One platform · one signal · one conviction language
See the next six weeks before the market prices them.
DIANA is the digital fusion of PhD energy-meteorology and 24 years of institutional trading — a daily research engine that reads weather regimes the way a senior desk does, and tells you how much to trust the call. Delivered every morning, before markets open.
The edge, in one line
+22–26pp
Direction edge over ECMWF · Weeks 4–5
| 73% | High-conviction hit rate |
| 13 | European energy markets |
| 45yr | ERA5 reanalysis, queryable |
Direction accuracy on observed weather · multi-year out-of-sample validation vs the ECMWF energy-weather forecast.
The proof, in one chart
DIANA pulls ahead at the horizons where positions are built.
Direction accuracy on observed weather, head-to-head against the ECMWF energy-weather forecast over multi-year validation. EC owns Week 1. By Week 2 its ensemble has converged toward climatology. Weeks 3–6 — where the multi-week trade is built — are where DIANA’s edge opens up.
DIANAECMWF energy forecast
|
61%
|
67%
|
66%
|
54%
|
61%
|
59%
|
56%
|
57%
|
||||
|
50% — coin flip
|
|||||||||||
|
42%
|
36%
|
31%
|
29%
|
||||||||
|
W1
−6pp
|
W2
+11pp
|
W3
+19pp
|
W4
+22pp
|
W5
+26pp
|
W6
+28pp
|
||||||
Where the trade lives · W3 to W6
45yr
Backtested weather
13
European markets
+22–26pp
Edge over ECMWF · W4 / W5
73%
High-tier hit rate
Why it beats consensus
A weather model can’t beat ECMWF. A regime engine can.
DIANA doesn’t try to out-forecast the ensemble at Week 1 — nobody does. It plays a different game past Week 2, where the ensemble has nothing left to say.
01 · The problem
ECMWF fades to average
Past Week 2 the ensemble spreads toward climatology. On direction, the energy-weather forecast drifts to a coin flip — exactly when the multi-week position is being built.
02 · The method
DIANA reads the regime
45 years of ERA5 reanalysis, pattern-matched. DIANA finds which historical regimes today resembles and projects the W0–W6 distribution — signal that holds where the ensemble has given up.
03 · The output
Conviction, not narrative
Every call ships with direction, magnitude and a High / Medium / Low conviction tier. Size up on High, lean on Medium, stand down on Low — decided before the week starts.
Conviction is the product
Hit rate climbs with the tier.
| High | 73% | |
| Med | 65% | |
| Low | 61% | |
Direction accuracy by tier · 5-year out-of-sample backtest (n = 8,439). The tiering is the trade: you know the odds before you put the position on.
On the weeks that move books
+43pp
Week 3 · regime-change weeks
DIANA 70% · ECMWF 27% on the extreme, regime-change weeks — the ones that actually move a book. That’s the tail; the chart above is the all-weeks average (+19pp at W3). The edge is widest exactly when it’s worth the most.
What DIANA stands for
Digital · Intelligent · Artificial · Numerical · Advisor
D
Digital
I
Intelligent
A
Artificial
N
Numerical
A
Advisor
|
Digital
|
45 years of ERA5 reanalysis — every storm, ridge and vortex on record — queryable on demand. |
|
Intelligent
|
Pattern matching that reads regimes the way a senior meteorologist does — faster, and on every market at once. |
|
Artificial
|
Self-calibrating each morning — yesterday’s observation feeds today’s conviction tier. |
|
Numerical
|
Direction, magnitude and HIGH / MEDIUM / LOW conviction — not narrative, not vibes. |
|
Advisor
|
Six connected reports, R0–R5 — one signal language across trader, hedger, planner and analyst. |
The daily chain
Six reports. One conviction language. Every morning.
Each report answers one question and hands its output to the next. Open the chain and the desk has a complete view — not six vendors to reconcile.
|
R0
|
R1
|
R2
|
R3
|
R4
|
R5
|
|||||
|
Status
|
The New Normal
|
Analogs
|
Hydrology
|
DIANA Weather
|
Trade
|
|||||
|
Pipeline health
|
DIANA analogue-year fit
|
Best-match years
|
Reservoir trajectory
|
W1 to W6
|
Synthesis · in build
|
Ask DIANA
Six reports — and a chat to query them all.
Ask any market, week or driver in plain English. DIANA answers with the number and the reasoning — then opens the report it came from.
● Ask DIANA
Quick — why is NP power bullish in W3?
A blocking high parks over Scandinavia — your Select analogs ran ~16% below-normal wind (a wind drought). Thin generation lifts prices and the engine agrees. Conviction HIGH. OPENS R4 →
Coverage
13 European energy markets. One signal language.
Nordic power
Sub-areas
Continental power
Germany · France · Spain
Energy-weather and hydrology per zone.
Gas
TTF · Henry Hub
Cross-commodity weather correlations into both hubs.
Trader, hedger, planner and analyst all read the same call — no translation layer between desks.
Behind DIANA
Meteorological science meets institutional trading.
Dr. Ivan Føre Svegaarden
CEO · Head of Energy Analysis & Weather Risk
PhD in Energy Meteorology · 13+ years applied research and energy trading. Architect of DIANA — the digital fusion of his Weather Regime Model research, trading practice and analyst workflow. Specialises in SRMC–power disconnect analysis, wind-drought dynamics and cross-market weather correlations.
Jan Erik Pedersen
Head of Portfolio Management & Risk
24+ years institutional commodity trading · MSc Finance, NHH. Senior positions at major European energy firms; fund founder and manager. Deep expertise in Nordic, German and French power, plus gas and cross-commodity dynamics.
Markets reward people who see first
Ready when you are.
A 30-minute briefing on how DIANA reads the next six weeks for your book.
Phone
Email
TradeWPower AS · Lillehammer, Norway · Since 2016 · Org.nr. 919 948 000
