Performance & methodology
Performance, with the assumptions attached
How to assess a Wealthoak result: distinguish backtests from live observations, inspect costs and data, and verify the source before drawing a conclusion.
Three kinds of evidence
A backtest applies a defined strategy to historical data using stated execution assumptions. Paper trading records simulated orders as new data arrives. Live trading records real broker activity. These answer different questions and must be labelled separately.
A backtest is sensitive to data quality, model selection and assumptions about fills. A live record also needs account coverage, cash-flow treatment, costs and an explanation of how the published sample was selected. Neither guarantees future results.
| Area | What the report must contain |
|---|---|
| Identity | Engine version, configuration, eligible instruments, venue, currency and report owner. |
| Time and capital | Exact start/end dates, initial capital, deposits/withdrawals and position sizing. Historical coverage is instrument-specific. |
| Data | Vendor, interval, timestamps, missing-data handling, contract rolls and corporate actions where applicable. |
| Evaluation | Training/selection window, untouched test window, walk-forward procedure and number of parameter trials. |
| Costs | Commissions, spread, slippage, financing, borrow costs and treatment of rejected or partially filled orders. |
| Results | Net return series, maximum drawdown, recovery time, turnover, exposure and a comparable benchmark. |
| Provenance | Source export, calculation code, checksum, publication date, limitations and independent-review status. |
Risk before a headline number
Inspect the full equity series, worst drawdown, time below the previous peak, tail losses and concentration. A strategy can report a positive return while requiring a loss tolerance or capital commitment that its user cannot sustain.
Sharpe ratios require a defined periodic return series, sampling convention and risk-free-rate assumption. Annualized figures require the period and formula. We do not infer these statistics from a screenshot or a list of gross trade profits.
Check the calculation yourself
The research laboratory provides a small synthetic dataset and reproducible drawdown calculations. It demonstrates the mathematics without representing Atlas, Falcon or a historical market.
Download the report disclosure template or read the evaluation protocol.
Request source-backed documentation
For a specific engine, market or broker, request the relevant report. A request is not a promise that an independently reviewed report is available. Published status is recorded in the evidence register.