Market scenarios
Plan for the market’s paths, not one number
A market forecast that gives one probability tells you which way to lean. A plan has to hold up on the paths nobody expected. ScenarioPilot turns a market or macro question into four evidence-based scenarios, gives each one strategies, and attaches indicators that show which one is arriving.
Run your market questionWhat do you put in?
A theme and a region: the market question your plan depends on. For example:
- The euro area interest-rate path to 2028, and what it does to our refinancing
- Brent crude and our freight costs over the next three years
- Demand for data-centre power equipment in Central Europe
What do you get back?
The full foresight cycle on that question. Weak signals from public sources, each scored for relevance, impact and novelty and fact-checked. Drivers of change rated for impact and uncertainty. Four scenarios on the two critical uncertainties, each with a narrative, its implications and its early warning signs. Two strategies per scenario. Indicators with thresholds, so you know which world is arriving. The methodology page describes every step.
How is this different from a one-number market forecast?
A one-number forecast answers “which way?”. A plan needs “what if not?”. The table sets ScenarioPilot against Previct.ai, a free market-forecast terminal, as its public pages describe it.
| Compared | Previct.ai | ScenarioPilot |
|---|---|---|
| What you get | A probability that an index closes higher, such as “P(up) 54%”, with a confidence label and an expected move. | Four scenarios on the two most critical uncertainties. None of them is presented as the forecast; each is a future to plan against. |
| Where it comes from | News, economic developments and company earnings combined with “model-based analysis”. | Signals from news, web, research, patent and open-data searches, each tagged with how many independent sources report it, and traceable from every scenario, driver and strategy. |
| What would change the view | A base and a tail scenario written per instrument. | Indicators with thresholds. When a reading you log crosses one, a trigger opens for the world it belongs to. |
| What to do about it | A bullish or bearish view per instrument. | Two strategies per scenario, each with its kind (no-regret, hedge, option or bet), an accountable role, a trigger and the assumptions it relies on. |
| Track record | The home page lists profitable signals only; the complete record is on a page that asks for a PIN. | Your own forecasts on dated, resolvable questions, scored with the Brier score when you resolve them. |
| Who it is for | Traders (the terminal greets “Good day, trader”). | Planning, finance, treasury and strategy teams who have to commit to a plan before the market decides. |
Previct.ai facts from its public pages (previct.ai, previct.ai/about and previct.ai/terminal), checked on 12 November 2026. Previct.ai is not affiliated with ScenarioPilot.
How do you keep score?
Turn a key uncertainty into a dated question that resolves yes or no, such as “Will the deposit rate be below 2% on 31 December 2027?”, and record your probability. When it resolves, the forecast is scored with the Brier score: 0 is perfect, 1 is the worst. An ambiguous resolution is recorded but not scored. Every forecast stays in the project’s decision record, the misses with the hits.
What is it not?
- It is not investment advice. It makes no buy or sell call on any security and does not tell you how to trade.
- It does not stream market prices. Indicators move when you log a reading; nothing is watched automatically.
- It does not put probabilities on scenarios. The four worlds are plausible futures to plan against, not a forecast to bet on.
- It does not claim to beat the market. The only score it keeps is the one on the forecasts you write and resolve yourself.