Worked research guide · Sources and examples checked 17 October 2026 · AllMarkets Trading
A trading journal and a portfolio tracker answer different questions. A journal can summarize completed round trips and their fees. A holdings tracker records what remains and how it is valued. A paper-trading or historical replay tool records simulated decisions. Combining their numbers without matching those definitions can create a plausible dashboard with the wrong interpretation.
Start with the trading journal catalogue, portfolio tracker comparison and paper-trading and replay directory. The following workflow concerns record quality and reproducibility, rather than selecting a strategy or promising a return.
Write down the unit of each record
A broker fill is not necessarily a completed trade. One position may open through several executions and close in parts. Before importing a completed-round-trip journal, reconcile those fills into the documented schema and preserve the original execution export. Do not infer a complete account return from a list of trades that excludes deposits, open positions or other costs.
The local journal uses explicitly entered completed linear long or short trades in one reporting currency. The holdings analyzer uses units, prices, exchange-rate assumptions and remaining cost. It does not resolve a broker’s tax lots or calculate a cashflow-adjusted total return. Their separate input schemas are a useful reminder to keep the records separate.
Reconcile a small closed-trade example
Consider an invented long trade: 10 units bought at 100 and sold at 110, multiplier 1, total fees 2. Gross result is 100; net result is 98. If planned risk was 50 in the same currency, the journal’s net R multiple is 1.96. These are arithmetic descriptions of the entered trade, not an account’s percentage return.
| Record | What it says | What it does not establish |
|---|---|---|
| Completed trade | Net 98 after the entered 2 of fees | Return on total account equity |
| Remaining holding | Units valued using a dated price | Realized trading profit or a tax basis |
| Simulated order | Outcome under the stated fill model | A live execution or future performance |
Use the local closed-trade journal for the first record. Use the holdings analyzer for a separate snapshot. Keep currency and the treatment of fees explicit. A current holding’s unrealized difference from remaining cost is a different quantity from a closed trade’s net result.
Keep simulated execution in its own file
The replay lab uses a bounded daily-bar cash model with explicit order timing and costs. Its assumptions are intentionally narrower than a live broker’s order lifecycle. Label a replay export as simulated, and keep it separate from real fills so a later journal review does not treat practice results as observed execution.
For historical strategy testing, inspect the separate backtesting directory. A current screener match is not historical evidence. An apparently strong result needs a compatible point-in-time dataset, an execution convention and a clear statement of excluded costs before any performance interpretation.
Finish each review with a portable checkpoint
Save the original source, applied filters, date range, currency and a readable report. A filtered CSV is useful for a selected view, while a full JSON backup can preserve records hidden by the current filters. Test restoration on a copied dataset and verify counts and a few known calculations before replacing a primary working record.
Reconcile one small example before importing an entire account. If a discrepancy appears, identify whether it comes from record grouping, remaining holdings, fees, currency, date basis or simulated execution. Changing the display until totals look similar does not resolve a difference in measurement.