5 AI Prompts Every Trader Should Be Using in 2026

1. The Pre-Trade Compression Prompt

What it does — When to use

Compresses 45 minutes of pre-trade research into a 90-second briefing. News, analyst sentiment, recent insider activity, earnings exposure, sector context — all in one structured response. Run it before opening any new position.

You are a senior buy-side analyst. I'm considering a position in [TICKER].

Build me a pre-trade brief covering:
1. Current consensus (bull/bear/neutral) with the strongest counter-argument to consensus
2. Three catalysts in the next 30 days (earnings, product, macro, regulatory)
3. Insider/institutional activity in the last 90 days — note any unusual filings
4. Sector tailwind/headwind summary (2 sentences each)
5. The single risk factor most likely to invalidate a long thesis

Format: bullet points. Cite primary sources. If data is uncertain or stale, say so explicitly.

Why it works: it forces the model into a specific role (buy-side analyst), defines the output structure, and includes a bias-correction layer (the counter-argument requirement). Use this with Perplexity Finance or Claude with web search for live data.

2. The Earnings Setup Decoder

What it does — When to use

Reads the current earnings setup and tells you whether it's a “show me” quarter or a “high bar” quarter — the difference between asymmetric upside and asymmetric downside. Run it one week before any earnings event in your portfolio or watchlist.

You are an earnings season strategist. The company [TICKER] reports earnings on [DATE].

Tell me:
1. Is the buy-side bar higher or lower than last quarter? Why?
2. What's the "whisper number" vs published consensus?
3. What guidance language would (a) trigger a 5%+ drop, (b) trigger a 5%+ rally?
4. Three specific Q&A questions analysts are likely to ask on the call
5. The non-consensus surprise (positive or negative) most likely to occur

Be specific. No hedging. If you don't have current data, mark it [STALE: verify].

Why it works:it frames earnings as a setup problem, not a prediction problem. You don't need to predict the earnings — you need to predict the bar the market has set. Trading the bar is repeatable. Trading the result is gambling.

3. The Risk-Reverse Stress Test

What it does — When to use

Forces the AI to argue againstyour trade — finding holes in your thesis you'd otherwise miss. Run it after you've decided to enter a position but before you click buy.

I'm planning to go long [TICKER] at [PRICE]. My thesis: [TWO SENTENCES].

You are a hedge fund risk officer. Your job is to argue against this trade.

Identify:
1. Three specific scenarios in the next 90 days where this trade loses 20%+
2. The data point or chart pattern that would prove me wrong fastest
3. Adverse selection: why might other smart money NOT be in this trade already?
4. Where would I be most likely to be exit-trapped (low-liquidity zones)?

Be brutal. Don't soften the analysis. Assume I'm overconfident.

Why it works:AI defaults to politeness. Forcing it into adversarial mode (and explicitly telling it to be “brutal”) flips that default. The result is the kind of red-team analysis that prevents catastrophic trades.

4. The Macro-to-Ticker Translator

What it does — When to use

Translates broad macro events (rate decisions, CPI prints, geopolitical shifts) into specific, actionable impacts on tickers in your watchlist. Run it before any FOMC meeting, CPI release, or major geopolitical event.

The Fed announces [DECISION] tomorrow at [TIME]. CPI prints [EXPECTATION].

You are a macro-to-equity translator. For each ticker below, tell me:
- Direction expected (up/down/flat)
- Magnitude (small / medium / large move)
- Mechanism — what specifically about this macro event drives this ticker?
- Hedging trade if I'm currently long

Tickers: [LIST 5-10 TICKERS]

Focus on second-order effects, not just "rates up = stocks down" surface logic.

Why it works: most traders read macro and trade equity separately. This prompt forces the model to connect the dots across asset classes — which is exactly what desks at major banks pay quants six figures to do manually.

5. The Position Sizing Calculator

What it does — When to use

Replaces gut-feel position sizing with a Kelly-criterion-adjusted, volatility-aware recommendation specific to your account size and edge. Run it every single time you size a position.

You are a portfolio risk manager.

Account size: $[AMOUNT]
Max risk per trade: [%]
Trade: long [TICKER] at $[ENTRY], stop at $[STOP], target $[TARGET]
Win rate (last 100 trades of this setup): [%]
Average winner / average loser ratio: [RATIO]

Calculate:
1. Kelly-optimal position size (and the half-Kelly you should actually use)
2. Position size in shares and dollars
3. Maximum drawdown if I take 5 losses in a row at this size
4. Position size if my win rate is 5% lower than I think (margin of safety)
5. The single edge metric I should track to know if my setup is degrading

Show your math.

Why it works: position sizing is where 90% of traders blow up — not stock picking. This prompt replaces emotion with arithmetic. The “win rate 5% lower” line is a margin-of-safety adjustment for overconfidence, which is the #1 cognitive bias in retail trading.

How to use these tonight

1. Pick the trade you're closest to taking right now.
2. Run prompt #1 (Pre-Trade Compression) in Perplexity Finance.
3. Run prompt #3 (Risk-Reverse Stress Test) in Claude.
4. Run prompt #5 (Position Sizing Calculator) with your real numbers.

If your trade survives all three, take it. If it doesn't, you just saved yourself a loss.

This is the workflow we deploy in every weekly Assistly edition: a tested AI tool, an actionable prompt, and a strategy you can run in 10 minutes. Nothing speculative. Just the edge, written down.