Your first live trade.

Before you place your first live trade

One pre-flight check that takes 60 seconds and protects you from 80% of disasters. Do this before the first signal arrives, then again any time the engine has been quiet for more than a day.

1. Wallet funded? Open the venue (Hyperliquid for most signals), check the margin balance. Must be at least 10x the maximum signal size. If the engine fires at $5 sizes, you want at least $50 free margin. More is fine.
2. BABA App authenticated? Open babacapital.app/app on your phone, confirm the brief card shows fresh data. If it's stuck on "Loading…", your license isn't being recognized — resolve before the next signal lands.
3. Telegram notifications on? Mute every channel that isn't BMI Premium. The signal alert is what you need to see; everything else is noise.
4. Read the latest /insights post. The morning intel tells you the regime the engine is operating in. Risk-off day? Expect more short signals. Quiet day? Expect fewer signals overall.
5. Decide your size policy. Default: take the engine's recommended size. If you have a clear reason to override (account in drawdown → smaller; high-conviction signal in a known regime → potentially larger), have the rule written down before the signal arrives, not after.

The signal arrives — step by step

Your phone buzzes. A BABA AI signal is in the Premium Telegram channel and in the BABA App.

1. Read the thesis. Five seconds. Does the AI's reasoning make sense given what you know about the asset? You don't have to be an expert — you just have to confirm nothing is obviously wrong. If "TON short, sustained downtrend, funding not crowded" sounds plausible, continue. If the thesis says "long because price went up a lot already" and Module 3 told you that's chase-prone, Skip.
2. Check the voter line. Two seconds. Are all three voters PASS, or only two? All three is highest conviction; two-of-three with one BLOCKED is borderline. Borderline trades fire at default size; don't size up borderline.
3. Check R:R. Two seconds. Max loss vs max gain. Engine targets ≥ 1.5:1. If you're seeing 0.8:1, something's off — the engine wouldn't normally fire that, so flag it (probably a calculation edge case) and Skip rather than fire on a malformed signal.
4. Approve or veto. The veto window is 60 seconds in AUTO_VETO mode. If you do nothing, the engine fires for you. If you Skip, no trade. If you tap Approve, the engine fires immediately.
5. Watch the fire confirmation. A second alert arrives within 5 seconds confirming the trade lands at the venue. If it doesn't arrive, something failed — open the venue UI to check.

During the trade — what to monitor

Most of the time: nothing. The engine has placed venue-side SL and TP orders. If the price hits either level, the position closes automatically and you get a Telegram notification. You don't need to babysit.

The exception: high-impact event windows. If the trade is open through an FOMC press conference, a CPI release, a Polymarket resolution event, or any other moment of expected sharp volatility, consider closing manually before the event. The engine's SL is good for normal volatility; it's not designed to survive 5% slippage through a print.

What to ignore: the minute-by-minute mark price. Profitable trades feel wrong half the time. The position spending two hours underwater before recovering to TP is normal. Watching the chart will not improve the outcome; it will only make you more likely to panic-close at the worst moment.

Closing a trade

Three ways a trade closes:

• At TP. The engine closes the position, you get a "trade closed +$X.XX" alert. Celebrate quietly; don't size up the next trade because you're feeling lucky.
• At SL. The engine closes the position, you get a "trade closed −$X.XX" alert. The thesis didn't work this time. This is normal — even great strategies have 40–50% loss rates. Do not revenge-trade the next signal.
• Manual early exit. Allowed but rarely correct. Two situations where it's justified: (1) the thesis explicitly broke (the news that drove the signal got reversed), (2) you need the margin for something else (e.g. another high-conviction signal is firing and you're at concurrent-position cap). "Feeling like" the trade isn't working is not a justification.

Post-trade review — 60-second debrief

After each closed trade, log three things in a notes file:

1. Outcome: +$X / −$X, hit TP / hit SL / closed manually.
2. Did the thesis play out? Yes / No / Partially. This is more important than the dollar outcome — a winning trade where the thesis didn't actually drive the move is luck; a losing trade where the thesis was correct but the timing was wrong is information.
3. Anything you did differently than the engine recommended? If yes, was it justified in hindsight?

After 20–30 closed trades you'll have enough data to evaluate your own discipline. Are you skipping signals you should be taking? Sizing up beyond defaults and losing on it? Moving SLs adversely? The log answers these. Without the log, you'll just have feelings — and feelings are reliably wrong about your own behavior.

Common first-trade mistakes

• Sizing 10x the recommended size on the first trade. The default is small for a reason. Earn the right to size up by closing 20 trades at the default first.
• Cancelling the venue-side SL after entry. "I'll watch it manually." No, you won't. Leave the SL.
• Trading the same alert twice from the channel and Telegram. The alert is the same signal; if you fire from both, you've doubled exposure.
• Approving in the last second of the veto window without reading. If you're not sure, Skip. The engine fires ∼48 setups per day at Tier 3; missing one is not catastrophic.
• Treating a winning trade as confirmation of intuition. Survivor bias. The thesis worked or it didn't — the outcome alone doesn't tell you which.