CEZ's Profit Jump Is a Tax Line, Not a Power Price
CEZ's Profit Jump Is a Tax Line, Not a Power Price
For three years the single most important line in a CEZ earnings deck was the one nobody in the utility business controls: the tax line. This quarter it finally snapped back, and the optics are spectacular in a way that flatters almost nothing about the underlying business.
The 24% that came from the tax office
Martin Novák did not bury it. “Our EBITDA dropped by 17% to CZK 86.1 billion. Our net income, on the other hand, has grown by 24% to CZK 26.6 billion.” — Martin Novák, Chief Financial Officer · 2026-11-12 The bridge is almost entirely fiscal: income tax of CZK 7.6 billion against nearly CZK 30 billion a year earlier, because there is no windfall tax in 2026. Operating cash flow rose 41% for the same reason. Net debt still climbed 14% to CZK 236 billion.
Watch what happened to the vocabulary. Windfall tax was CEZ's number one theme in 20243 and a fixture near the top of its keyword stack for the next two years. This quarter it has slid down the list to a near-afterthought. In its place, physically adjacent in the transcript, sit corporate tax surcharge and Income taxes — both inside the company's top ten extracted themes. That is the tell. The levy that gutted CEZ's 2023–2025 earnings has vanished from the P&L and been replaced, in Prague, by a new option on the table.
Arthur Sitbon asked directly about the government's proposal to reinstate a generation revenue cap, and whether the corporate tax surcharge follows it. Pavel Cyrani was careful and, reading between the lines, unrattled: “this is just a tool and that there was no intention announced to use it now” — Pavel Cyrani, Chief Sales and Strategy Officer · 2026-11-12. He stressed it has only cleared the government, not parliament, and pointed out that next year's retail prices are roughly 30% below the level of the 2023 cap — the crisis machinery exists, the crisis does not.
Contrast that with August. Asked then about windfall taxes returning, Novák said plainly: “I do not think it is on the table” — Martin Novak, CFO · 2026-08-11, because energy profits are far below where they were when the tax was introduced. Three months later a surcharge question is back in the top-ten keyword stack. Not because the tax exists — because the political option now does, permanently, as an amendment sitting in the statute book waiting for a crisis.
The guidance raise tells the same story. Adjusted net income was lifted from CZK 31–35 billion to CZK 34–37 billion, and the largest single driver was not operations but the disposal of CEZ's 49% stake in a Slovak nuclear project company to the Slovak state for €189 million, booking CZK 2.2 billion of net income that never touches EBITDA.
Hedged at €91 into a €150 market
Generation and Mining earned CZK 48.4 billion, down 25%, with nuclear alone accounting for CZK 11.8 billion of the decline. The average achieved power price falls from roughly €120/MWh in 2025 to €108–111 this year. That is the structural story: a baseload generator with no carbon cost on its nuclear fleet, watching its revenue per megawatt hour grind lower.
We do hedge our nuclear portfolio in a standard manner because you really never know what is the high price, right? EUR 100 seems to be peak, and suddenly within a few weeks, we -- past few weeks, we got to EUR 150 actually for next year. So we are hedging at a standard pace. We somewhat accelerated hedging of our coal exposure for 2027, 2028 when the opportunity opened up.
That is the most consequential paragraph in the call, and it cuts both ways. CEZ has an average fixed price of €91 for 2027 with about 80% of output sold by the end of September. Since then the front year has run to roughly €150. The company is therefore unusually well protected on the downside and unusually capped on the upside — only a fifth of 2027 remains exposed. Management resists the obvious temptation to sell more forward at these levels, which is either admirable discipline or a missed trade, and there is no way to know yet which.
The opportunistic trade they did make was in coal. Lignite plants ran 7% harder over nine months and are guided to 9% higher output for the year, with CCGT generation up almost 70%. The logic is a temporary arbitrage: after the Iran-driven move, gas and spot power re-rated while carbon allowances did not, opening a window for dirtier generation that was not in the original plan. Cyrani confirmed there are no volume constraints next year — coal burn will be purely a function of spot prices. But note the punchline: fossil-fuel generation EBITDA fell 40% to CZK 2.2 billion. More megawatt hours, less money. Volume is not the same as value.
The keyword at the top that management keeps pushing to the bottom
data center is CEZ's number one extracted keyword this quarter, exactly as it was a year ago. It is the single loudest thing in the company's own language. And then Petr Bartek asked about it, and the answer was a lesson in lead-time mismatch: “they basically have something like 24 months outlook and everything beyond 24 months is not really of interest. And with the SMRs, we are looking at the first SMRs only coming in mid-next decade, right, '35 or '36” — Pavel Cyrani, Chief Sales and Strategy Officer · 2026-11-12. CEZ has sites — old lignite footprints — and is offering them to developers, but the two sides are speaking different languages about time.
This is not a new caution. Twelve months ago the same management team said: “we really didn't see much real, real kind of projects in the region so far” — Martin Novák, Chief Financial Officer · 2025-11-11, and noted that European power simply is not cheap enough to make the region competitive with the US for compute. Meanwhile, the tape has been voting against the theme: across the global mover set, AI data center names show 13 advancers against 57 decliners over the past thirty days. So CEZ's highest-ranked keyword is one where the market is de-rating the cluster, the company is openly sceptical of near-term economics, and the realistic first delivery is 2035. It is optionality, not a story — and the keyword ranking overstates it.
CEZ Energy: still Q1 2027, still no details
CEZ Energy was the company's top keyword just two quarters ago. Piotr Dzieciolowski asked about the carve-out, the debt transfer, the rating target, and whether the new entity issues its own bonds. The answer is a timetable and nothing else — preparation complete by the end of Q1 2027, with the perimeter, the capital structure and the rating all still under analysis.
That is worth flagging as a pattern rather than a problem. On the August call, pressed on exactly the same questions, Cyrani said: “we are not ready to answer them as yet” — Pavel Cyrani, Head of Strategy and Sales and Strategy · 2026-08-11. The keyword has now fallen out of the top-30 stack entirely while the project behind it has not visibly progressed. Investors have stopped asking, which is not the same as the question being answered.
Also worth watching: the derivative revaluation of derivative swing that produced most of the CZK 3.9 billion trading variance and turned a near-zero proprietary result into a CZK 2.3 billion drag. Novák called it temporary and self-clearing within the year — the identical explanation he gave in August, when the swing was pointing the other way. With two consecutive quarters of derivative noise, "temporary" is doing a lot of work.
Strip it all back and CEZ is a regulated network business quietly compounding through a favourable new regulatory period — distribution EBITDA up 9%, or 17% normalised once correction factors and the South Bohemian gas acquisition are handled — attached to a generating fleet whose realised price is in structural decline. The 24% profit growth is real; it just came from the absence of a tax, not the presence of a market. The things that could change that — a 2027 parliament vote, an open 20% position, a datacentre signed in 2035 — are all still ahead.