ING’s David Havrlant expects Czech inflation to soften further in 2026 as food prices decelerate, with headline inflation seen near 1% in summer and core inflation easing in the second half. However, sticky services inflation keeps the Czech National Bank cautious, with only a roughly 55% probability of a single rate cut between May and August and a base rate of 3.5% possibly maintained.
"Core inflation picked up to 3% annually at the start of the year, according to our estimate. Annual dynamics in imputed rents marginally quickened to 5.1% YoY in January, mainly on the back of price gains for new properties. Prices of goods fell by 0.4% YoY, while prices of services gained 4.7% YoY. In any case, the absence of a deflationary process in the services segment will prevent the CNB from easing monetary policy, despite headline inflation crawling well below the target."
"In our base case, we still expect services and the core rate to soften in June and July. So, when there is no sign of acceleration in services but an outlook for a slowdown, the cut will come, which is our base case. However, that outcome is only marginally different from seeing no slowdown in services at all, effectively a no‑change scenario. It’s roughly a 55% chance of one cut between May and August, and 45% for staying on hold."
"So, there is already a chance of a single rate reduction in May if the Bank Board is forward-looking and our hypothesis on softening core and services inflation comes through. And yes, if the rate-setters are about to wait for evidence in the data, then more clarity on the potential core/services price deceleration would be available by the June/July inflation print. Ergo, in the matter of timing, you can throw a dice. The CNB is in a comfortable situation, not really being pushed into action by anything. Yes, June is a good middle-ground choice. But the evidence will enter the courtroom only in early August, escorted by July's inflation print."
(This article was created with the help of an Artificial Intelligence tool and reviewed by an editor.)