CEVES Observatory – Q2 2026

The Fiscal Strategy expects GDP growth to accelerate from 2.0% in 2025 to 3.0% in 2026, and then to considerably higher rates over the medium term. Such a scenario is difficult to reconcile with what current data show about the structure of growth. The estimated growth rate of 3.2% in Q1 2026 appears, at first glance, to indicate an acceleration, but in fact largely reflects the low base from Q1 2025, when growth was only 1.8% (Table 1). Seasonally adjusted data show that the economy is not accelerating, but is instead moving at a pace closer to 2–2.5% annually. Some improvement by the end of the year is possible, owing to fiscal expansion and a better agricultural season, but this still does not explain where more growth would come from. The decline in registered employment, which has been ongoing since September last year, continued into 2026. In the first quarter, employment was 0.5% lower year-on-year, while in April and May the decline slightly accelerated, to 0.54% and 0.63%, respectively.

                                Table 1:

The structure of growth further weakens the argument that the economy has entered a new cycle of acceleration. On the production side, the largest contributions come from trade (0.9 percentage points), professional services and the government sector (0.6 percentage points each), while the financial sector records strong growth of 6.9%, although its contribution remains modest at 0.2 percentage points. This combination is not impossible, but it is difficult to reconcile with the broader data on consumption, imports, investment and sentiment. If private consumption provided the main impulse to growth, one would expect to see a clearer footprint across a wider set of production sectors. One possibility is that the increase in professional services output, despite weak growth in their exports, is linked to EXPO. This could then generate spillovers into retail trade in luxury goods and services, and possibly into “intermediary” services as well. Instead, the data look more like an economy in which nominal incomes and public spending are still sustaining demand, while the real basis for future growth is narrowing. This is more important than the difference between growth of 2.5% and 3.0%, because it suggests that even the existing growth is becoming increasingly difficult to explain through a healthy expansion of productive capacity.

Investment is the weakest point in this picture. Gross fixed capital formation increased by only 1.4% in Q1 2026 and remained practically at the level recorded in Q1 2024, despite a roughly 5.1% increase in government capital expenditure excluding the Rafale payment. This means that public investment has not triggered a broader investment cycle, while private investment is clearly weakening. Foreign direct investment in January–April was 44% lower than a year earlier and around two-thirds lower than in the same period of 2024. This time, inflows into construction are almost absent, German and Chinese investment is negative, while most of the modest inflow is concentrated in manufacturing and trade. By country of origin, the Netherlands again stands out, which should be interpreted with caution: CEVES has previously noted that part of these flows probably reflects Chinese companies registered in the Netherlands, meaning that the immediate investor country does not necessarily reveal the ultimate origin of capital. Table 2 below shows FDI developments in the first quarters of the last four years by country and sector. Furthermore, new investment loans to the corporate sector paint a less clear picture: after falling by 19% in Q1, they recorded a strong and unusual increase in April and May, bringing the five-month cumulative figure 23% above the previous year. This increase, however, does not change the overall assessment, as it may reflect one or a few large arrangements rather than a broad recovery in investment demand.

                                Table 2:

Goods exports continue to grow, but their base has narrowed significantly in 2026. In 2025, goods exports excluding electricity increased by 6.4% in real terms and 8.2% in nominal terms, with growth still supported by several sectors. The largest contributions came from motor vehicles, rubber and plastics, electronics, basic metals, chemicals and machinery. However, even then, agriculture and food products were already dragging real export growth down, while mining looked much stronger in nominal than in real terms. In January–April 2026, goods exports excluding electricity increased by 5.1% in real terms and 10.7% in nominal terms, but motor vehicles alone contributed 5.3 percentage points to real growth. Without this sector, real goods exports would have been slightly negative. Nominal data for January–May confirm the same picture: exports excluding electricity increased by 9.4%, but excluding motor vehicles and mining, growth falls to only 1.8%.

Data on real goods exports show that nominal growth in 2026 partly masks weaker underlying dynamics. In the first four months, mining made a large nominal contribution, but in real terms it added only 0.2 percentage points to growth, because export unit values increased sharply. Basic metals show the opposite pattern: they grew in nominal terms, but declined in real terms and subtracted 1.2 percentage points from real goods export growth. Food, agriculture and other transport equipment also weighed on real exports. The remaining positive contributions from machinery, electronics, chemicals, rubber and plastics are important, but not large enough to change the overall picture. In the first half of 2026, Serbia’s goods exports are growing mainly because one large pillar in the automotive industry is expanding extremely rapidly, while a significant part of the rest of the export base is stagnating or declining.

Services exports remain an important stabilizer, but they are no longer growing fast enough to offset weaknesses in goods exports and investment. In 2025, services exports increased by 5.3%, with telecommunications, computer and information services contributing 2.9 percentage points, and other business services 1.6 percentage points. These two groups therefore explained most of the growth. Tourism made a smaller positive contribution, while transport was the main negative category. In January–April 2026, services export growth slowed to 3.8%. The structure also changed: professional business services became the largest source of growth, transport recovered, while the contribution of ICT services fell to only 0.6 percentage points. Without other business services, transport and ICT, services exports would have been practically stagnant.

One should also point to the latest issue of MAT (Macroeconomic Analyses and Trends), which estimates that GDP growth in the first five months of 2026 amounted to 3.6%, although without a clear description of the methodology used. Most of the analysis is devoted to manufacturing, which is reported to have grown by 1.6% year-on-year, even though the decomposition shows that most of this increase came from higher production at Stellantis, while almost half of the observed manufacturing branches recorded a decline. The same applies to exports. At the same time, MAT states that the largest contribution to overall GDP growth will not come from industry, but from “other services” with 2 percentage points, “net taxes” with 0.7 percentage points, and “wholesale and retail trade” with 0.6 percentage points. Precisely for this reason, it would be important for MAT to explain more clearly what lies behind the growth of “other services” and how such a production-side structure of growth fits with other indicators that still point to weak conjuncture — above all investment, employment, exports and economic sentiment.

                                Figure 1:

The latest economic sentiment data do not support the conclusion that economic activity in Serbia is recovering: overall sentiment is weakening again, and the decline is broad-based across components, although retail trade remains positive. After the downturn that began in Q1 2025, overall sentiment reached its lowest point in July; the partial recovery that followed still left it below the EU-27 average and in negative territory (Graph 1). It is not surprising that both Serbian and European sentiment fell sharply in February 2026, as the survey was conducted after the outbreak of the war in the Gulf. However, in May and June the two indicators diverged: while European sentiment recovered slightly, sentiment in Serbia weakened sharply again. According to the latest data, Serbia is among the lowest-ranked European countries, with only Denmark, Austria and Romania below it. Weakness across the components has largely stabilized, but construction and consumer sentiment remain in negative territory, industry is fluctuating around zero without a clear recovery, and services have lost momentum after a partial rebound. Retail trade is the only component that has remained positive throughout 2026, although it too shows signs of weakening. Overall, survey indicators do not confirm assessments of an imminent recovery in economic activity; instead, they point to weak expectations among firms and consumers and limited scope for a stronger acceleration of growth in the short term.

This is why the projected acceleration of growth increasingly rests on consumption and the state. Wages and pensions continue to support consumption, especially after the increases in the minimum wage and public-sector wages, but this channel has limits. Employment growth is no longer broadening the base of consumption as it did earlier, while part of the increase in average wages reflects a change in the structure of employment, as the lowest-paid jobs are disappearing more quickly. Public investment may temporarily support growth through EXPO, but this contribution is time-limited. Once the project is completed, sustaining high growth will require a recovery in private investment, a broader export base and productivity growth. Current data do not show that such a foundation is forming.