Today, Ukraine’s electricity market is alive. It functions, fluctuates, delivers results, and trades megawatts. But to be honest, it operates largely “on intuition.”
There is no electricity price forecast as a systemic tool. What we have instead is something you might call: “Guess right – good job. Guess wrong – expect losses big enough to knock you out of the market for half a year… or entirely.”
And we’re not talking about minor fluctuations — we’re talking about substantial costs paid by businesses, municipalities, hospitals, schools. In other words, by all of us.
Why has this happened?
It’s convenient to blame the war: limited data, high risks, unstable conditions. And that’s partly true. But as an industry association, we must highlight a deeper problem: the absence of a system-wide forecast allows each market player to create their own “version of the truth.” Sometimes — in their own interest. Without a transparent forecasting mechanism, we no longer have a civilized market, but something closer to an energy casino.
We’re not here to sow distrust — but we won’t sugarcoat it either. Today, each supplier, trader, or producer tries to forecast on their own — based on what data and methods they have access to.
And when it comes to artificial intelligence — there’s a new boom. Some still can’t tell a chatbot from a real analytical model but are already promoting urgent AI implementation.
Yes, some companies are genuinely trying. But AI without reliable input data isn’t analytics — it’s just well-packaged distortion, creating an illusion of precision. And everyone working in the market knows this well.
Maybe it’s time to approach the issue from a public policy perspective — not through manual regulation, but with a European, business-oriented approach: where price is the output of a model, not a result of survival tactics in the dark.
That’s why we want to shift from criticism to practical suggestions:
– What data is used for price forecasting?
– Who generates and maintains that data, and ensures its quality?
– Who is actually responsible for forecasting — since serious models aren’t built “on a napkin”?
How is this done in European countries? And why has the time come to create a proper platform in Ukraine?
We are talking about building a systemic, transparent, and independent forecasting model — one that serves the entire market, not just select players. A model based on trust, not assumptions.
What data is required for forecasting? It’s not just 7 factors — it’s a full energy data ecosystem
When people talk about electricity price forecasts, many imagine a simple list: demand, generation volumes, fuel prices, CO₂ emissions (which may apply in Ukraine from 2026 — though likely to be postponed), generation outages, and weather forecasts.
But that’s just the tip of the iceberg. True energy forecasting involves a complex, multi-level data system — where each variable interacts with dozens of others. Even a small error in one input can distort the entire outcome.
We’re not talking about a table of numbers — but a fully integrated data infrastructure. Without this, high-quality, reliable forecasting is impossible.
An example of a systematic approach: Poland’s ARE SA
Since 2023, the Polish company ARE SA has been publishing long-term electricity price forecasts quarterly, through 2040. Their analytical ecosystem includes:
1. Methodology: not just one model, but interconnected modules
Models like STEAM-PL, MESSAGE-PL, ARENA-POL, and ORCED-PL form a full chain from demand to market — factoring in climate policy, changing generation structures, new technologies, and cross-border trade.
2. Assumptions database — the “energy foundation”
• Macroeconomic and demographic projections
• Fuel prices (gas, coal, biomass)
• CO₂ emission allowance prices
• Planned commissioning/decommissioning of generation units
• Scheduled cross-border interconnectors
• Legislative initiatives (including those not yet adopted)
3. Demand forecast — not just “3% annual growth”
Using the STEAM-PL model, they account for:
• Changing lifestyles (EVs, air conditioning)
• Electrification of industry and heating
• Energy efficiency measures
• Consumer response to price fluctuations
4. Generation mix forecast — detailed by source
Coal, gas, nuclear, renewables, CHP, hydro, biomass, DSR, storage.
5. Final block — wholesale price forecast
This is the final product of a sophisticated data dance. ARE uses mathematical models (ARENA-POL and ORCED-PL) to forecast hourly prices up to 20 years in advance. In Ukraine, we’d be happy to start with monthly or quarterly projections…
Their models include:
– Marginal costs for each generating unit (hourly)
– Technical-economic characteristics of plants
– Fuel costs
– Environmental costs
– Grid constraints
– Real-time demand-supply balance
– Available system reserves
– Cross-border trade volumes and available infrastructure
The model builds a “generation stack” — from cheapest to most expensive — and defines the market price based on which block “closes” the demand in a given hour. That “marginal” block sets the market price.
ORCED-PL also uses stochastic modeling to assess volatility — how much prices can swing, and whether such swings can be predicted. Because real markets don’t follow textbooks — they react to weather, geopolitics, outages, and more.
Conclusion: Yes, it’s complicated. But it’s possible. And it works in Europe — so it can work in Ukraine.
Forecasting isn’t an Excel sheet with 7 columns. It’s a complex system of interlinked data and models. That’s why, in the EU, forecasting is handled by institutions with access to all market, technical, and regulatory inputs. Without that, any AI becomes a “black box of random results.”
Ukraine’s electricity market must be based on an open, transparent, and mandatory forecasting model — a true foundation for trust, investment, and stability.
We urge the NEURC, Ministry of Energy, and other relevant bodies to address this critical gap: the absence of a systemic electricity price forecast in Ukraine.
Its consequences include:
– Direct financial losses for consumers and the state budget
– Destabilization of industrial operations
– Impossibility of strategic planning for banks and investors
– Chaotic and unpredictable market behavior
We call on the regulator to consider developing or implementing an open, mandatory forecasting model — following European best practices, adapted to Ukrainian realities, while maintaining the standards of transparency, reliability, and integrity accepted in the EU.
After such a model is developed, it is vital to ensure:
– Access to forecasts via a secure portal on NEURC or Market Operator websites
– Usage rights for suppliers, producers, traders, and consumers participating in the market
– If full data cannot be public due to wartime constraints — then at least an official forecast figure or range should be published monthly or quarterly, based on the approved model
This is not a mere recommendation — it is an urgent, strategic need for the stability of the energy market and the broader economy.
Sincerely,
Energy Union