The 60-Second Morning Risk Scan
What You Need
- Your current portfolio positions (even a rough list works)
- A news source: Reuters, Bloomberg terminal, or even a well-curated X feed
- Any LLM with web access (Claude with web search, ChatGPT with browsing, or Perplexity)
The Prompt
I hold these positions as of this morning:
[PASTE YOUR POSITIONS — ticker + approximate weight is enough]
Scan overnight and pre-market news for any of the following:
1. DIRECT HITS: Any news specifically about my holdings (earnings surprises, management changes, lawsuits, product recalls, analyst actions)
2. SECTOR RISK: Sector-wide events that affect 3+ of my names (tariff announcements, regulatory actions, commodity shocks)
3. MACRO SHIFTS: Overnight moves in rates, currencies, or commodities that change the thesis on my largest positions
4. PEER SIGNALS: If a peer of one of my holdings reported overnight, what does it imply for my position?
For each finding:
- Severity: 🔴 ACT NOW / 🟡 WATCH / 🟢 NOTE
- What happened (one sentence)
- What it means for my position (one sentence)
- Suggested action (one sentence)
What You Get
A prioritized risk report before the market opens. The 🔴 items get immediate attention. The 🟡 items go on your watchlist. The 🟢 items are context for later.
Most mornings it returns nothing critical — and that is fine. The scan confirming "no new risks" is itself valuable. You are not worrying all morning about what you might have missed.
Making It Automatic
The power move: save this prompt as a template with your positions pre-filled. Every morning you paste it, hit enter, and have your risk report in 60 seconds.
Even better: if you use Claude Code or any scriptable LLM tool, schedule it to run at 6:30 AM and email you the results. My system does exactly this — the morning brief arrives before I finish coffee.
Here is a simplified version of what that looks like:
import subprocess
from datetime import datetime
POSITIONS = """
AAPL 8.2%, MSFT 7.1%, GOOGL 5.3%, AMZN 4.8%, NVDA 6.2%,
JPM 3.1%, V 2.8%, UNH 3.5%, XOM 2.1%, LLY 4.4%
"""
prompt = f"""
Today is {datetime.now().strftime('%B %d, %Y')}.
I hold these positions: {POSITIONS}
[... rest of the scan prompt from above ...]
"""
# Send to your preferred LLM API and email the result
The exact implementation depends on your setup. If you use Python, this is 20 lines of code. If you use a no-code tool like Zapier, it is a 3-step automation.
The Compounding Advantage
Most portfolio managers check news reactively — they find out about problems when the position moves. This scan flips it: you find problems BEFORE they show up in your P&L.
Over time, this builds an information edge. You are not smarter than the market. But you are faster than the version of yourself who skips the morning check.
One catch avoided per quarter justifies the 60 seconds daily. Over a year, that is 6 hours of scanning that prevents one or two decisions you would regret.
What is Coming Next Week
Issue #3: The trade journal that writes itself. How to use AI to capture your decision-making process — and actually learn from it during your quarterly review.
AI Finance Brief is written by a team running live algorithmic trading systems. Every workflow is tested on real market data.
Free every week. Subscribe at aifinancebrief.com