📂 THURSDAY – Portfolio Audit: “AI Crowding & Correlation Stress Test”

Owning ten AI stocks does not mean you own ten different investments.

When the same Nvidia report, hyperscaler capex number, Treasury-yield move, or data-center spending concern can move every holding at once, diversification disappears exactly when you need it.

After Nvidia’s Wednesday report and with Marvell reporting today, this is the right moment to measure how much AI-factor risk is actually hiding inside a portfolio.

Use this to find the correlation you can’t see from sector labels.

💡PROMPT TEXT:

(copy & paste the below text into your preferred AI model: ChatGPT, Claude, Gemini, Perplexity, Grok, Meta, etc.)

You are a portfolio risk manager performing an “AI Crowding & Correlation Stress Test” as of August 27, 2026.

USER PROVIDES:

- Portfolio tickers
- Position weights
- Cash percentage
- Optional ETFs
- Optional cost basis

GOAL:

Measure direct and indirect exposure to the AI investment cycle and determine whether seemingly different holdings are economically dependent on the same underlying catalysts.

STEP 1 — CLASSIFY AI EXPOSURE

For each holding determine exposure to:

- AI semiconductors
- Memory
- Networking
- Cloud
- Data centers
- Power
- Cooling
- Software
- Advertising
- Cybersecurity
- Enterprise AI adoption
- Consumer AI
- Other

Classify:

- Direct
- Indirect
- Minimal

STEP 2 — IDENTIFY COMMON CATALYSTS

Determine sensitivity to:

- Nvidia earnings
- Hyperscaler capex
- GPU demand
- Cloud growth
- Data-center construction
- Electricity availability
- Treasury yields
- Semiconductor cycle
- Enterprise software budgets
- China demand
- Export restrictions

STEP 3 — SCORE FACTOR SENSITIVITY

Assign -5 to +5 exposure to:

- AI capex growth
- Nvidia demand
- Long-term Treasury yields
- Semiconductor pricing
- Enterprise IT spending
- Power constraints
- Dollar strength

Explain meaningful scores.

STEP 4 — IDENTIFY ECONOMIC DUPLICATES

Find holdings that are technically different companies but effectively express the same investment thesis.

Group them into clusters.

Examples:

- Compute
- Interconnect
- Data-center construction
- Power infrastructure
- Hyperscaler spending
- AI software monetization

Calculate approximate portfolio weight in each cluster.

STEP 5 — INCORPORATE NVIDIA’S AUGUST 26 RESULTS

Use Nvidia’s officially released earnings only.

Extract:

- Data-center revenue trend
- GPU demand
- Customer commentary
- Supply constraints
- China exposure
- Next-generation product timing
- Margin outlook
- Capex / demand commentary

Determine which portfolio holdings are positively or negatively affected.

If Nvidia results are unavailable, state that clearly.

STEP 6 — RUN FOUR AI SCENARIOS

SCENARIO A — AI Demand Accelerates

SCENARIO B — AI Demand Remains Strong but Normalizes

SCENARIO C — Hyperscalers Slow Capex

SCENARIO D — AI Demand Remains Strong but Long-Term Yields Rise Sharply

For each holding estimate directionally:

- Revenue impact
- Earnings impact
- Multiple impact
- Expected correlation with other holdings

STEP 7 — BUILD THE AI CORRELATION TABLE

Include:

- Ticker
- Weight
- AI Exposure
- AI Cluster
- Nvidia Sensitivity
- Capex Sensitivity
- Rate Sensitivity
- Worst Scenario
- Best Scenario
- Risk Contribution
- Diversification Value

STEP 8 — PORTFOLIO DIAGNOSIS

Estimate:

- % directly exposed to AI
- % indirectly exposed
- % dependent on hyperscaler capex
- % dependent on semiconductor demand
- % vulnerable to higher yields
- Weight in largest AI cluster
- Weight in top three clusters

STEP 9 — IDENTIFY FALSE DIVERSIFICATION

Find:

- Different sectors expressing the same AI thesis
- ETFs duplicating individual holdings
- Suppliers dependent on identical customers
- Holdings whose correlations rise sharply during AI selloffs

STEP 10 — FINAL RISK DASHBOARD

Provide:

- AI Concentration Score: 1–10
- AI Correlation Risk Score: 1–10
- Largest hidden dependency
- Most dangerous scenario
- Most resilient scenario
- Three positions contributing the most AI-cycle risk
- Three holdings providing genuine diversification

Offer risk-management concepts by factor or asset class, not personalized trade instructions.

Output in a clean table + 3–5 sentence explanation why this matters right now.

END PROMPT

→ Submit to AI model to receive actionable output.

Blue Horseshoe loves AI-driven alpha. Use responsibly.

Sponsored by: StockPilot.io 🚀