H1 2026 RESULTS

Six months on the EGX100, every decision tracked live.

January–June 2026. Two strategies ran on the Egyptian Exchange in real time — every trade decision graded, tracked, and logged with identical exit logic.

9,345 trades tracked · 6,667 filled and closed · EGX100 universe

THE TAPE

Every closed trade. No cherry-picking.

1,095 day trades and 282 position trades, January–June 2026 — each bar is one real trade from the log: wins up, losses down.

Day-trading engine · 545 wins / 550 losses (49.8%) · +17.1% at 3% per trade

Position-trading engine · 188 wins / 94 losses (66.7%) · +32.5% at 3% per trade

The judgment

Cortex saw every signal. It skipped most of them.

The measure of an AI is not what it trades — it is what it refuses. Same engine, same period, identical exit logic: what Cortex kept vs everything it filtered out.

Day-trading engineCortexSkipped
Signals kept1,095 / 3,6402,545
Win rate49.8%42.0%
Avg R per trade+0.47R+0.25R
Max drawdown−0.4%−1.8%
Position-trading engineCortexSkipped
Signals kept282 / 3,0272,745
Win rate66.7%51.7%
Avg R per trade+0.74R+0.26R
Max drawdown−1.1%−25.0%

Both pools were live-tracked over the same period with identical exit logic — the only difference is Cortex’s selection.

The forward test

Rules froze in June. July was the exam.

Cortex’s selection rules were fixed before July — every signal that month was graded forward, in real time, with no refitting. A full out-of-sample month.

Day-trading · JulyCortexSkipped
Closed trades286434
Win rate52.8%47.7%
Avg R per trade+0.58R+0.35R
Profit factor2.771.96
Position-trading · JulyCortexSkipped
Closed trades69115
Win rate91.3%75.7%
Avg R per trade+1.09R+0.70R
Profit factor14.94.2

July return at 3% per trade: day +4.5% · positions +11.9% — vs EGX30 +6.2%, with under a tenth of the drawdown (−0.20% worst dip vs −1.81%).

Still a short record: part of the fills are simulated, figures are gross of costs, and a 91.3% month won’t repeat every month. August is month three — already running.

DAY TRADING — POWERS THE AUTOMATION

The engine behind fully automated execution.

Intraday to short holds. Of 5,434 trades tracked, Cortex selected 1,748 (1,095 filled and closed) — fewer, better trades, with a fraction of the drawdown.

49.8%

win rate — break-even is 29.9%wins avg +1.64R vs −0.70R losses · a 20-pt cushion

+0.47R

avg per trade1.9× the skipped pool · profit factor 2.32

−0.4%

max drawdownvs −1.8% for the skipped pool

+17.1%

H1 return1M EGP · 3% per trade · gross of costs

BREAK-EVEN MATH

Wins average +1.64R vs −0.70R losses — Cortex only needs a 29.9% win rate to break even. It delivered 49.8% — a 20-point cushion.

ACCOUNT GROWTH BY CALENDAR DAY — VS EGX30 / EGX70

All lines show % growth from 1 Jan 2026 (indices from their end-2025 close), so the account and the indices are directly comparable.

{{ l.name }}

WHY CORTEX BEATS THE INDEX

The index needs 100% of capital exposed all six months to earn its return; Cortex risked at most 3% per trade, was positive in 6 of 6 months, and spread its edge across 99 symbols. In the shaded correction it kept climbing while the EGX30 gave back ≈17 points.

POSITION TRADING — OFFERED TO USERS BY THEIR BROKER

The recommendations a broker puts in front of clients.

Multi-day holds with defined targets and stops. Of 3,911 trades tracked, Cortex selected just 360 (282 filled and closed) — two thirds of them winners.

66.7%

win rate — break-even is 38.3%wins avg +1.61R vs −1.00R losses · a 28-pt cushion

+0.74R

avg per trade2.8× the skipped pool · profit factor 3.22

−1.1%

max drawdownvs −25.0% for the skipped pool

+32.5%

H1 return1M EGP · 3% per trade · gross of costs

BREAK-EVEN MATH

Wins average +1.61R vs −1.00R losses — Cortex only needs a 38.3% win rate to break even. It delivered 66.7% — a 28-point cushion.

ACCOUNT GROWTH BY CALENDAR DAY — VS EGX30 / EGX70

All lines show % growth from 1 Jan 2026 (indices from their end-2025 close), so the account and the indices are directly comparable.

{{ l.name }}

WHY CORTEX BEATS THE INDEX

1.6× the EGX30 and 1.8× the EGX70 from just 282 trades, with at most 3% of capital at risk per trade versus 100% exposure to match the index. Its worst losing stretch cost ≈1.1% of the account — and in the shaded correction it kept climbing while the index fell.

HONEST READ

How to read these numbers.

FILL RATE

Unfilled entries are lost edge

Only ~63% of day-trading and ~77% of position-trading entries ever filled. Improving fill capture is an open workstream.

H2 — LIVE NOW

The next six months, pre-labelled

H2 tracking is already underway — every new decision flagged in real time, reported in the same formula-driven format every month.

FULLY TRACEABLE

Every number has a log line

Each figure on this page traces to a per-trade record — timestamps, grades, and exits — reviewable line by line in a briefing.

Walk through the full trade logs with us.

Every number on this page traces to a per-trade log — we'll open it live in a 30-minute briefing.

Book a briefing