Strategy Track Record

109

Closed trades

59%

Win rate

+2.5%

Avg / trade

+0.8%

Median / trade

+61.1%

Best trade

−27.1%

Worst trade

Historical backtest · sequential compounding by trade order (not calendar time) · refreshed from the daily pipeline.

Backtest equity curve equal-weight, one trade at a time, 1.00 → 7.30×

Historical

x-axis = closed-trade order (sequence), not calendar dates · gross of costs & slippage

Each step is one closed trade applied to a single unit of capital in sequence. Flat-to-rising for most of the window, then steepening as a late cluster of semiconductor and quantum-sector signals (QCOM +61%, TXN +47%, CRWV +31%) compounds.

Live forward track record every published signal, marked to market

Live · daily

Loading the live ledger…

1,148

Closed trades

55%

Win rate

+2.0%

Avg / trade

+1.5%

Median / trade

293 / 71 / 784

Target / stop / horizon

2026-09-17

Ledger as of

—

Open positions

—

Currently green

—

Avg mark

—

Best open

—

Worst open

—

Since

Each dot is one published buy, placed on the day it fired (x) and its current 20-day mark-to-market return (y). Green above the line, red below. This record grows forward every trading day and cannot be hindsight-fit.

01 / How to read this

Two views, one discipline.

The backtest takes every past buy signal under one fixed rule set and compounds it sequentially — the honest “if you took every signal” number, but ordered by trade, not by date. The live record is the real thing: every signal the site actually published, priced at the open the next day and marked to market on a 20-day horizon. It is younger and humbler, and it is the one that cannot be fit after the fact.

Entry. Enter the bar after the signal fires — never on the signal bar itself.
Management. Scale-in adds, adaptive sells, and a hard −25% stop on every position.
Backtest accounting. Equal-weight, one trade at a time, compounded from 1.00.
Live accounting. Marked to market daily; 20-day hold; grows forward from launch.

Read this before the number.

No costs, no slippage. Backtest returns are gross — transaction costs, spread and slippage are not modelled. Real execution would be lower.
Sequence, not calendar. The 7.30× compounds one trade at a time in trade order. It is not a two-year buy-and-hold, and the curve flatters a small number of large winners.
The live record is shorter than the backtest. The forward ledger starts 2026-05-29 and now holds 1,148 closed trades; open positions are marked to market, not realised. The managed exit rules (stop −25% / target +15%) were introduced 2026-08-09 and applied retroactively, so managed stats before that date are in-sample — the plain 20-day-hold track is shown alongside on the ledger page for continuity.
Backtest ≠ live. Past signals simulated after the fact will always look cleaner than the forward record. We show both on purpose.
Not investment advice. A demonstration of signal behaviour for research only. No recommendation, no offer, no guarantee of future results.

02 / Backtest top performers by compounded return per symbol

Symbol	Trades	Win	Total
QCOM	1	100%	+61.1%
TXN	1	100%	+47.4%
CRWV	2	50%	+30.7%
HTZ	1	100%	+28.6%
MCHP	1	100%	+27.4%
QUBT	1	100%	+24.2%
AVGO	1	100%	+23.8%

Concentration is honest to note: the sample leans on semiconductor and quantum-sector names — exactly the universe the strategy is built to read.