Negative electricity prices in Hungary: the first year of 15-minute trading

Negative electricity prices in Hungary came in 894 of the 35,040 day-ahead quarter-hours between 1 October 2025 and 30 September 2026. Of those, 88.9% fell between 10:00 and 16:00, and the median negative price was -2.61 EUR/MWh. Most came when residual load was below nuclear output, and 59 sat in hours whose average was zero or above.

This was the first year of 15-minute day-ahead trading in Hungary. The study counts every quarter-hour priced below zero in the Hungarian (HU) bidding zone, the area that shares one wholesale price, and relates the counts to solar output, residual load and cross-border flows.

What changed on the HUPX day-ahead market on 1 October 2025?

The single day-ahead coupling (SDAC) moved from hourly to 15-minute market time units (MTUs) for delivery from 1 October 2025, as the NEMO Committee, the body of nominated electricity market operators (NEMOs), confirmed. HUPX, the Hungarian Power Exchange that runs the Hungarian day-ahead market, published its first 15-minute results shortly after 13:00 on 30 September, covering the 96 quarter-hours of 1 October. Every price in this study is the clearing price of one quarter-hour.

How many quarter-hours had a negative price in Hungary?

In 894 of the 35,040 quarter-hours, the day-ahead price was below zero, which is 2.6% of the year. Another 267 quarter-hours cleared at exactly 0 EUR/MWh, so 1,161 quarter-hours (3.3%) were zero or below.

No quarter-hour was negative in the four months from October 2025 to January 2026. October 2025 came closest, with a lowest price of 0.09 EUR/MWh.

Which months had the most negative prices?

April 2026 had the most negative quarter-hours: 314, on 20 different days. May followed with 202 on 11 days and June with 148 on 7 days. March had 76 on 5 days, and February had 14 on a single day. The count then fell to 67 in July, 28 in August and 45 in September.

Negative quarter-hours per month

Day-ahead quarter-hours with a price below zero, Hungary, October 2025 to September 2026.

Monthly average7575October 2025 to September 2026

0100200300400Avg 0

OctNovDecJanFebMarAprMayJunJulAugSep

Negative quarter-hours, October 2025 to September 2026

Month	Negative quarter-hours
Oct 2025	0
Nov 2025	0
Dec 2025	0
Jan 2026	0
Feb 2026	14
Mar 2026	76
Apr 2026	314
May 2026	202
Jun 2026	148
Jul 2026	67
Aug 2026	28
Sep 2026	45

Month	Quarter-hours	Below zero	Zero or below	Lowest price (EUR/MWh)	Days with a negative price
Oct 2025	2,980	0	0	0.09	0
Nov 2025	2,880	0	0	14.4	0
Dec 2025	2,976	0	0	47.28	0
Jan 2026	2,976	0	0	10.69	0
Feb 2026	2,688	14	19	-0.45	1
Mar 2026	2,972	76	92	-7.11	5
Apr 2026	2,880	314	373	-500	20
May 2026	2,976	202	265	-500	11
Jun 2026	2,880	148	186	-55.18	7
Jul 2026	2,976	67	117	-10.96	5
Aug 2026	2,976	28	47	-1.59	2
Sep 2026	2,880	45	62	-3.08	3

Source: Day-ahead prices for the HU bidding zone, Energy-Charts API (Fraunhofer ISE), from Bundesnetzagentur | SMARD.de, CC BY 4.0

The year therefore had one spring peak rather than a summer one. The sections below look at what the power system was doing in those quarter-hours.

At what time of day do negative prices occur?

Negative prices clustered at midday. Of the 894 negative quarter-hours, 795 (88.9%) fell between 10:00 and 16:00 Budapest time. The hour from 13:00 to 14:00 had the most, with 176. None fell between 18:00 and 07:00.

Negative quarter-hours by hour of the day

All negative quarter-hours of the year, grouped by the Budapest clock hour they fall in.

Hourly average3737October 2025 to September 2026

050100150200

036912151821

Negative quarter-hours, October 2025 to September 2026

Hour	Negative quarter-hours
00:00 to 01:00	0
01:00 to 02:00	0
02:00 to 03:00	0
03:00 to 04:00	0
04:00 to 05:00	0
05:00 to 06:00	0
06:00 to 07:00	0
07:00 to 08:00	3
08:00 to 09:00	15
09:00 to 10:00	32
10:00 to 11:00	84
11:00 to 12:00	118
12:00 to 13:00	162
13:00 to 14:00	176
14:00 to 15:00	158
15:00 to 16:00	97
16:00 to 17:00	39
17:00 to 18:00	10
18:00 to 19:00	0
19:00 to 20:00	0
20:00 to 21:00	0
21:00 to 22:00	0
22:00 to 23:00	0
23:00 to 00:00	0

Show the data table

Hour (Budapest time)	Negative quarter-hours	Share of that hour's quarter-hours (%)
00:00 to 01:00	0	0
01:00 to 02:00	0	0
02:00 to 03:00	0	0
03:00 to 04:00	0	0
04:00 to 05:00	0	0
05:00 to 06:00	0	0
06:00 to 07:00	0	0
07:00 to 08:00	3	0.2
08:00 to 09:00	15	1
09:00 to 10:00	32	2.2
10:00 to 11:00	84	5.8
11:00 to 12:00	118	8.1
12:00 to 13:00	162	11.1
13:00 to 14:00	176	12.1
14:00 to 15:00	158	10.8
15:00 to 16:00	97	6.6
16:00 to 17:00	39	2.7
17:00 to 18:00	10	0.7
18:00 to 19:00	0	0
19:00 to 20:00	0	0
20:00 to 21:00	0	0
21:00 to 22:00	0	0
22:00 to 23:00	0	0
23:00 to 00:00	0	0

Source: Day-ahead prices for the HU bidding zone, Energy-Charts API (Fraunhofer ISE), from Bundesnetzagentur | SMARD.de, CC BY 4.0

The month-by-hour map shows how the shape changes through the year. April has the tallest midday block, June is flatter and starts earliest, in the 07:00 hour, and February's few negative quarter-hours sit in a narrow window around noon.

When negative prices happen: month by hour

Count of negative quarter-hours in each month from October 2025 to September 2026 (rows) and Budapest clock hour (columns). Darker cells have more.

OctNovDecJanFebMarAprMayJunJulAugSep0246810121416182022

063negative quarter-hours in that month and hour

Show the data table

Month	00	01	02	03	04	05	06	07	08	09	10	11	12	13	14	15	16	17	18	19	20	21	22	23
Oct 2025	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Nov 2025	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Dec 2025	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Jan 2026	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Feb 2026	0	0	0	0	0	0	0	0	0	0	1	3	4	4	2	0	0	0	0	0	0	0	0	0
Mar 2026	0	0	0	0	0	0	0	0	1	3	12	17	20	15	7	1	0	0	0	0	0	0	0	0
Apr 2026	0	0	0	0	0	0	0	0	4	12	23	36	56	61	63	39	17	3	0	0	0	0	0	0
May 2026	0	0	0	0	0	0	0	0	0	5	22	25	36	36	36	30	11	1	0	0	0	0	0	0
Jun 2026	0	0	0	0	0	0	0	3	10	12	12	15	21	28	21	12	8	6	0	0	0	0	0	0
Jul 2026	0	0	0	0	0	0	0	0	0	0	11	8	11	12	13	9	3	0	0	0	0	0	0	0
Aug 2026	0	0	0	0	0	0	0	0	0	0	0	4	6	8	6	4	0	0	0	0	0	0	0	0
Sep 2026	0	0	0	0	0	0	0	0	0	0	3	10	8	12	10	2	0	0	0	0	0	0	0	0

Source: Day-ahead prices for the HU bidding zone, Energy-Charts API (Fraunhofer ISE), from Bundesnetzagentur | SMARD.de, CC BY 4.0

The position inside the hour makes little difference. Each of the four quarters of the clock hour holds between 218 and 232 negative quarter-hours.

Are negative prices more common at weekends and on holidays?

Yes. 6.8% of weekend quarter-hours were negative, against 0.9% of weekday quarter-hours, and weekends held 76% of all negative quarter-hours.

The first negative quarter-hour fell on a Saturday and the last on a Sunday. Both days with a price at the -500 EUR/MWh minimum were also days off: 26 April 2026 was a Sunday, and 1 May 2026 was a Friday and Labour Day. 1 May is a public holiday under section 102(1) of the Labour Code.

The weekday figure includes public holidays such as 1 May, so it may overstate the rate on ordinary working days. Lower demand on days off is a likely contributor, but this analysis does not measure it.

How far below zero did prices go?

Most negative prices were shallow. The median negative price was -2.61 EUR/MWh, but the mean was -33.68 EUR/MWh, because a small number of quarter-hours cleared far lower. Prices between 0 and -1 EUR/MWh made up 37.2% of negative quarter-hours, and prices between -1 and -10 EUR/MWh another 30.3%.

Each band includes its deeper edge: a price of exactly -1 EUR/MWh falls in the -1 to -10 band, and exactly -10 EUR/MWh in the -10 to -50 band.

How far below zero prices went

Negative quarter-hours by price band, EUR/MWh.

Price band (EUR/MWh)	Quarter-hours	Share of negative quarter-hours (%)
Below 0 to −1	333	37.2
−1 to −10	271	30.3
−10 to −50	164	18.3
−50 to −100	45	5
−100 to above −500	65	7.3
−500 (the price floor on those dates)	16	1.8

Source: Day-ahead prices for the HU bidding zone, Energy-Charts API (Fraunhofer ISE), from Bundesnetzagentur | SMARD.de, CC BY 4.0

The deepest quarter-hours hit the price floor. In 16 quarter-hours the price cleared at -500 EUR/MWh, all on 26 April and 1 May 2026. That value is the reference harmonised minimum clearing price for SDAC in the methodology approved by the EU Agency for the Cooperation of Energy Regulators (ACER) on 4 February 2026. The methodology lowers the limit by 100 EUR/MWh when the price in at least one bidding zone falls below 70% of it in at least two market time units on at least two different days within 30 rolling days.

The NEMO Committee reported on 7 May 2026 that this condition was met for delivery on 26 April and again for 1 May 2026, in several European countries. The minimum became -600 EUR/MWh for delivery from 29 May 2026. From June to September the lowest Hungarian price was -55.18 EUR/MWh, so no quarter-hour came near either limit.

A price at the floor shows that the market reached its limit, not how far below it a free price would have cleared.

How many negative quarter-hours does an hourly average hide?

59 negative quarter-hours (6.6% of the total) sat inside an hour whose average of the four quarter-hour prices was zero or above, so an hourly average would not show them as negative. These are brief dips between higher-priced quarter-hours.

This average is a calculation on quarter-hour prices, and an hourly auction would not have cleared at it.

How long do negative price runs last?

Negative prices came in 80 runs, where a run is a stretch of consecutive negative quarter-hours. The median run lasted 7 quarter-hours. The longest lasted 41 consecutive quarter-hours and started at 07:45 on 7 June 2026, a Sunday.

A run ends at the first quarter-hour priced at zero or above, including exactly 0 EUR/MWh. A stretch broken by one zero price therefore counts as two runs.

What was the power system doing during negative quarter-hours?

Negative prices clustered when residual load, meaning load minus solar and wind output, was below nuclear output. That condition held in 4,771 quarter-hours, and 16.1% of them were negative. When residual load was at or above nuclear output, only 0.4% of 30,269 quarter-hours were negative.

Residual load below nuclear output means solar, wind and nuclear together exceeded domestic load, so the surplus likely had to be exported, stored or reduced. The study of Hungary's duck curve and residual load looks at residual load by hour over a longer period.

Hungary was a net exporter in 90.8% of negative quarter-hours, against 19% of all quarter-hours. Scheduled trade comes out of the same market clearing as the price, so this describes co-occurrence, not independent evidence. In the negative quarter-hours the mean net import was -1,217 MW, which is a net export of 1,217 MW, and mean public solar output was 2,595 MW. The study of how midday solar turned Hungary's imports into exports covers that shift.

Does more solar output mean more negative prices?

Not in a straight line. The share of negative quarter-hours rose with public solar output up to 3,000 MW and then fell.

Negative prices and solar output

Share of quarter-hours with a negative price, by the public solar generation in that quarter-hour.

Public solar generation	Quarter-hours	Negative	Share negative (%)
Under 1,000 MW	24,484	0	0
1,000 to 2,000 MW	3,352	132	3.9
2,000 to 3,000 MW	3,787	580	15.3
3,000 to 4,000 MW	3,059	182	5.9
4,000 MW and above	358	0	0

Source: Day-ahead prices for the HU bidding zone, Energy-Charts API (Fraunhofer ISE), from Bundesnetzagentur | SMARD.de, CC BY 4.0; Actual generation, load and scheduled cross-border trade for Hungary, ENTSO-E Transparency Platform via the Energy-Charts API

No quarter-hour with public solar output under 1,000 MW was negative, out of 24,484. The share was 3.9% at 1,000 to 2,000 MW and peaked at 15.3% at 2,000 to 3,000 MW. It fell to 5.9% at 3,000 to 4,000 MW and to zero at 4,000 MW and above, a band of only 358 quarter-hours.

This does not mean that more solar output lowers the risk of a negative price. The public solar series (reported to ENTSO-E by MAVIR) covers utility-scale plants and leaves out most household and other small rooftop systems, so solar figures here are lower than total solar output. The detail is in the study of why published Hungarian solar capacity figures differ.

The study does not test why the share falls above 3,000 MW. Two likely explanations are output reduction at negative prices (the series is recorded after curtailment) and higher weekday load, since weekends held 76% of negative quarter-hours.

How this analysis was done

Data. Day-ahead prices for the HU bidding zone come from the Energy-Charts API of Fraunhofer ISE, which carries Bundesnetzagentur | SMARD.de data under CC BY 4.0. Solar, wind and nuclear generation, load and scheduled cross-border trade come from the ENTSO-E Transparency Platform of the European Network of Transmission System Operators for Electricity (ENTSO-E), also through the Energy-Charts API. Data refreshed: 10 October 2026. The CSV datasets behind the study are listed at the top of the page.

Period and time. The period runs from 1 October 2025 00:00 to 30 September 2026 24:00 Budapest time (Central European Time in winter, Central European Summer Time in summer), which gives 35,040 quarter-hours. Hours and weekdays follow Budapest local time. The clock changes explain why March 2026 has 2,972 quarter-hours and October 2025 has 2,980.

Definitions.

Limitations.

• The study covers day-ahead prices only. Intraday and balancing prices are not included.
• The public solar series leaves out most household and other small rooftop systems, so solar figures here are lower than total solar output. MAVIR's data page labels its generation by production type as excluding household-scale power plant and prosumer solar, which it lists separately. The same gap affects residual load.
• Cross-border values are scheduled commercial trade, not physical flows.
• It is one year, the first of 15-minute trading, so it has no hourly-auction year to compare with, and it shows association, not cause.

The methods page lists the data sources and definitions used across this site.

What this means for solar park owners, battery developers, traders and grid planners

Solar park owners

• Number to use: 88.9% of negative quarter-hours fell between 10:00 and 16:00, the hours when a solar park produces most.
• Question to ask: what share of a park's output falls in those quarter-hours, and at what depth?
• Trap to avoid: quoting the mean of -33.68 EUR/MWh as typical, when the median negative price was -2.61 EUR/MWh.

Battery developers

• Number to use: the median negative run lasted 7 quarter-hours and the longest 41, and 76% of negative quarter-hours fell on weekends.
• Question to ask: does the model's time step resolve a run of 7 quarter-hours, or average it away?
• Trap to avoid: sizing against the 16 quarter-hours at -500 EUR/MWh, which fell on two days.

Traders

• Number to use: the minimum clearing price in force on the delivery date, -500 EUR/MWh until delivery on 28 May 2026 and -600 EUR/MWh from 29 May 2026.
• Question to ask: does the contract or position reference quarter-hour prices or an hourly average? 59 negative quarter-hours sat in hours whose average was zero or above.
• Trap to avoid: reading a price at the limit as a measure of the surplus. It shows only that the limit was reached.

Grid planners

• Number to use: a mean net export of 1,217 MW in negative quarter-hours, and 16.1% of quarter-hours negative when residual load was below nuclear output.
• Question to ask: how does that export compare with the interconnector capability available at midday?
• Trap to avoid: treating the export link as independent evidence, since scheduled trade comes out of the same market clearing. Solar figures here are lower than total solar output, so true residual load was likely lower than the one used here.