Serbia’s managed float transfers volatility from the exchange-rate screen to the central bank’s balance sheet — and into the margins, inventories and timing decisions of companies.
A float inside a narrow corridor
The euro bought RSD118.47 at the end of 2017 and RSD117.282 at the end of 2025. Over eight years of inflation shocks, a pandemic, war in Europe and an energy crisis, that is remarkable stability. It is also policy. Serbia formally operates a managed float and the National Bank of Serbia says it intervenes to smooth excessive short-term volatility, safeguard price and financial stability and preserve adequate reserves.
Between 2017 and 2025 the NBS bought a net €11.28bn in the foreign-exchange market. Those purchases prevented structural inflows from pushing the dinar sharply higher and accumulated a reserve buffer. In 2025 the direction briefly reversed: the bank sold a net €580mn and the dinar weakened just 0.2 per cent, even though market demand exceeded supply by €405mn. The small price move was the intended output of a large balance-sheet operation.
The rationale is more than cosmetic. Serbia’s savings, loans, rents, capital goods and corporate contracts remain substantially linked to the euro. Exchange-rate changes pass quickly into inflation and debt service. A visibly weaker dinar can provoke household conversion and corporate pre-buying, amplifying the move. Stability is therefore a form of macroprudential policy — but it is not free price discovery.
Serbia’s currency risk has not disappeared. The NBS warehouses part of it and companies carry the rest in their margins.
The 2025 market showed who moves the rate
From January to May 2025, energy demand, uncertainty around NIS and weaker investment flows created pressure; the NBS sold about €1bn. From June to September, euro-indexed corporate lending, company FX sales and lower energy needs reversed the market, allowing the bank to buy €1.43bn. In the fourth quarter, renewed concern over NIS sanctions, energy imports, dividend payments and household demand forced it to sell about €1.01bn.
The concentration is striking. The largest energy importers bought a net €2.45bn of foreign currency in 2025. Other resident companies sold a net €1.37bn, 65 per cent less than in 2024. Non-residents bought €1.27bn, their highest net demand since 2017. In late November and December households alone bought more than €660mn net. These flows are the Serbian FX market: energy invoices, export receipts, group dividends, loan disbursements, portfolio exits and household confidence.
Calling this a set of ‘games’ can imply manipulation where ordinary treasury behaviour is enough. Exporters choose when to convert euro revenue and may delay if the dinar looks vulnerable or if they have imported inputs to pay. Importers accelerate purchases before energy or inventory bills. Banks price liquidity and match client orders. Non-residents hedge dinar bonds. The central bank observes the aggregate and leans against a one-way move.
The quiet winners and the squeezed exporters
A stable nominal rate helps importers, euro borrowers, retailers and households plan. It anchors the dinar cost of machinery, fuel and consumer goods and reduces the chance that an FX-linked loan suddenly becomes unserviceable. The government also benefits because most public debt is in foreign currency. Price stability and fiscal stability reinforce each other.
Exporters receive certainty but not an automatic advantage. When Serbian wages and local prices rise faster than the dinar moves, their euro-denominated labour and operating costs increase. A foreign-owned factory importing much of its input base may be naturally hedged; a domestic software, furniture or food exporter with dinar salaries and local materials feels a real appreciation more directly. The response must be productivity, pricing or higher-value products, not an assumption that devaluation will restore competitiveness.
Formal hedging remains unusually small. In the first quarter of 2026 residents’ forward FX purchases were only 0.6 per cent of their total currency purchases. New corporate lending, especially investment lending, remained predominantly FX or FX-indexed even as the dinarisation of outstanding corporate and household receivables reached a record 39.7 per cent. Most companies still hedge with deposits, invoice timing, matching inflows and outflows, or simply trust the NBS corridor.
The stability bargain can hold — until flows align
By July 2026 gross reserves stood near €30.5bn and net reserves near €26bn, while inflation was 2.7 per cent inside the 3 per cent plus-or-minus 1.5 point target band. That is a substantial defence. Since August 2012, the NBS says the dinar appreciated about 1 per cent while the bank bought a net €9.9bn. Credibility is now one of the intervention regime’s strongest assets.
The risk is a shock that turns several groups into buyers at once: weaker foreign direct investment, large fiscal imports, a renewed energy or NIS disruption, portfolio outflow and household conversion. Reserves can absorb a temporary imbalance; defending a misaligned level indefinitely would tighten dinar liquidity and eventually become costly. The NBS therefore needs discretion rather than a public peg promise.
Corporate treasurers should treat RSD117 as a policy range, not a law of nature. Exporters should budget real cost appreciation and convert according to cash needs rather than speculation. Importers should stagger purchases and use forwards for large committed bills. The central bank’s achievement is genuine, but the calm screen is the end of the process, not evidence that no market forces are operating behind it.
