Methodology

Effective: July 22, 2026, Last updated: July 23, 2026

This document sets out the methodology TVIEW LAB (the "Company") applies to the research, indicators, backtests, and market analysis it publishes. If the Research Principles describe where we aim, this document describes how we get there in practice. It is the minimum standard we hold ourselves to on every published output.

The methodology below is applied consistently across all of the Company's public research. Readers should interpret our results with the assumptions and limits described here in mind.

§1. Scope

This document applies to the following outputs:

• Strategy reports we publish
• Indicator analysis and combination testing
• Backtest results and their performance metrics
• Market reviews and data summaries

It does not apply to research users conduct themselves or to third-party data.

§2. Data sources

• Market data: TradingView, exchange public APIs, and licensed data vendors.
• Macro data: National statistical offices, central banks, and public data portals.
• Data timing, adjustment methodology, and how we handle missing data are noted in the footer of each report, and can also be verified through our contact channel.

§3. Backtest design principles

• Period: A minimum of three years, ideally spanning multiple market regimes (bull, bear, and low-volatility periods).
• Universe: We include only instruments that pass a minimum liquidity threshold within their asset class.
• Trading costs: We include commissions, slippage, and spread. Where these are not included, we say so.
• Money management: We report R-multiple, maximum drawdown, win rate, and Sharpe ratio together.
• Reproducibility: Rules and parameters are published with the results so that a third party can reproduce them.

§4. Statistical significance

• Results derived from a small sample or from a regime-biased window may not be statistically significant. We note this where applicable.
• Where needed to distinguish chance from real signal, we run auxiliary tests such as bootstrap or walk-forward analysis.

§5. Bias management

We recognize that the following biases cannot be fully eliminated, and we call out any applicable bias within each report:

• Survivorship bias
• Look-ahead bias
• Overfitting
• Sample selection bias

§6. Disclosure of limits

• Past performance does not guarantee future results.
• Backtest results do not fully reflect the latency, liquidity, and psychological factors of live execution.
• The statistics we publish are reference material. They do not replace a user's own judgment.

§7. Related documents

• Research Principles — the Company's research principles
• Editorial Policy — editorial and AI-usage principles
• Risk Disclosure — investment-related risks
• Disclaimer — the nature of the Service and the limits of our responsibility
• Terms of Service — general terms of use

§8. Revisions

This document is revised when the methodology improves or a new verification procedure is introduced. On each revision the "Last updated" date is updated, and material changes are announced through the Company's notice channels.