The Commitment You Can't Choose, and the Four Ways LPs Get It Wrong

The problem most commitment models skip

A limited partner sets a target — say 20% of the portfolio in private equity — and then makes the only decision actually available: how much to commit. What gets deployed is the GP's call, spread over years the LP doesn't control. The two numbers are not the same, and the gap between them is not noise — it is structural.

Call it the slack: the difference between the capital an LP commits and the capital that is ever actually at work at the program's peak. It arises from the ordinary mechanics of the call-and-distribution cycle — capital is drawn gradually while earlier distributions are already flowing back, so a committed dollar is rarely a deployed dollar. An LP that commits exactly its target will, almost by construction, sit under its target exposure for most of the program's life.

Four ways LPs respond — and what each one costs

Strip the decision to its essence and there are four benchmark policies. Each is a real institutional posture; each has a precise failure mode:

• Commit at target. The intuitive choice — commit the 20%. It systematically underinvests, because slack guarantees deployment falls short of the commitment. Safe on liquidity, but it quietly leaves return on the table, every year.
• Commit to the liquidity limit. Commit as much as the program could fund if calls came fast and distributions slow — the maximum exposure that still never breaches the liquidity threshold. This is the right reference point: it achieves the most exposure with zero shortfall risk.
• Overcommit to offset slack. Scale the commitment up to compensate for expected slack, so deployment lands near target. It reaches the exposure — but it now operates above the liquidity limit in adverse states, accepting real shortfall risk to do so.
• Optimize to the institution's own risk posture. Choose the commitment that balances the cost of under-exposure against the probability of a liquidity shortfall, weighted by the LP's actual tolerance. This is the only one of the four that uses the institution's preferences rather than a rule of thumb.

The result that should change the decision

Here is the part that matters for governance. When the liquidity threshold is high enough — when the program can absorb a fast draw — the commit-to-the-limit policy dominates all the others outright, with no need to estimate any probabilities at all. The decision isn't close, and it doesn't depend on a forecast. Only when liquidity is genuinely binding does the trade-off become real — and then the right commitment depends on the institution's risk posture, not on a market view.

That reframes the commitment question. It is usually treated as a forecasting problem ("what will the GPs draw?"). It is better treated as a structural one: where does this program sit relative to its liquidity limit, and which of the four postures does that imply? Get that right and the forecast often doesn't matter; get it wrong and no forecast saves you.

The four-policy framework and the conditions under which one dominates are grounded in PrivateMetrics's proprietary research and the commitment-modeling engine behind the platform — built on a decade-plus of production private-markets portfolio optimization. The full analysis is available on request.