Portfolio Monitor diagnoses your portfolio’s risk using Modern Portfolio Theory (Markowitz, 1952). A 50% position can contribute 66% of total risk if highly correlated with your other holdings — covariance-based decomposition surfaces those weight-vs-risk mismatches and pairs them with correlation, concentration, and benchmark-relative performance views.
更新于—
内置技能
Driven
内置技能
Portfolio risk is not the weighted average of individual vols — it depends on how holdings move together. Covariance-based marginal contribution to variance surfaces positions whose risk share exceeds their weight share: a 50% position can contribute 66% of risk when it moves together with the rest of the book, and that mismatch is invisible in a standard weights table.
Pairwise correlations flag hidden co-movement in seemingly diversified positions. Pairs above 0.8 offer minimal diversification benefit and effectively behave as a single position during drawdowns; low cross-sector correlations indicate genuine spread. The matrix answers "am I really diversified, or just holding more names?"
Herfindahl-Hirschman Index at both the holding level and the sector level. A portfolio can have a low holding HHI but a high sector HHI — ten equal-weighted tech names look diversified by ticker count but are effectively a concentrated tech bet. Both views surface different failure modes.
Excess return per unit of volatility: (annualized return − risk-free rate) / annualized vol. A portfolio with higher raw return but lower Sharpe than its benchmark is delivering returns through risk, not skill — and will likely give them back the first time the market stops cooperating.
Peak-to-trough decline from 1-year daily cumulative returns. Captures the tail risk that volatility misses — the number that actually describes the lived experience of a selloff. A -24% drawdown vs