Risk Budget Under Correlated Selloff
A foundation portfolio holds 55% global equities, 35% investment-grade bonds, and 10% gold. Its target volatility is 10%; trailing 12-month realized volatility is 9.8%. The portfolio is in a 6.2% drawdown. Over the past 20 days the equity-bond correlation has moved from -0.25 to +0.45, the equity index's 200-day trend has turned negative, and the 10-year yield has risen 40 bp. Gold's trend is positive and its correlation to equities is -0.1. The monthly 95% VaR budget is 1.2% of assets; current VaR is estimated at 1.35%. The CIO wants to sell equities and buy bonds to restore the policy weights after the equity decline. The board is concerned about locking in losses. The investment committee asks for a risk-based recommendation with specific allocation bands.
1. Assess whether the correlation shift and current VaR breach should change the portfolio's risk budget and policy allocation. Be explicit about what VaR does and does not capture.
Read the worked answer
The correlation shift matters more than the small VaR breach. With equities and bonds positively correlated at +0.45, bonds are no longer a reliable hedge; portfolio volatility for a given equity weight rises because the covariance term turns positive. The 1.35% VaR against a 1.2% budget means the portfolio is consuming about 112% of its monthly risk allowance. But VaR is a threshold measure, not a loss ceiling; it says nothing about gap risk, liquidity, or tail dependence. The 200-day equity trend turning negative and the 10-year yield rising 40 bp both argue that the risk is persistent, not a one-day anomaly. The CIO's instinct to rebalance back to policy weights would buy more equities after a decline and sell bonds, which increases risk exactly when correlation is adverse. The correct move is to treat the policy weights as long-run anchors, not a tactical command, and to de-risk within bands.
2. Propose a specific de-risking path with allocation bands, using trend and correlation, that brings estimated VaR back inside the 1.2% budget without wholesale liquidation.
Read the worked answer
A staged path: cut global equities from 55% to 47-48%, raise gold from 10% to 12-13%, and hold bonds at 35% or reduce only modestly to 34% if yield momentum is negative for bonds. Cash can be 3-4%. Rationale: reducing equity by 7-8 percentage points cuts the largest risk contributor; gold has a -0.1 correlation to equities and positive trend, so it improves diversification; bonds still carry duration risk, so do not add aggressively while yields rise. Use bands of +/-5 percentage points around policy weights, with a hard trigger: if the equity 200-day trend remains negative and 20-day equity-bond correlation stays above +0.3 for another month, move to the lower equity band (45%). If the equity trend turns positive and correlation falls below 0, restore toward 55% in steps. Recalculate VaR after each step; target 1.0-1.1% to leave a buffer.
3. How would you monitor whether the de-risking was too early, too late, or about right?
Read the worked answer
Set ex-ante criteria. 'Too late' evidence: the portfolio breaches 1.2% VaR again within the next month, drawdown exceeds 8-9%, or equity-bond correlation rises further while equities fall. 'Too early' evidence: equities rally above the 200-day trend, breadth improves, and the portfolio's up-capture lags the benchmark by more than the intended risk reduction. 'About right': realized volatility falls toward 8-9%, VaR stays below 1.2%, and the portfolio's downside capture improves without a large sacrifice in upside capture. Track the de-risking against a shadow portfolio held at policy weights; if the shadow outperforms by more than the risk saved over a full quarter, revisit the trigger. Also monitor liquidity and transaction costs, since repeated band adjustments can create whipsaw.