The Denominator Breach: Rebalancing an Illiquid Policy Portfolio

Start by separating the measured weight from the economic weight. The 26.4% figure rests on lagged NAVs after a sharp public-market decline, so part of the drift is a denominator artifact. If the office's view that marks are 5-8% rich is right, the economic weight is closer to 24.5-25.2% - at or just inside the band. Either way, a 1.4-point technical breach is not worth crystallizing a 12-15% secondary discount.

Better sequencing:

1. Restate the weight using a mark-to-model adjustment and a lagged-NAV convention written into policy, and report to the board with a remediation plan (policy breaches are governance events even when small).
2. Rebalance in liquid assets. The PE overweight was funded by an underweight in public equity; buy public equity with fixed income rather than selling PE.
3. Address the flow, not the stock. New commitments add exposure for years. Suspending new commitments for 2-3 years is the cleanest, cheapest way to let the weight drift down as NAVs are distributed.
4. Only sell secondaries if the weight exceeds the band by a wide margin after mark adjustment, or if liquidity coverage fails.

Selling a 12-15% discount to close a 1.4-point gap destroys more value than the drift it corrects.

Current coverage: requirement = (2 x $105m) + $300m = $510m against $1.05bn of Tier 1 assets, or 2.06x. That passes, but barely, and the pass is fragile for two reasons.

Stress it:

• Assume public equity falls another 20%. If Tier 1 is roughly $300m cash, $450m public equity and $300m bonds, the equity sleeve loses $90m and Tier 1 falls to about $960m.
• In a downturn capital calls accelerate, because GPs buy assets from distressed sellers and draw down facilities. Assume calls of $400m.
• Coverage becomes $960m / ($210m + $400m) = 1.57x, a breach.

The binding constraint is not spending - it is the call pace. Two fixes follow:

• Size the backstop to the calls. A committed credit line equal to roughly one year of stressed calls (around $400m) is cheaper than holding that much cash and directly addresses the tail.
• Model the second-order effect. Falling public markets raise the PE weight and raise calls and cut the denominator, all at once. The rule should be tested on the stressed denominator, not the current one, and re-run quarterly.

The reasoning confuses trailing returns with a forward-looking policy decision, and it ignores what the recent numbers actually measure.

• The return record is partly an artifact. Lagged, appraisal-based marks understate PE volatility and overstate risk-adjusted returns in a smoothed series; beta is being under-counted, not avoided.
• The illiquidity budget is already stretched. A 30% target against a 26% current weight and $640m of unfunded commitments implies a commitment pace well above the current one. The endowment would be systematically adding exposure during drawdowns - the opposite of a liquidity buffer.
• Spending is the liability. A 4.5% spending rule drawn from a portfolio with 30% in assets that cannot be sold for 8-10 years raises the probability of forced selling or borrowing in a bad year.
• Peer comparison is not a policy rationale. Neighboring endowments running 30%+ targets have different gift flows, debt and spending needs.

A defensible process: commission a liquidity/return study that models spending volatility under both targets, including a stressed path, and require the board to approve the target against an explicit illiquidity budget, not a return chart.