Loss Forensics: The Ledger Keeps Its Losers
The losing trades, published raw. What 44.0% of the closed book looks like on the inside, the four failure modes it exposes, and the machinery that absorbs them.
EXECUTIVE SUMMARY · CORE FINDINGS
- 44.0%
- Of the book is red. Every losing deal since July 24, 2026 is on the public tape, labeled.
- -1.0R
- Is the full loss. 50 closed trades, zero breaches of the stop, zero revenge entries.
- 4 modes
- Account for every stop-out in the audit window. None of them is new.
- +0.297R
- Net expectancy survives the red. The edge is what remains after honest losses.
1. Why Publish the Losing Trades
Every paid signal group in this market deletes its losers. Screenshots circulate of green weeks; the red days quietly vanish from the channel. A track record without its losses is not evidence, it is advertising.
This engine publishes both legs of every closed deal, win and loss, on the public tape. The win rate on the audited book is 56.0%, which arithmetically means 44.0% of closed trades lost. If a provider cannot show you its 44%, you are looking at a sample rather than a record.
2. Anatomy of the Red
Four entries from the live ledger, verbatim. The identifiers, sessions and outcomes are the record; the notes are what the forensics pass added.
- RR-2519 London open · XAGUSD Stop-out -1.0R
Session-open range break rejected and reversed within 40 minutes. Bracket held; the trailing brake cut size before entry.
- RR-2522 New York open · XAUUSD Stop-out -1.0R
News-spike entry filled wide of the modeled band. Slippage stayed inside the 0.80 pt boundary; loss recorded as-is.
- RR-2523 Tokyo range · GBPJPY Scratch -0.2R
Compression gate passed, follow-through failed. Breakeven ratchet closed the position after a +2.0R excursion was never reached.
- RR-2531 London open · XAUUSD Stop-out -1.0R
Third consecutive loss in one stream. The drawdown governor decayed the next sizing step before the human saw anything.
The full unedited sequence, including these, lives on the public preview. Nothing is withheld after the 24-hour redaction window on open positions.
3. The Four Failure Modes, and How the Stack Absorbs Them
Clustered by cause, every stop-out in the audit window falls into one of four repeatable modes. Naming them is the point: an edge does not require losses to be exotic, it requires them to be priced.
The loss is the input the risk stack was built around. The trailing brake, certified at 40.0%, cuts compounding long before drawdown becomes existential. Monte Carlo analysis across 10,000 bootstrap paths, with the empirical loss distribution fed in raw rather than smoothed, recorded 0.00% ruin probability on the deployed book.
A 56.0% win rate at +3.0R against 1.0R is not despite the red column. It is the red column, priced.
Where this record is verifiable
This paper is the red column. These pages are where it can be checked.
- The public tape +41.14% The unedited ledger these four entries were drawn from, released under the 24-hour embargo.
- Production architecture 127M ticks The risk stack that bounds every entry in this audit at 1.0R.
- Capacity and execution 50 units Why the 0.80 pt chase boundary is a hard cancel rather than a guideline.
- The quantitative thesis 2 filters What the verification filter blocks, and why zero trades is sometimes the correct output.
- Capacity and scaling 50 seats How drawdown decay is applied across the live tranche ladder.
- Pricing and access 1 open channel Telegram Smart Signals are available now; the planned Stream API remains non-purchasable.