A financial prediction should be recorded before the outcome and objectively verifiable afterward.
Predict → Record → Evaluate → Verify
Day 1
Predict
The AI system generates its prediction before the relevant market outcome is known.
Day 1
Record
The prediction is preserved and timestamped so that the original prediction can later be independently compared with the actual result.
Day 2
Evaluate
Once the relevant market result is known, the original prediction is compared with actual market data.
Day 2
Verify
Anyone can compare the original, timestamped prediction with the actual outcome and the evaluation.
Public cross-validation
Where appropriate, selected predictions may also be published publicly before the outcome is known, including through LinkedIn. This can provide an additional, independent timestamped public reference.
The objective is to allow comparison between:
Don’t ask people to trust the prediction.
Give them the ability to verify it.
Transparency standard
Every result is labelled with what it actually is. We clearly distinguish between:
Prediction
A forward-looking model output, recorded before the outcome is known.
Actual market outcome
The observed market data after the relevant period has ended.
Historical evaluation
Comparison of past recorded predictions with what actually happened.
Backtesting
Applying a model to historical data it was not deployed on at the time.
Simulation
Hypothetical scenarios and modelled results — not real trading.
Live results
Outputs produced in real time, after deployment, and recorded as they occurred.
Realized performance
Actual results achieved with real capital. Never inferred from the categories above.
We never represent simulated, historical, backtested or research-stage performance as realized client performance.
Future concept · Subject to legal review · Not an offer
Future commercial principle — Pay for Success
Our intended future philosophy is alignment between our economic interests and those of our clients. For an applicable overnight use case:
Day 1
The tool generates and records the prediction.
Overnight
The relevant market outcome develops.
Day 2
The result is measured against the original prediction.
The intended future model is that an applicable performance-based charge would be associated with a verified successful result. If the defined result is unsuccessful, there would be no performance-based charge for that result.
This section describes a future commercial concept only. It is not a currently available service, a regulated financial or investment-advisory service, a finalized pricing model, or an offer. Any future model remains subject to review by company legal counsel and to all applicable licensing, regulatory and disclosure requirements.
Our principles
- Prediction must be measurable.
- Performance must be verifiable.
- Results must be transparent.
- Our interests should be aligned with the interests of our clients.
Predictions are probabilistic and may be wrong. Historical results do not guarantee future outcomes. Content on this site is research and decision-support information, not personalized investment advice.