Ruby Capitora data-analysis workspace displaying historical market charts and risk indicators

Evidence-Based Intelligence for First-Time Investors

Markets now generate more signals than any individual can reasonably weigh in real time. Ruby Capitora applies backtested statistical models to historical price and volume data, giving you a calibrated reference point before you commit capital — rather than a reaction to the latest headline.

Explore the Platform Review the Backtested Data

Every recommendation is traceable to a documented historical pattern, shown with its test period and risk classification — not presented as an unqualified prediction.

Why It Feels Harder Than It Should

Why Volatility Is Harder to Read Than the Headlines Suggest

The average UK retail investor now has access to more live market data than a professional trading desk had two decades ago. Access, however, is not the same as clarity. Faced with thousands of indicators updating by the second, most first-time investors default to one of two responses: analysis paralysis, or an emotional reaction to a short-term price movement.

Analysis paralysis: too many indicators, with no consistent framework for weighing them against one another.
Recency bias: the most recent price swing is given more weight than long-run historical patterns.
Reactive decision-making: portfolio changes driven by a headline rather than a documented rule set.
Inconsistent risk assessment: risk tolerance is estimated once, then rarely re-checked against actual market behaviour.

Ruby Capitora does not remove volatility — no model can. It replaces ad hoc judgement with a documented, testable process: the same inputs produce the same class of recommendation, on any given day, regardless of the news cycle.

Ruby Capitora workspace showing historical market data charts used for pattern analysis

Methodology

How the Analysis Engine Reaches a Recommendation

Rather than treat the model as a black box, Ruby Capitora sets out each stage of the process, from raw data to a validated, risk-adjusted suggestion.

01 — INGESTION

Data Ingestion

The engine ingests decades of historical price, volume and macroeconomic data from licensed market data providers, synthesising it into a single, cleaned dataset before any analysis begins.

02 — MODELLING

Predictive Modelling

Statistical models identify recurring correlations between historical market conditions and subsequent price behaviour, calibrating each pattern against multiple historical periods to reduce reliance on any single market cycle.

03 — CALIBRATION

Risk Calibration

Every identified pattern is weighted against your stated risk profile and time horizon, adjusting position sizing and volatility exposure before any recommendation reaches your dashboard.

04 — VALIDATION

Validated Output

Recommendations are validated against out-of-sample historical data — periods the model was not trained on — with the underlying reasoning shown alongside the suggestion, not hidden behind it.

Platform Tools

Functional Tools Built for Decision Support

Each tool is designed to reduce noise, not add another dashboard to monitor.

Real-Time Risk Scoring

A continuously updated score reflecting how far your current portfolio sits from your stated risk profile, recalculated as new market data arrives rather than once a quarter.

Backtested Strategy Library

A catalogue of documented strategies, each tested against historical market data across multiple market cycles, with the underlying assumptions and limitations disclosed alongside the results.

Automated Portfolio Rebalancing

Rebalancing suggestions triggered by measurable drift from your target allocation, not by discretionary judgement or market sentiment.

Tailored Volatility Alerts

Dynamic alerts that notify you only when data signals a significant deviation from your risk profile, reducing notification noise rather than adding to it.

Historical Evidence

What the Historical Data Shows — and What It Doesn't

Backtesting is a tool for evaluating consistency, not a promise about the future. The evidence below is presented in that spirit.

Illustrative Consistency Across Historical Test Periods

The illustration below represents how a documented rules-based strategy tracked relative to a broad market benchmark across three separate historical test periods, including periods of market drawdown. It is a structural illustration of the comparison method, not a chart of actual returns.

Test Period 1 Test Period 2 (Drawdown) Test Period 3

Illustrative representation only. Figures are used to explain the comparison method and do not represent actual investment returns or performance figures.

Methodology Disclaimer

Historical simulations are constructed using past data and reflect how a strategy would have performed under prior market conditions. They are a tool for evaluating a strategy's structure and statistical consistency — not a projection of future results. No investment decision should rely on backtested performance alone.

Data Source Transparency

Backtests draw on licensed historical market data from established financial data providers, covering UK and major international equity indices. Full methodology documentation, including test periods and assumptions, is available to registered users within the platform.

Questions Before You Start

Decision Support: Common Questions

Direct answers on regulation, data handling and how the platform behaves when markets move outside historical norms.

Is Ruby Capitora regulated in the same way as a traditional broker?

Ruby Capitora provides data analysis and decision-support tools rather than personal financial advice, and is not a broker-dealer. We do not hold or move client funds. If you place trades through a separate broker, we recommend checking that firm's regulatory status directly, and treating our output as one input into your own decision, not a substitute for it.

How is my data kept secure?

Account and portfolio data is encrypted in transit and at rest. We do not sell personal data to third parties, and access to raw historical datasets is restricted to the systems that generate your risk score and recommendations.

What is the difference between an AI model and a human broker?

A human broker applies judgement, which can vary day to day and is difficult to audit after the fact. Our models apply a fixed, documented set of rules to the same data, so comparable market conditions produce a comparable class of output, and each recommendation can be traced back to the historical pattern that informed it.

What happens when markets behave in ways the model hasn't seen before?

No historical dataset contains every possible future scenario. When current conditions fall outside the range the model was tested on, the platform flags the recommendation as lower-confidence and widens the suggested risk buffer, rather than presenting an unqualified suggestion.

Do I need trading experience to use the platform?

No prior trading experience is required. The platform is built for people who understand the general concept of risk and return but have not necessarily used a trading terminal before.

Data encrypted in transit and at rest