PowerTradingLab

What do you measure your short-term trading portfolio against? PowerTradingLab publishes free daily benchmarks for European short-term power trading. The buy-or-sell decision for each interval is published before the day-ahead auction, scored after delivery, and never changed afterwards.

Where do the benchmarks come from? The benchmarks grew out of a six-year study, A Retrospective Look at 2020–2026 Short-Term Power Trading Opportunities in Europe by Irina Chuchueva. The report shows which bidding zones held "speculative fat" and how it decayed over time.

Next trading action (Trailing Bias Value)

Forecast action of the “simplest trading strategy” for each interval: +1 means buy on DA, −1 means sell on DA, NA means there is not enough data for a decision.
ZoneDelivery dayPublishedBuySellFile
AL2026-11-102026-11-09 10:44 UTC096forecast
BE2026-11-102026-11-09 10:44 UTC7224forecast
CH2026-11-102026-11-09 10:44 UTC3660forecast
CZ2026-11-102026-11-09 10:44 UTC6036forecast
DE_LU2026-11-102026-11-09 10:44 UTC6036forecast
EE2026-11-102026-11-09 10:44 UTC4848forecast
HR2026-11-102026-11-09 10:44 UTC6432forecast
NL2026-11-102026-11-09 10:44 UTC3660forecast
NO52026-11-102026-11-09 10:44 UTC924forecast
RO2026-11-102026-11-09 10:44 UTC7224forecast
SE12026-11-102026-11-09 10:44 UTC2472forecast
SI2026-11-102026-11-09 10:44 UTC5244forecast

Latest scored day

EUR per 1 MW of trading quantity for each 15-minute interval, summed over the delivery day.
ZoneDelivery dayDaily “Speculative Fat”
(Extractable)
Daily “Simplest trading strategy” PnL
(Trailing bias)
15-minute intervals
scored / forecasted
File
AL2026-11-081,766.06no decision96 / 0score
BE2026-11-08581.0898.7796 / 96score
CH2026-11-082,548.80−1,124.1996 / 96score
CZ2026-11-08786.49−141.8796 / 96score
DE_LU2026-11-08985.13−12.0096 / 96score
EE2026-11-082,460.99509.4896 / 96score
HR2026-11-083,549.13705.5596 / 96score
NL2026-11-08390.82−370.0496 / 96score
NO52026-11-0856.9149.5596 / 92score
RO2026-11-084,164.46−245.1296 / 96score
SE12026-11-08231.07−46.8896 / 92score
SI2026-11-08727.90−167.1296 / 96score

Current month, running total

EUR per 1 MW of trading quantity, summed over the scored days of the month so far. Running: the values change every day until the month settles on the 20th of the following month.
ZoneMonth“Speculative Fat”
(Extractable)
“Simplest trading strategy” PnL
(Trailing bias)
Days scored /
days in month
File
ALNovember 202619,041.008,824.467 / 30monthly
BENovember 20267,816.23325.868 / 30monthly
CHNovember 202619,785.90−1,089.418 / 30monthly
CZNovember 202612,638.49167.078 / 30monthly
DE_LUNovember 202611,025.321,418.898 / 30monthly
EENovember 202619,665.521,553.498 / 30monthly
HRNovember 202633,461.56409.698 / 30monthly
NLNovember 20265,883.70−903.498 / 30monthly
NO5November 2026632.61168.118 / 30monthly
RONovember 202637,050.27−899.678 / 30monthly
SE1November 20262,947.42456.358 / 30monthly
SINovember 202610,348.973,094.218 / 30monthly

Last settled month

EUR per 1 MW of trading quantity, summed over the scored days of the month. Settled: every day has its final pass and the values will not change. File.
ZoneMonth“Speculative Fat”
(Extractable)
“Simplest trading strategy” PnL
(Trailing bias)
Days scored /
days in month
ALSeptember 202626,436.31—14 / 30
BESeptember 202613,205.75—14 / 30
CHSeptember 202632,190.54—14 / 30
CZSeptember 202617,289.55—14 / 30
DE_LUSeptember 202623,448.78—14 / 30
EESeptember 202617,000.19—14 / 30
HRSeptember 202664,022.54—14 / 30
NLSeptember 202610,561.06−3,330.6114 / 30
NO5September 20262,345.24—14 / 30
ROSeptember 202644,807.451,073.9114 / 30
SE1September 20268,859.54—14 / 30
SISeptember 202635,977.16—14 / 30

Two key benchmarks

“Speculative Fat”: how much a market holds — extractable_value
Imagine a trader with a crystal ball who knows in advance both the day-ahead (DA) and the imbalance (IM) prices. For every 15-minute interval this trader buys or sells 1 MW at DA and closes the position at the IM price. The profit of such a trader is extractable_value.
“Simplest trading strategy”: how much of market fat is repeatable — trailing_bias_value
Imagine a trader with no model, only a habit. For each hour of the day the trader looks at the previous three days, sees whether buying or selling at DA paid better in that hour, and follows the same pattern bidding for tomorrow. The profit of such a trader is trailing_bias_value.

Zones details · Methodology and Data Interface

Latest news

2026-11-02

Methodology 4 is live. From delivery day 3 November 2026, I compute the trailing bias value on a window tied to timestamps: for each hour, the three most recent complete observations of the same hour, including today’s if the zone has already published it in full. Today’s hours in zones that do not settle in euro are valued at the last exchange rate known at the moment of the decision. The decision now runs 20 minutes before the day-ahead auction closes, at 11:40 CET all year round, so in winter it sees one more hour of data than before. Files published earlier keep methodology 3; every file states its version in the methodology_version field.

All news