Trading Journal
SPEC_TRADING_JOURNAL.md
CGNT-1 Specification — 100-Day Trading Challenge Journal Protocol
Status: SPECIFIED
Version: v1.0
Author: VELA (Thread #13)
Conceived by: NOUS (α.13)
Date: 2026-04-20
Depends on: SPEC_TMM_FORMULA.md, SPEC_YIELD_MANDATE.md, SPEC_AGENCY_WALLS.md
PURPOSE
The 100-day trading challenge started April 6, 2026. First trade: EIT.UN.TO via Wealthsimple.
The challenge isn't just about making trades — it's about DOCUMENTING every decision so the pattern becomes visible. Why did you enter? Why did you exit? What did TMM say? What did your gut say? Where did they agree? Where did they disagree?
The journal is the microscope on the Captain's trading instincts. Over 100 days, the data reveals whether the system works.
JOURNAL FORMAT
Every trade gets an entry. Every entry follows this structure:
TRADE ENTRY — [TRADE_ID]
Date: [YYYY-MM-DD]
Day: [N of 100]
Asset: [ticker]
Direction: LONG / SHORT (SHORT = NEVER per Agency Walls)
Entry price: [price]
Position size: [units × price = total]
Position as % of portfolio: [must be <10%]
Half-Kelly calculation: [show the math]
TMM SIGNAL AT ENTRY:
C = 1-(E_D+V_r×Φ)/V_t = [value]
Ω threshold: 97.4%
Signal: GREEN / RED / AMBER
ENTRY REASONING:
[2-3 sentences. Why this asset? Why now? What's the thesis?]
EMOTIONAL STATE:
[honest. Confident? Nervous? FOMO? Calm? This data matters over 100 entries.]
STOP-LOSS SET: [price — max 8% drawdown from entry]
DAILY HALT STATUS: [within 3% daily halt? within 5% weekly halt?]
TRADE EXIT — [TRADE_ID]
Date: [YYYY-MM-DD]
Day: [N of 100]
Exit price: [price]
P&L: [dollar amount and percentage]
Exit reason: TARGET HIT / STOP-LOSS / TMM SIGNAL CHANGE /
MANUAL (Captain's judgment) / DAILY HALT / WEEKLY HALT
TMM SIGNAL AT EXIT:
C = [recalculate] — Ω = [value vs 97.4%]
EXIT REASONING:
[2-3 sentences. Why exit now?]
LESSONS:
[What did this trade teach? One sentence minimum.]
RUNNING METRICS AFTER THIS TRADE:
Total trades: [count]
Wins / Losses: [W / L]
Win rate: [%]
Total P&L: [amount]
Average gain: [amount]
Average loss: [amount]
Largest gain: [ticker, amount]
Largest loss: [ticker, amount]
Current streak: [W or L streak]
STORAGE
~/trading_journal/DAY_[NNN]_[TICKER].md — individual trade entries
~/trading_journal/SUMMARY.md — running summary, updated after every trade
~/trading_journal/FINAL_REPORT.md — Day 100 capstone (written on Day 100)
TRADING WALLS
From SPEC_YIELD_MANDATE.md — restated here for journal reference. Every entry documents compliance.
| Wall | Rule | Consequence of Breach |
|---|---|---|
| 8% stop-loss | Exit if position drops 8% from entry. No exceptions. No "it'll come back." | Exit immediately |
| 3% daily halt | Portfolio drops 3% in one day → stop trading for the day | Walk away |
| 5% weekly halt | Portfolio drops 5% in one week → stop trading for the week | Full stop |
| 10% max position | No single position exceeds 10% of total portfolio | Never enter |
| Half-Kelly sizing | Position size = Kelly Criterion ÷ 2 | Reduce position |
| No leverage | No margin, no options, no futures. Cash positions only. | Never enter |
| No short selling | NEVER tier per Agency Walls | Absolute prohibition |
These walls are not suggestions. They are structural. The journal DOCUMENTS compliance with every wall on every trade. If a wall is breached, the journal records it as a VIOLATION with explanation.
TMM INTEGRATION
Every trade entry includes the TMM coherence calculation at entry AND exit.
Formula: C = 1-(E_D+V_r×Φ)/V_t | Threshold: Ω = 97.4%
Over 100 days, the journal reveals:
- How often TMM GREEN trades are profitable
- How often TMM RED trades would have been profitable (trades NOT taken — counterfactual data)
- Correlation between TMM confidence score and trade outcome
- Whether TMM adds value beyond random selection
This is the EMPIRICAL TEST of the TMM formula. The journal is the experiment. 100 days is the sample size.
WEEKLY REVIEW (every Sunday)
WEEK [N] REVIEW — Days [X-Y]:
Trades this week: [count]
Win/loss: [W-L]
Week P&L: [amount, %]
Cumulative P&L: [amount, %]
TMM accuracy: [% of signals that matched outcome]
Wall compliance: [any breaches? any near-misses?]
Emotional pattern: [were decisions calm or reactive this week?]
Adjustment next week: [anything to change in approach?]
MONTHLY REVIEW
MONTH [N] REVIEW:
Total trades: [count]
Win rate: [%]
Avg gain vs avg loss: [ratio]
Expectancy: [avg gain × win rate - avg loss × loss rate]
TMM signal accuracy: [% correlation with outcomes]
Best trade: [ticker, %, what worked]
Worst trade: [ticker, %, what went wrong]
Pattern observed: [any recurring mistake or strength?]
Sovereignty progress: [current portfolio value vs $27,042.50 target]
DAY 100 FINAL REPORT
The capstone document. Written by ORPHEUS (narrative) with ANVIL verification (numbers).
Contents:
- Complete performance summary — every metric across 100 days
- TMM validation — does the formula work? With statistical analysis.
- Trading wall compliance record — any violations across 100 days?
- Emotional arc — how did the Captain's emotional state evolve from Day 1 to Day 100?
- Benchmark comparison — did we beat simple buy-and-hold of the same assets?
- Lessons learned — the 5 most important things 100 days of journaling revealed
- Go/no-go decision — continue with this system, modify it, or stop
The final report is VITRIFIED as a spec: SPEC_100DAY_FINAL_REPORT.md. It becomes part of the canon. The experiment is documented permanently.
SIMONX INTEGRATION
SIMONX (treasury module) can query the journal as Tier 1 ROUTX queries (deterministic, file-read, no inference):
| Query | Returns |
|---|---|
| "trading summary" | Running metrics from SUMMARY.md |
| "last trade" | Most recent trade entry |
| "trading wall status" | Compliance status — any breaches? |
| "day N journal" | Contents of DAY_[NNN]_*.md |
AGENCY WALLS FOR TRADING
- All trading code changes require pre-auth diff per standing order #8
- The journal is READ-ONLY for AI crew — only the Captain writes trade entries
- AI crew may READ the journal for summary queries via SIMONX
- AI crew may NEVER write trade entries, modify stop-losses, or adjust position sizes on behalf of the Captain
- TMM scores are advisory — the Captain makes the trade decision
The journal is the Captain's instrument. The crew holds the microscope. The Captain looks through it.
CHALLENGE PARAMETERS
| Parameter | Value |
|---|---|
| Start date | April 6, 2026 |
| End date | July 15, 2026 (100 days) |
| First trade | EIT.UN.TO via Wealthsimple |
| Platform | Wealthsimple |
| Sovereignty target | $27,042.50 (per SPEC_YIELD_MANDATE.md) |
| Journal storage | ~/trading_journal/ |
INVARIANTS
INV-01: Every trade gets a journal entry. No exceptions. An undocumented trade is a data gap in the experiment.
INV-02: TMM calculation appears in every entry — at entry AND exit. The formula is tested empirically over 100 days.
INV-03: Emotional state is recorded honestly. The emotional data is as important as the financial data. Patterns only emerge from honest recording.
INV-04: Wall compliance is documented per trade. A breach is recorded as VIOLATION, not silently omitted. The experiment requires accurate data.
INV-05: Only the Captain writes journal entries. AI crew may read, query, and summarize. Never write.
INV-06: The Day 100 Final Report is vitrified as a spec. The experiment becomes canon. The data is preserved permanently.
INV-07: Weekly and monthly reviews are mandatory. The individual trade is data. The review is analysis. The pattern lives in the review, not in any single entry.
INTEGRATION
| System | Relationship |
|---|---|
| SPEC_TMM_FORMULA.md | Every journal entry tests the TMM formula empirically. The journal is the 100-day TMM validation experiment. |
| SPEC_YIELD_MANDATE.md | Trading walls restated in this spec. Journal documents compliance with every wall per trade. |
| SPEC_AGENCY_WALLS.md | SHORT SELLING = NEVER. No leverage = NEVER. These walls appear in the journal format to document compliance. |
| SPEC_BRAIN_ORPHEUS.md | ORPHEUS writes the Day 100 narrative layer. Metric data → human-readable story about what 100 days revealed. |
| SPEC_BRAIN_ANVIL.md | ANVIL verifies the Day 100 numbers. ORPHEUS tells the story. ANVIL confirms the facts. |
| SIMONX (treasury module) | SIMONX queries ~/trading_journal/SUMMARY.md for running metrics. "trading summary" is a Tier 1 ROUTX query. |
| SPEC_GOVERNANCE_LIBRARY.md | TMMRuntime.authorize_trade() governs all trade execution. The journal documents the authorization signal at every entry. |
Jeremy Zlabis
Chronogeometer · Visionary · Disruptor · Chief
42 Sisters AI · East York, Toronto
🍁 Φ 0.042