDP. Public procurement in 2024 was worth 10.87 per cent of GDP. No private group can match that breadth of demand.
The order book runs through ministries, municipalities, road and rail companies, hospitals and the national health fund, and state-controlled groups in electricity, gas, telecoms and banking. Roads, bridges, railways, wastewater plants, hospitals, schools, defence, IT systems, medicines, vehicles and energy equipment all meet the state as buyer, sponsor or guarantor. Even when a private contractor builds the asset, public money or a public promise usually anchors the revenue.
This scale has a developmental logic. Serbia still has an infrastructure catch-up requirement; EU-related environmental and transport standards demand expensive networks; EXPO 2027 accelerates urban and national projects; and state-owned enterprises control natural-monopoly assets. Shallow municipal and project-finance markets make central government the borrower able to secure long tenors. The political logic is just as strong: visible construction supports growth, employment and a narrative of modernisation.
The state dominates because it can combine taxes, regulation, land and sovereign borrowing in a way no Serbian company can.
Who actually buys
The buying system has several centres. The central budget finances ministries and large projects. Roads of Serbia and rail entities commission transport works. EPS and electricity distributors buy generation, grid and metering equipment; Srbijagas buys and develops gas infrastructure. Health authorities procure medicines, medical devices and hospital construction. Cities buy transport, utilities and public services. Telekom Srbija and other state-controlled companies operate commercially but can also advance public infrastructure objectives.
For suppliers, the opportunity is therefore enormous but segmented. Civil contractors need bonding capacity and project references; technology vendors need security, localisation and maintenance; pharmaceutical companies need reimbursement and tender access; foreign investors may need an intergovernmental or export-credit structure. The state is often both the rule-maker and the counterparty, which can reduce demand risk but increase procedural and political risk.
Competition in purchasing remains the weak point. The European Commission reported an average of 2.5 bids per procurement in 2024 and a single-bid rate of 50.75 per cent. Contracts outside the standard procurement framework were estimated at €5.7bn, including €1.2bn under intergovernmental agreements. Special laws and strategic-partner procedures for selected projects can accelerate execution, but they also reduce comparable bidding and public scrutiny.
Taxes fund the state; creditors fund the timetable
It is misleading to assign one lender to one road. Tax, VAT, social-contribution and other revenue finance the majority of public activity, while borrowing covers the overall deficit, refinancing and the timing gap between receipts and projects. At end-June 2026 central-government debt was €41.29bn, or 43.8 per cent of GDP; general-government debt was 44.1 per cent. The ratio is moderate by European standards, giving Belgrade room to maintain high investment if growth and fiscal discipline hold.
The creditor map is diversified. Serbia had €8.67bn of domestic government securities and €12.42bn of Eurobonds outstanding. External obligations also included €2.95bn to international commercial banks, €2.83bn to China Exim Bank, €2.24bn to the World Bank’s IBRD, €2.11bn to the IMF and €1.49bn to the EIB, plus foreign-government loans and smaller exposures to the Council of Europe Development Bank, AFD, KfW and EBRD. Local banks, insurers and pension funds are natural buyers of dinar paper; international asset managers, banks, official lenders and bilateral partners supply foreign currency.
In May 2026 Serbia placed three tranches: a €1bn five-year bond at 4.25 per cent, a $1.25bn ten-year bond at 5.50 per cent and a €900mn 12-year bond at 4.875 per cent. Development finance can be cheaper or longer and usually comes with procurement, environmental and governance conditions. One example is the EIB package of more than €192mn for health and wastewater projects, combining a €157mn loan with €35.5mn of EU grants.
The model’s constraint is execution, not access to money
Debt sustainability looks manageable, but the composition deserves attention. At end-June, 79.1 per cent of central-government debt was in foreign currency and 32.3 per cent carried a variable rate. Euro-linked revenues and reserves reduce immediate risk, yet a weaker dinar, higher reference rates or cost overruns would raise the burden. Public enterprises can create contingent liabilities even when their borrowing is not initially visible in the budget.
The IMF’s June 2026 review kept a 3 per cent-of-GDP deficit ceiling and projected public debt around the mid-40s per cent of GDP. It also highlighted fiscal risks around Roads of Serbia and Belgrade and called for stronger appraisal, prioritisation and monitoring of public investment, including special-interest and government-to-government projects. The warning is not that Serbia should stop building. It is that a high-investment state needs a high-quality filter.
The corporate implication is double-edged. Public demand gives contractors, banks and suppliers a long pipeline and can crowd in factories around infrastructure. It can also crowd out private borrowers, favour incumbents with political and tender experience, and leave the economy dependent on the next budget cycle. Serbia’s state-led model will remain financeable while debt is moderate and growth continues. Its long-term return will be decided by procurement competition, project selection and whether public assets raise private productivity after the ribbon is cut.
