Daily guide · FI · 2026-08-24
The number your Finland price model has to beat this week
All numbers computed from the week's published auction results at publication and frozen — this article is an honest snapshot, not a page that rewrites itself. Reproduce via the API.
The number to beat this week is 51,98 EUR/MWh: that is the persistence baseline's mean absolute error across the six days on record. Any Finland price model earning its place has to sit comfortably below it. For context, our own scorecard over the last 290 days shows an MAE of 20,56 EUR/MWh against a baseline of 21,94 EUR/MWh, so the bar this week is roughly double our recent working error.
What stands out is how uneven the naive baseline is day to day. Persistence does its worst job at the extremes: 65,73 EUR/MWh on 18 August, when the day mean was 81,87, and 63,40 on 23 August, when the day mean collapses to 4,64. Its cleanest day is 21 August, with an error of 25,84 against a mean of 108,55. The lesson is that persistence struggles when levels shift sharply between days.
The concrete takeaway: target the transition days. If your model can beat 65,73 on 18 August and 63,40 on 23 August, you will carry the week, because those two days do most of the damage to the 51,98 average.
Before you train anything on FI prices, you need the number your model has to beat. In day-ahead forecasting that number is naive persistence: predict that tomorrow's curve equals today's. It costs nothing, needs no features, and it is embarrassingly hard to beat — because power prices are dominated by daily and weekly rhythms that persistence gets for free. We computed it on this exact week:
51.98 €/MWh
persistence MAE on this week's FI auctions — the bar. A model that can't get under this adds negative value.
20.56 vs 21.94
our live production model vs its persistence baseline, trailing 290 scored days — the same scoring, in public, on /accuracy
Why the bar moves: this week, day by day
| Target day | Day mean €/MWh | Persistence MAE |
|---|---|---|
| 2026-08-18 | 81.87 | 65.73 |
| 2026-08-19 | 131.27 | 50.96 |
| 2026-08-20 | 88.49 | 42.78 |
| 2026-08-21 | 108.55 | 25.84 |
| 2026-08-22 | 68.04 | 63.19 |
| 2026-08-23 | 4.64 | 63.4 |
Calm days make persistence look unbeatable; transition days (weather swings, weekend boundaries) are where a real model earns its keep. Averages hide this — always look at the daily distribution.
The part everyone gets wrong: the features
The model is the easy half. The silent killer is lookahead: training on weather reanalysis instead of the forecast that existed at issue time, on revised load data, on same-day flow averages that hadn't finished flowing at the auction. Backtests built that way look brilliant and die in production. Our Quant feature matrices serve one row per (issue day, 15-minute slot) where every column is documented with when it became knowable — the same rows our own models train on:
curl -H "Authorization: Bearer YOUR_KEY" \
"https://voltcast.com/api/v1/features/FI?issue_from=2026-07-01&horizons=1"
# columns include: price lags · previous-run weather (as forecast AT issue)
# · grid drivers (lagged to last COMPLETE day) · day-ahead load/RES forecasts
# · ensemble spread · target_price label (null until matured, never imputed)
Start with Home: the persistence bar above needs nothing but GET /v1/prices/FI
(Home covers one zone of your choice). When your walk-forward backtest beats it honestly, prove it in public on the
Forecast Bench — forward-blind, open scoring code, and our
own model is auto-entered under identical rules. If you'd rather buy the plumbing than build it,
contact us about ML-grade data access.
Method & citation. Prices are hourly means of published day-ahead auction results (native 15-minute periods averaged; ENTSO-E/SMARD, attributed). Wholesale-price component only — grid fees and taxes come on top and vary by supplier. Cite as "Voltcast Research, voltcast.com/guides/quant-features-fi-2026-08-24".
The monthly European power roundup
Negative-price records, the biggest spreads, which zones were hardest to forecast — every number computed from our production data, on the 2nd of each month. No filler, unsubscribe anytime.