Claude · Investing Prompt

Mag 7 Overexposure Check

Most people holding SPY, QQQ, and a few tech stocks are massively concentrated in the same 7 companies without knowing it. This prompt exposes hidden overlap and shows you how to fix it.
1
Copy the prompt below
2
Open Claude.ai and paste it in
3
List all your ETFs and individual stock holdings with approximate allocation %
The Prompt
You are a portfolio construction specialist focused on concentration risk and diversification analysis. I want to find out my true exposure to the Magnificent 7 stocks (Apple, Microsoft, Nvidia, Alphabet, Amazon, Meta, Tesla) across my entire portfolio — including hidden exposure through ETFs. **1. Hidden Concentration Map** For each ETF I hold, estimate its approximate Mag 7 weighting. Then calculate: - My total effective Mag 7 exposure as a % of my whole portfolio - Which of the 7 I'm most concentrated in (accounting for direct holdings + ETF overlap) - A simple table: [Stock] → [Direct %] + [ETF Exposure %] = [Total %] **2. Concentration Risk Assessment** - Is my Mag 7 exposure above or below typical benchmarks (S&P 500 is ~32%, QQQ ~45%)? - What's the downside scenario if Mag 7 sells off 30%? How much does my portfolio drop? - Am I getting false diversification — thinking I'm spread across many funds but really just owning the same stocks? **3. What I'm Missing** Based on my holdings, identify: - Sectors I have little to no exposure to - Geographies I'm underweight (ex-US, emerging markets, etc.) - Asset classes missing entirely (bonds, commodities, real estate) **4. Rebalancing Options** Give me 3 concrete ways to reduce Mag 7 concentration without abandoning growth exposure: - Option A: Low disruption (small ETF swap) - Option B: Medium rebalance (trim + reallocate) - Option C: Structural change (for those wanting real diversification) **5. Final Concentration Score** Rate my portfolio concentration: 🟢 Healthy / 🟡 Elevated / 🔴 Dangerously concentrated Please share your full holdings list with approximate % allocations and I'll run the analysis.