Our Approach
How we forecast
Futuratty produces forward-looking intelligence by synthesising prediction market signals, fundamental macroeconomic data, and proprietary scenario models. Every forecast carries a probability estimate and a confidence interval.
Why prediction markets matter
Prediction markets have emerged as the most accurate public forecasting mechanism available. Platforms like Polymarket, Kalshi, and Metaculus now process over $10 billion in monthly volume, consistently outperforming polls, pundit forecasts, and traditional survey-based models.
The mechanism is straightforward: participants stake capital on the likelihood of specific outcomes. The resulting market price reflects the aggregated beliefs of thousands of informed participants, each with their own information edge. Markets that misprice outcomes attract arbitrageurs who correct them. The result is a continuously updated probability estimate that incorporates all available information.
However, raw prediction market data has limitations. Markets are thin on niche topics. Prices can reflect liquidity constraints rather than true probabilities. And market participants may have systematic biases. Futuratty addresses these gaps.
Our three-layer approach
Layer 1: Market signal aggregation
We aggregate real-time prices from major prediction markets (Polymarket, Kalshi, Metaculus, PredictIt where available) and calculate volume-weighted consensus probabilities. We track bid-ask spreads, trading volume, and price momentum to distinguish genuine signal from noise. For events not covered by prediction markets, we use superforecaster panels and structured estimation techniques.
Layer 2: Fundamental data modelling
Prediction markets are supplemented with fundamental data analysis. For monetary policy forecasts, we model central bank reaction functions using employment, inflation, and output gap data from FRED, ECB Statistical Data Warehouse, and ONS. For real estate signals, we integrate land registry data, building permit trends, and capital flow indicators. For regulatory forecasts, we analyse legislative pipelines, consultation documents, and precedent analysis.
Layer 3: Scenario analysis and portfolio translation
Raw probabilities are necessary but insufficient for investment decisions. We build if/then scenario trees that map forecast outcomes to portfolio-relevant consequences. For example: if the UK raises CGT to 28% (61% probability), then Mediterranean property demand from UK investors rises 15-22% within 18 months. This translation layer is where Futuratty intelligence becomes actionable.
What makes our forecasts different
Most wealth intelligence platforms look backward: portfolio aggregation, performance reporting, compliance monitoring. Futuratty is forward-looking by design. We don't tell you what happened - we tell you what's likely to happen and what it means for your portfolio.
Every Futuratty forecast includes:
- A specific probability estimate — not vague sentiment, but a number you can reason about
- A confidence interval — acknowledging the range of uncertainty around our central estimate
- Source transparency — every data source and prediction market cited, so you can verify independently
- Scenario trees — if/then analysis mapping outcomes to cross-asset portfolio impact
- A defined timeframe — every forecast is falsifiable and time-bound
Calibration and track record
A forecasting methodology is only as good as its calibration. We track every forecast we publish against actual outcomes and report our accuracy systematically. Well-calibrated forecasts mean that events we assign 70% probability should occur approximately 70% of the time. Overconfidence and underconfidence are both systematic errors we actively correct for.
As Futuratty builds its public forecast track record, calibration data will be published on this page. We believe transparency about accuracy — including our misses — is essential to credibility.
Frequently asked questions
How accurate are prediction market-based forecasts?
Prediction markets have consistently outperformed polls and expert panels across political, economic, and sporting events. Meta-analyses show they are well-calibrated on average, though individual market prices can be temporarily distorted by liquidity constraints or speculative activity. Futuratty mitigates these risks through multi-source aggregation and fundamental data cross-validation.
How often are forecasts updated?
Active forecasts are updated when material new information changes the probability estimate by more than 5 percentage points, or when a significant data release occurs. High-priority forecasts (monetary policy, major regulatory events) are monitored in near-real-time during active market hours.
Can I commission bespoke research?
Yes. Futuratty produces tailored intelligence reports for family offices, UHNWI, and institutional allocators. Bespoke engagements cover specific scenarios, geographies, asset classes, or time horizons that are not part of our standard coverage. Contact us to discuss your requirements.
Is Futuratty a regulated financial advisor?
No. Futuratty provides informational intelligence and scenario analysis. We are not authorised by the FCA or any other financial regulator to provide personal financial advice. Our forecasts are probabilistic estimates, not investment recommendations. Always consult qualified financial professionals before making investment decisions.
What data sources do you use?
We aggregate data from prediction markets (Polymarket, Kalshi, Metaculus), macroeconomic databases (FRED, ECB SDW, ONS), property registries (Cyprus Land Registry, UK Land Registry), trade databases (UN Comtrade), and proprietary alternative data sources. Every forecast cites its sources.
Ready for forward-looking intelligence?
Tell us what decisions you're facing and we'll assess fit.