In spite of the YoY contraction in GDP, demand appears to have been brisk during the quarter, with domestic demand – largely a reflection of consumption and investment – up 3.9%. Weighing more heavily, however, was a strongly negative contribution from net trade to output growth, which SI estimates to have lowered GDP by 5.8 percentage points during the quarter.

Strong import growth a salient factor in H1 …

As we have discussed recently, combined goods and services trade generated a sizeable deficit in H1. But this deficit stems largely from a massive increase in investment relating to data centres. According to the newly issued Monetary Bulletin, if these imports are set aside, this year’s trade balance will probably be close to neutral. Furthermore, investment goods imports grew 72% YoY at constant exchange rates in Q2, after a first-quarter growth rate of 84%. If the data centre projects are set aside, however, investment goods imports contracted marginally YoY in Q2.

Total imports grew 13.4% in Q2, roughly the same as in Q1. This seems fairly strong relative to growth in demand – particularly in investment – and as we discuss more fully below, it probably has an outsized effect on the national accounts.

Goods exports had a difficult quarter, contracting by 4.5% YoY, according to SI’s accounts. This is probably due in part to the effects of the Trump trade war with the rest of the world. It is pointed out in Monetary Bulletin, however, that SI’s distribution of pharmaceuticals company Alvotech’s export revenues between goods and services, which differed from that assumed by the CBI, could also have been a factor. Services exports grew more strongly than goods exports during the quarter, growing by 4.2% YoY.

… but does not appear to fully reflect investment growth

Given the above-mentioned boom in investment goods imports, Q2 investment growth figures seem rather modest. Total investment grew by 8.3% in real terms during the quarter, led by a nearly 14% real jump in business investment. At the same time, public investment grew by more than 6%, while residential investment shrank nearly 9%, after shooting up 20% in Q1.

Because SI’s investment figures fluctuate widely from quarter to quarter, it is more illustrative to examine longer periods of time. In H1/2025, total investment grew by 13%, driven by just over 16% growth in business investment, while residential investment was up by slightly more than 5% and public investment nearly 8%.

Although these figures represent handsome growth, they do not appear to capture in full the surge in investment goods imports during the half. The discrepancy in Q2 data is particularly stark.

In this context, it is useful to recall an article written by the Deputy Governor for Monetary Policy and released in the CBI’s online publication Kalkofninn about three years or so ago. The article, entitled “Domestic inventories – the missing link in Iceland’s national accounts?”, (in Icelandic only) focuses on a shortcoming in SI’s expenditure accounts. More specifically, the item called inventory changes, which is supposed to reflect goods that have been imported or manufactured but have not been sold on to domestic or foreign buyers, only captures a portion of inventories. What it does capture is inventories of imported oil products, on the one hand, and export companies’ inventories, on the other; i.e., it mostly represents unsold export goods manufactured by companies in the energy-intensive and fishing industries.

The result of this, to give an example, is that if a portion of imported investment and consumer goods are still unsold at the end of the quarter, they are (correctly) measured in full in developments in imports for that quarter; however, they are not recognised as inventories, private consumption, or investment for that period. This discrepancy rights itself during the following quarter, when the imported goods in question are used for investment or consumption.

We think it likely, given the discrepancy between import growth and domestic demand in Q2, that this shortcoming was unusually prominent in the national accounts for the quarter. If so, growth in demand could turn out far stronger in Q3 than can be explained by the quarter’s increase in imports of goods for investment and consumption.

Household consumption still robust

Private consumption has been brisk year-to-date, although fortunately, it has kept pace with real wage gains and population growth. This means Icelandic households are not taking on debt to finance their consumption, as they still have a decent-sized stock of accumulated savings.

According to SI’s press release, there was a general increase in domestic consumption, partly because of an uptick in households’ purchases of service-related consumer goods. In addition, new motor vehicle purchases by households rose for the second quarter running, after contracting in the recent term, and Icelanders’ travel-related spending abroad grew markedly.

But there was an unusual lull in public consumption, which reflects consumption relating to functions such as education and healthcare, items that are paid for with public funds. In Q2, public consumption grew in real terms by only 0.3% YoY, its slowest growth rate in nine years. It generally fluctuates less than other key national accounts items, and presumably it will resume its usual pace in the quarters ahead – unless SI revises its preliminary figures to align more closely with recent quarters.

No substantial economic contraction in Iceland

So if we take into consideration the reservations described above, what can we conclude from the newly published national accounts figures, and what is the outlook for 2025 as a whole?

As we have noted, we believe that temporary factors and data problems complicate the interpretation of preliminary Q2 figures. A contraction of nearly 2% for the quarter is therefore not a sign that the economy is suddenly cooling at a rapid pace, as other indicators suggest that domestic demand is still chugging along at a good clip.

GDP growth measured 2.7% in Q1, and in real terms it measured 0.3% YoY in H1. This sluggish growth comes on the heels of a 1.0% contraction in 2024, according to revised SI data. If we look at the first half of the 2020s as a whole, we can see that in real terms, GDP was 11.3% higher in 2024 than in 2019. This translates to an average growth rate of 2.2% per year, although of course, the period in question included a sharp contraction followed by a sizeable boom. In historical context, such a growth pace is relatively moderate, particularly in view of the surge in population thus far in the 2020s.

If our suspicions are correct and what we are seeing is a temporary effect caused by the above-described shortcomings in SI’s expenditure accounts, GDP growth could turn out that much stronger in H2/2025. The deduction for imported investment and consumption goods will then be exceeded by the corresponding imported consumption and investment. In addition to this, the peak tourist season exceeded expectations, although the winter ahead is still an unknown.

As we see it, then, the outlook is for solid GDP growth in H2, with growth for 2025 as a whole perhaps in the neighbourhood of the scant 2% we forecast in May. It is also worth noting that in its August forecast, the CBI projected year-2025 output growth at 2.3%.

In our opinion, the big picture is that the CBI’s tight interest rate policy has begun to dampen demand growth, and headwinds in goods exports have affected the growth rate there. On the other hand, the domestic economy remains reasonably strong, and as yet there are few signs of an imminent hard landing with the attendant difficulties in company operations, widespread unemployment, and significant financial distress among households.

The situation is probably somewhat to the CBI’s liking, however, at a time when the short-term inflation outlook has brightened a bit after September’s favourable CPI measurement. We hope that sooner rather than later, the CBI will see cause to begin unwinding interest rates again after last month’s pause. If developments pan out as they look set to do, rate cuts could start again early in 2026.