I asked Claude to analyze my stock portfolio risk. Here's what it found.

Most people have no idea what their actual portfolio risk looks like. Here's how to find out in Claude — with real numbers, in under a minute.

The problem with "diversification"

You might own 5 different stocks and think you're diversified. But if they're all in tech, or all correlated to the same macro factor, you're not. The correlation between AAPL and MSFT is ~0.82. Owning both is barely better than owning just one of them for risk purposes.

Portfolio analysis tools that show you this clearly — in plain language — don't exist for regular investors. Until now.

The setup

Install FinanceKit MCP and ask Claude:

claude mcp add financekit -- uvx --from financekit-mcp financekit

Or use the free tier (no terminal): mcpize.com/mcp/financekit-mcp

Example portfolio analysis

Prompt: "I have a portfolio: 35% AAPL, 25% MSFT, 20% NVDA, 10% GOOGL, 10% AMZN. What's my actual risk exposure? Am I diversified? What should I know?"

What Claude returns (from FinanceKit's portfolio_analysis + correlation_matrix + risk_metrics tools):

Portfolio composition:

Total: 5 positions
Sector breakdown: 100% Technology (zero sector diversification)

Correlation matrix (30-day):
| | AAPL | MSFT | NVDA | GOOGL | AMZN |
|--|--|--|--|--|--|
| AAPL | 1.00 | 0.81 | 0.74 | 0.76 | 0.69 |
| MSFT | 0.81 | 1.00 | 0.78 | 0.85 | 0.73 |
| NVDA | 0.74 | 0.78 | 1.00 | 0.71 | 0.65 |

These 5 positions are highly correlated. A sector-wide sell-off would affect all of them simultaneously.

Risk metrics (annualized):

VaR (95%, 1-day): -2.8% (there's a 5% chance of losing more than this in a single day)
Beta vs S&P 500: 1.32 (30% more volatile than the market)
Sharpe Ratio: 1.45 (good risk-adjusted return, but concentrated)
Max Drawdown (12mo): -23.4%

Claude's assessment:
"Your portfolio is concentrated in a single sector with high inter-asset correlation. The VaR suggests meaningful daily downside risk. While the Sharpe ratio is reasonable, a significant tech sector correction would impact all 5 positions simultaneously — there's no meaningful hedging happening here. Consider adding positions in healthcare, energy, or financials to reduce sector correlation."

That analysis — with specific numbers, correlation matrix, and actionable recommendation — would cost $500/hour from a financial advisor. Claude produces it in 8 seconds.

What the tools actually compute

FinanceKit uses:

VaR (Value at Risk): Monte Carlo or historical simulation to estimate downside probability
Sharpe Ratio: (Return - Risk-Free Rate) / Standard Deviation
Correlation matrix: Rolling covariance computed on yfinance historical data
Beta: Covariance(portfolio, SPY) / Variance(SPY)
Max Drawdown: Peak-to-trough decline in the lookback period

All computed from free market data (yfinance + CoinGecko for crypto). Zero API keys.

Important disclaimer

This is informational analysis, not financial advice. The numbers are real and the analysis is technically sound, but investing decisions should involve qualified professionals.