The National Bureau of Economic Research has updated its August 2026 forecasts for two key indicators of Kazakhstan’s economy: the tenge exchange rate and inflation. Both models point in the same direction: the coming month should bring an improvement, but the tenge is expected to weaken in the autumn, while inflation will increasingly split between disinflationary food prices and stubbornly double-digit growth in non-food goods and services.
The tenge has returned to spring levels — and the August forecast has improved significantly
July brought an unexpected strengthening of the tenge. The average monthly USD/KZT exchange rate fell from 487.9 tenge per dollar in June to 471.7 in July, effectively returning to May’s level of 469.2. The average exchange rate for 2026 to date stands at 483.9 tenge per dollar.
The Bureau’s previous forecast put the average July rate at 476.7 tenge, meaning the model missed the actual figure by 5.0 tenge, or about 1.0%. By comparison, the actual rate was 12.5 tenge, or 2.7%, above the optimistic scenario of 459.2, and 22.4 tenge, or 4.5%, below the pessimistic scenario of 494.1. Once again, the actual trajectory was closest to the baseline rather than either extreme scenario, which the Bureau sees as further evidence that the model is sound.
At the time of publication, the average monthly exchange rate is estimated at 473.0 tenge per dollar.
The updated forecast for the coming months is as follows:
August: optimistic scenario — 453.1 tenge per dollar, baseline — 470.5, pessimistic — 488.0
September: optimistic — 468.5, baseline — 486.0, pessimistic — 503.4
October: optimistic — 482.3, baseline — 499.7, pessimistic — 517.2
November: optimistic — 479.1, baseline — 496.6, pessimistic — 514.0
December: optimistic — 476.7, baseline — 494.2, pessimistic — 511.7
The baseline scenario assumes that the tenge will remain close to its July level on average in August, with the model showing no immediate reversal of its recent gains. The trajectory then changes: the average monthly rate rises to 486.0 tenge in September and reaches a local peak of 499.7 in October, before strengthening slightly to 496.6 in November and 494.2 in December.
The forecast revisions vary considerably across the horizon. The previous version put the baseline August rate at 484.7 tenge, well above the new and more optimistic estimate of 470.5. The September forecast is almost unchanged. The outlook for the fourth quarter, however, has become more cautious. The previous baseline estimate for October through December was around 485-490 tenge, whereas the model now assumes a range of 494-500. In other words, the update does not simply extend July’s appreciation through the rest of the year. It combines a stronger short-term position for the tenge with continued expectations of depreciation by the autumn.
The Bureau lists global oil prices, the international strength of the dollar, export foreign-currency earnings, import payments and domestic demand for foreign currency among the main risks through the end of the year. A combination of falling Brent crude prices and a stronger dollar could push the market towards the pessimistic scenario, while favourable oil-market conditions and a steady supply of foreign currency would keep the exchange rate closer to the baseline or optimistic outcome.
The forecast is based on average monthly USD/KZT observations since January 2011, transformed into a stationary series, and represents a consensus of four independent econometric models. The average relative forecast error for 2026 was 0.3%, while the average absolute error was 1.2 tenge. In July, the deviation from the baseline scenario was 1.0%, or 4.9 tenge.
Inflation will fall into single digits for the first time in August — but not across every category
Annual inflation slowed to 10.2% in July from 10.3% in June, while monthly price growth eased from 0.8% to 0.6%. Disinflation is continuing on both an annual and monthly basis. The composition of inflation, however, remains uneven: food prices rose by 10.1% year on year and 0.4% month on month; non-food goods increased by 11.7% and 0.7%, respectively; and paid services were up 9.2% year on year and 0.7% over the month.
The Bureau’s previous July forecast projected annual inflation of 10.3% and monthly inflation of 0.6%. The actual annual reading was just 0.1 percentage point below the estimate, while the monthly figure matched the forecast exactly. At the component level, the differences were as follows: food inflation was forecast at 10.0%, compared with an actual reading of 10.1%, a deviation of 0.1 percentage point; non-food inflation was forecast at 11.5%, compared with 11.7%, a difference of 0.2 percentage point; and paid services were expected to rise by 9.6%, against an actual 9.2%, the largest deviation at 0.4 percentage point. On a monthly basis, the food and non-food forecasts matched the actual figures exactly at 0.4% and 0.7%, while service prices rose less than expected, by 0.7% rather than the forecast 0.9%.
The Bureau notes that the end of fuel-price monitoring and the removal of some restrictions on utility tariffs have yet to produce a noticeable acceleration in the headline index. The pass-through of higher fuel and utility costs into transport, logistics and service expenses may emerge with a lag, however, leaving the risk in place for the second half of the year. Monetary easing provides additional context: the base rate was cut from 18.0% to 16.75% in June, gradually changing borrowing conditions and domestic demand, although such a decision does not affect prices immediately.
The direction of the baseline forecast remains unchanged: inflation is expected to enter single-digit territory in August and stay there through the end of 2026. The trajectory itself, however, has been revised lower. The August forecast remains at 9.9%. The September estimate has been cut from 9.6% to 9.5%, October from 9.9% to 9.8%, November from 9.7% to 9.6%, and December from 9.6% to 9.4%. The revisions are even more pronounced over the longer horizon: the February 2027 forecast has been lowered from 9.0% to 8.7%, while the June 2027 estimate has also been cut from 9.0%, to 8.6%.
The month-by-month forecast through the end of the year is as follows:
August: annual inflation of 9.9%, with monthly price growth of 0.6%; food inflation of 9.9% (0.5% over the month); non-food inflation of 11.2% (0.7%); paid services inflation of 8.6% (0.7%)
September: annual inflation of 9.5% (0.7% over the month); food inflation of 9.1%, with monthly growth of only about 0.2%; non-food inflation of 10.3%; services inflation of 8.5%
October: annual inflation will temporarily rise to 9.8%, not because disinflation has ended, but because of base effects and a projected increase in monthly non-food inflation to 1.2%; annual services inflation may accelerate to 10.3%
November: annual inflation will ease to 9.6%
December: annual inflation — 9.4%, including food inflation of 8.0%, non-food inflation of 10.3% and paid services inflation of 10.4%
The most significant revision concerns the food segment. The December food inflation estimate has been cut from 9.9% to 8.0%, which the Bureau attributes to a stronger seasonal disinflationary effect. The revisions to non-food goods and services are smaller: the December estimate for non-food inflation has been lowered from 10.7% to 10.3%, while services inflation has been cut from 10.9% to 10.4%. The overall forecast has therefore become more favourable, but both categories are still expected to remain above 10% at the end of the year.
By December, the composition of inflation will have changed fundamentally. Food, previously one of the main sources of price pressure, will become the principal disinflationary force, while non-food goods and paid services will continue to post double-digit growth. The headline index falling below 10% should not therefore be interpreted automatically as the end of the inflation cycle. It is more accurate to describe it as a change in the structure of inflation: pressure is easing in food, but remains persistent in categories where accumulated costs, tariff decisions and second-round effects are more pronounced — namely non-food goods and services.
The forecast is based on data since January 2011, transformed into a stationary series, and also represents a consensus of independent econometric models. The average forecast error for 2026 was 0.2 percentage point on an annual basis and 0.02 percentage point on a monthly basis. In July, the deviation from the forecast was just 0.1 percentage point annually and 0.0 percentage point monthly.
What the two forecasts mean together
Both forecasts describe an economy moving towards a degree of normalisation, but not in a straight line. The tenge gets some breathing room in August before weakening into the autumn and reaching a local exchange-rate peak in October. Inflation falls below 10% in August for the first time since the current cycle began, but the autumn brings a temporary setback, particularly in October, driven by base effects and faster non-food price growth. These are the same categories that are sensitive to the tenge exchange rate, fuel prices and import costs. Both reports arrive at a similar conclusion: the short-term picture has improved, but the quality of that improvement is uneven. The main risk for the second half of the year remains the pass-through of exchange-rate and tariff costs into consumer prices outside the food segment.
Disclaimer: both reports are analytical in nature and do not constitute individual investment advice. The exchange rate is sensitive to news, external shocks and changes in the behaviour of market participants, while the inflation trajectory may deviate from the forecast because of supply shocks, changes in tariff policy, external price pressures and shifts in the behaviour of economic agents. Actual outcomes may therefore differ from the scenario estimates.
National Bureau of Economic Research specifically for EconomyKZ.org