Volatile is a forecasting lab for power markets. We build probabilistic forecasts and trading signals that tell you when to close, where to bid and how to split your volume.
Day-ahead · Intraday · Balancing API, SDK & Platform
Solutions
Our solutions tackle the most challenging problems that trading desks face every day. Each one is built on our own forecasts.
01 · Intraday
In continuous trading, what counts is when and at what price you close your positions.
See where prices are heading.
Price forecasts from your execution time show the expected path and its range.
Time your closing.
Decide when and at what price to close, based on the latest forecast and market liquidity.
02 · Multi-market
Where does your volume earn more tomorrow: in balancing or in the day-ahead auction? And what do you keep for intraday?
Compare markets directly.
Day-ahead price bands and aFRR/mFRR acceptance probabilities side by side.
Adjust through the day.
Intraday forecasts show when the market has moved since the day before.
03 · Day-ahead
With the full price distribution in view, your optimizer weighs every outcome and finds the optimal bidding strategy.
Optimize across scenarios.
Feed quantiles into your stochastic optimization. Bids, schedules and hedges are planned for the range of outcomes.
Know your risk per quarter-hour.
Quantify market uncertainty in every quarter-hour, from evening peaks to negative prices, and decide how much risk to take.
04 · Balancing
In capacity auctions, every bid is a trade-off between price and the chance of acceptance.
Know your bid price.
For aFRR and mFRR, we forecast the bid price for your volume at different acceptance probabilities. For FCR, you get price quantiles.
Bid according to your risk appetite.
Choose the bid that fits how much risk you want to take, whether you prioritize acceptance or price.
Forecasts
Every solution builds on our forecasts. You can also use them directly, as input for your own models, algos and optimizers.
Market
Balancing capacity
FCR · aFRR · mFRR · DE
Day-ahead auction
DE-LU · FR · BE · AT · NL
Intraday continuous
DE
Balancing energy
aFRR · mFRR · DE
How certain is tomorrow's price?
Where are intraday prices heading?
At which price will my volume be accepted?
Where will balancing energy prices clear?
See how Volatile performs on your portfolio.
Request a backtestPlatform
Our platform turns forecasts into decisions. It explains what drives each price, flags the hours that matter and supports your bidding. Every feature is also available via API.
Explainability
For every forecast, the platform shows the drivers behind it, such as wind, solar, load and cross-border flows. You know why a price is expected, not just which one. That makes forecasts easier to check, trust and act on.
Intelligence features
Ready-made services on top of our forecasts.
Get warned when negative prices become likely, per quarter-hour and for 3-, 4- or 6-hour blocks. Plan curtailment ahead instead of reacting on the day.
Derive aFRR and mFRR bids from acceptance probabilities. Set how much risk you want to take, and see the bid price that fits your volume.
Data richness
Behind every forecast is a pipeline that models the energy system itself.
Weather, outages, cross-border coupling and market behavior, combined with our own physics-based fundamental models that stack analytical modeling and machine learning. One data and model pipeline, built by people who know power markets inside out.
12
bidding zones
>1500
model runs per day
24/7
live operation, built for production
1 min
intraday update cycle
How it works
Use our forecasts as input for your algorithms, or work with them in the platform. Or both.
For systems & algorithms
Feed forecasts into your trading bots, algorithms, optimizers or EMS. Get started in a few steps, using our example scripts.
from volatile import Client client = Client(api_key="YOUR_KEY") da = client.day_ahead.latest(zone="DE-LU") for update in client.intraday.stream(zone="DE"): strategy.on_forecast(update)
For traders & analysts
For teams who need more than inputs: analysis, explanations and decision support in the browser. Live forecasts and backtests in one place.
Getting started
01 · Backtest
Historical forecasts for your markets and periods. Evaluate them on your own metrics before you commit to anything.
backtest_DE-LU.csv · illustrative data
02 · Live trial
The forecasts run live in your setup and for your portfolio. You see how they perform in your daily operations.
live stream · illustrative data
03 · Subscription
Start with the markets and products that proved themselves in the test. Add more whenever you need them.
Use cases
Find your role and see which solutions and forecasts fit your daily decisions.
Operators and optimizers of batteries and flexible assets
"In which market does my battery earn the most tomorrow?"
Trading houses, utility trading teams, quant traders
"How do I position in day-ahead and intraday, and what do I feed my algos?"
Power plants, CHP and utilities with own assets
"When does my plant run, and what do I bid in the auction and in balancing?"
FAQ
Many forecasts stop at a single number. We forecast the full price distribution with our own fundamental and market models, and build solutions for concrete trading decisions on top, together with the desks that use them.
If you need one number, use the median. The quantiles show how likely the price is to cross your threshold and how uncertain the market is, for adaptive thresholds, probability-weighted bids or stochastic optimizers.
By whether it improves your decisions. We check calibrated quantiles, the right direction and the P&L of a strategy in the backtest, because averages like MAE miss the spikes and negative prices that matter.
Every historical forecast comes from a model trained only on data available at that time. No look-ahead, no training data in the test period. The backtest shows what to expect live.
Both. Our forecasts and signals feed your own strategies and optimizers, and our intraday timing accounts for liquidity and order book depth. On request, we develop strategies together with your desk.
Yes, and we recommend it. Start with a backtest, then run our forecasts live next to your current setup and compare both on the same decisions before you decide.