Kazakhstan Forecasts Versus Reality

The 2026 war became a rare test for economic forecasts for Kazakhstan. Usually, forecasts can only be checked after the fact, through annual statistics. Here, the situation unfolded differently. By midyear, it was already clear which estimates described the shock better and which remained too general.

But it is not only GDP figures that need to be compared. The Kazakh case is more complex. The external shock tested several forecasting channels at once: oil production, export routes, the budget, the National Fund, inflation, the exchange rate, sectors, and the economy’s ability to compensate for the oil shortfall through the non-oil sectors.

If one looks only at GDP, the picture is incomplete. If one looks only at the oil price, it becomes even more misleading. In 2026, Kazakhstan benefited from high oil prices, but at the same time faced declining production, restrictions on the CPC, an accident at Tengiz, and costly alternative routes. So the key question was not: “How much is Brent?” Far more important was this: “How much oil can Kazakhstan actually produce, export, and turn into revenue?”

In one respect, the forecasters were almost entirely right. The main threat to Kazakhstan was not in Hormuz. Kazakh oil does not pass through Hormuz, so the war in the Gulf did not create a direct blockade of exports. The main vulnerability lay elsewhere — in the route through the Black Sea and in the operation of the largest fields.

Various observers pointed this out. International institutions spoke of dependence on the CPC and the oil sector. S&P had already built in slower growth due to lower production, a weaker fiscal impulse, and heavy dependence on pipeline infrastructure. Halyk Finance expected an economic cooling and a weaker oil contribution. AERC revised its forecasts downward, factoring in oil production, investment, and the external sector. Industry experts, including Olzhas Baidildinov, directly linked Kazakhstan’s losses to attacks on the CPC, export restrictions, and fiscal consequences.

In this sense, the forecasting consensus was strong. Almost everyone saw that Kazakhstan would not automatically benefit from high oil prices. The price gain could be eaten up by lower production, logistical disruptions, OPEC+ quotas, and the budget mechanism.

But beyond that, differences began to emerge. It is one thing to identify the CPC as the main risk. It is another to assess the depth of the decline in production, the duration of repairs, the speed of redirecting volumes, the impact on the National Fund, and the ability of non-oil sectors to offset the blow.

The most pronounced divergence was in GDP growth. By mid-2026, forecasts for Kazakhstan ranged widely: from cautious 4.1–4.6% to an optimistic 5.5%. Actual growth for January–May came in at 3.7%.

A direct comparison between the five-month actual and annual forecasts is not possible: the second half of the year may accelerate if production recovers and the CPC operates more steadily. But as an early stress indicator, 3.7% matters. It shows that the oil slump in the first half of the year was deeper than most annual estimates had assumed.

It is especially important to distinguish between types of forecasts. The government’s 5.4% forecast was the baseline budget scenario before the shock, not a new post-war estimate. The National Bank’s 4.5–5.5% forecast in June was already a reactive forecast based on part of the actual data. The National Bank’s consensus survey and the estimates of international institutions lay between these two poles. S&P and some local analysts were closer to the cautious end.

Thus, everyone saw the slowdown, and almost everyone overestimated the resilience of the first half of the year. The more accurately a forecast accounted for physical production, the CPC, and Tengiz, the closer it was to the actual trajectory.

The Kazakh shock is best explained not by GDP, but by oil. Before the crisis, the baseline macro framework assumed economic growth against the backdrop of relatively stable production and manageable export flows. But by spring 2026 it had already become clear: the key variable was not Brent, but physical volume.

The Ministry of Energy lowered its forecast for oil and gas condensate production in 2026 after incidents on the CPC and the accident at Tengiz. In the first quarter, Kazakhstan produced 19.7 million tons of oil and gas condensate, which was about 80% of the level a year earlier. Exports over the same period amounted to 15.3 million tons, or 79% of the level a year earlier.

This changes the entire logic of the forecast. Prices could rise, but the country could not fully capitalize on that increase if production and shipments had declined. Kazakhstan found itself in a situation where the global market was paying more, but export volumes were lower.

Therefore, the more accurate forecasts were not those that simply looked at Brent, but those that asked the production question: what is happening with the CPC, Tengiz, Kashagan, Karachaganak, export logistics, and alternative routes?

In this sense, industry experts were more useful than conventional macro models. Baidildinov and other oil analysts had pointed out in advance that disruptions at the CPC would cause not only logistical but also financial damage: lost export revenue, pressure on KMG, the budget, the National Fund, dividends, and the tenge. Not all of their foreign exchange estimates had been confirmed by mid-June, but they identified the cascade of route-and-fiscal effects more accurately than many.

Another weak spot in the forecasts was the fiscal channel. In public debate, expensive oil is often automatically translated into “more money for the budget.” The Kazakhstan case showed that this link is more complex.

First, oil revenues are distributed through the tax system, export duties, subsoil user payments, the National Fund, and transfers. The Brent market price is only an input variable. It is not equal to direct budget revenue.

Second, part of the oil margin remains with the operators of major fields and participants in the export chain. The state does not receive revenue from the full price of a barrel, but from its share of taxes, payments and dividends.

Third, the National Fund simultaneously receives oil revenues and finances budget transfers. In the first quarter of 2026, receipts from oil companies increased, but withdrawals for transfers were comparable or higher. This is the transfer trap: a high oil price softens the blow, but does not automatically make the budget sustainable.

The fiscal forecast would have been stronger if it had accounted not only for price and GDP, but also for three linkages: production → oil-related payments → the National Fund → budget transfers. Without this, expensive oil looks like a windfall, although in practice it may only partially offset losses from lower volumes.

Inflation forecasts were more accurate. Most estimates converged around a double-digit, but gradually declining, level. By May, inflation had slowed to 10.4%. This confirmed that inflationary pressure remained high, but the peak had already passed.

In June, the National Bank lowered the base rate to 17%, explaining the decision by slower inflation, a stronger tenge, and an improved forecast trajectory. At the same time, monetary policy remained tight: inflation was still above the target, and inflation expectations remained elevated.

Forecasts for the tenge exchange rate were less accurate. Some analysts expected a sharper weakening, to as much as 600–610 tenge per dollar by the end of 2026. But by midyear, the tenge proved stronger than pessimistic expectations. High oil prices, foreign-exchange sales from the National Fund, a tight policy rate, and foreign-currency flows supported the exchange rate more strongly than stress scenarios had assumed.

Here it is important to distinguish the time horizon. By mid-June, the pessimistic currency scenario had not materialized. But that does not mean it was definitively wrong. If oil prices retreat, the CPC again faces constraints, and budgetary needs persist, pressure on the exchange rate may return in the second half of the year.

The most important blind spot in Kazakhstan’s forecasts is routes. Many forecasts spoke about dependence on the CPC, but rarely explained what diversification means in practice.

On the map, Kazakhstan has several routes: the CPC, Atyrau–Samara, Baku–Tbilisi–Ceyhan, China, the Caspian, the Middle Corridor, and southern routes. But in a crisis, what matters is not the map, but available capacity.

An alternative route may exist, but it may be more expensive, slower, and more constrained in terms of oil quality, tanker logistics, port capacity, or political approvals. Therefore, the simple phrase “Kazakhstan will redirect its oil” is insufficient in a forecast. It is necessary to calculate how many tons can be redirected, over what period, at what discount, through which ports, under what insurance coverage, and with what impact on the budget.

Here, sectoral forecasts are once again more useful. They more often capture the physical realities of the route: the port, tanker, pipeline, oil quality, loading schedule, contract, and insurance risk. A macro model is more likely to see the final GDP figure, but it does a poorer job of showing the path through which the shock reaches it.

The second blind spot is the internal structure of the economy. Almost all forecasts discussed GDP, inflation and the exchange rate. But far less attention was paid to how the shock is transmitted within the country.

The oil sector declined, but transport, construction, services, and manufacturing partially offset the drop. This is an important signal: Kazakhstan’s economy proved to be more than one-dimensional. The blow to production was severe, but overall GDP did not collapse because other sectors continued to grow.

However, such compensation does not eliminate the cascading effects. In oilfield services, downtime and the risk of workforce reductions could emerge. In aviation and logistics, fuel and insurance costs were rising. In regions tied to oil production, the fiscal and employment impact could be stronger than the national average. In transport policy, the importance of the Caspian, the BTC pipeline, the Middle Corridor, and the Chinese route increased. In energy strategy, the case for uranium and the nuclear agenda gained strength.

These effects were hardly integrated in advance into a single forecast framework. Forecasts captured the aggregate picture, but were less able to capture the cascades.

If we assess not a single GDP figure, but the entire structure of the shock, the picture looks as follows.

S&P Global, Halyk Finance, and AERC were closer to the cautious scenario. They better captured the slowdown, the weakness of the oil contribution, high interest rates, fiscal constraints, and external sector risks.

International institutions (the IMF, the EBRD, the World Bank, and the ADB) clearly identified the CPC as a key risk and issued a moderate growth forecast. Their weakness is that the analysis often remained at the aggregate level: GDP, inflation, the current account, and the budget.

The National Bank was more accurate than others in its work on inflation and the monetary policy path. But its June forecast can no longer be considered an early prediction. It was a response after part of the shock had already become visible in the statistics.

The EDB remained on the optimistic edge of the forecast range. Its scenario may be partially confirmed only if production recovers quickly and the second half of the year is strong.

Industry experts were better than most at identifying the route-and-fiscal mechanism: the CPC, export losses, the cost of alternatives, the National Fund of Kazakhstan, KMG, dividends, and the tenge. But some of the foreign-exchange stress scenarios had not materialized by mid-June: the exchange rate proved more resilient than pessimists had expected.

The main takeaway: it was not those who named one correct figure who were more accurate. More accurate were those who saw several interconnected contours — production, route, price, budget, currency, and sectors.

The quality of forecasting for Kazakhstan in 2026 hinged on one simple but difficult variable: the physical operation of oil infrastructure. Everyone understood that the CPC was important. Not everyone was able to assess how long disruptions at the CPC and Tengiz would weigh on production, exports, GDP, and the budget.

For the second half of the year, the key question remains the same: will production recover quickly enough for full-year growth to reach 4.5% or higher? If growth remains below 4% based on the results of the first nine months, that will confirm the cautious forecasts. If the economy reaches 4.5%, it will mean the initial shock was deep but temporary. If new disruptions at the CPC or at the fields recur, the forecasts will have to be revised again.

For users of forecasts, the practical takeaway is this: for Kazakhstan, one cannot rely solely on the average GDP forecast. A map of conditions is needed.

• What is happening with the CPC?
• How quickly is Tengiz recovering?
• How much oil is actually being transported through alternative routes?
• What share of high-priced oil turns into state revenue?
• What transfers come from the National Fund?
• Where is the tenge supported by underlying fundamentals, and where is it supported administratively and through the budget?
• Which sectors are offsetting the oil shortfall, and which are bearing hidden costs?

The 2026 war became a test not only for Kazakhstan’s economy, but also for the ways it is forecast. It showed that standard macro models see aggregates, but capture routes and cascades less well. And for a landlocked resource-rich country, resilience is determined not only by the oil price, but by the ability to extract, transport, and monetize the resource amid several wars around export corridors.

The materials reflect interim results as of June 17, 2026. The war is not over; the study continues.

This article is part of TALAP’s series on the 2026 oil shock and its consequences.

1. Part 1
2. Part 2
3. Part 3