Introduction
The California State Teachers' Retirement System (CalSTRS) wants 14% of its portfolio in private equity, but it cannot buy 14%. A limited partner (LP) can only sign commitments to funds, which a general partner (GP) calls over several years while older funds send money back, so the figure a board sets (a share of portfolio value, with private holdings counted at net asset value, or NAV) and the figure the investment team controls (dollars committed) never line up. Keeping them close is the job of commitment pacing, and doing it at scale means deliberately committing more than the target.
The target has a second weakness: it is a percentage, and a percentage has a denominator. When listed stocks and bonds fall faster than private marks, an LP can drift above its range without signing anything new. How LPs respond, by slowing commitments, moving targets, or selling fund interests, decides how much supply reaches the secondary market, which is why a private capital advisory (PCA) banker reads an LP's allocation position before its fund list.
Strategic Asset Allocation: Private Equity Targets and Policy Ranges
A pension's strategic asset allocation (SAA) sets a long-term weight for each asset class, and a policy range around each weight tells staff when to act. For listed assets the range is a trading instruction; for private equity it is closer to a tolerance, because a fund interest cannot be sold with a phone call.
How a Target and Band Work in Practice
CalSTRS' investment policy statement of May 2025 sets private equity at a 14% long-term target with a band of five percentage points either side, adopted in January 2024. It concedes that appraisal-based valuations and illiquidity can make breaching a private-asset range unavoidable, in which case a rebalancing plan goes to the investment committee, and it invests the uncalled portion of the private equity target in other asset classes until the calls arrive.
Policy weights vary by institution and change over time, as CalSTRS, the California Public Employees' Retirement System (CalPERS), and the Dutch pension fund ABP show:
| Investor | Private equity policy | Set or announced |
|---|---|---|
| CalSTRS (US public pension) | 14% target, range 9% to 19% | Adopted January 2024 |
| CalPERS (US public pension) | Target raised from 13% to 17%, range 12% to 22% | March 2024 |
| CalPERS, from July 1, 2026 | No asset-class targets; judged against a 75% equity, 25% bond reference portfolio | Adopted November 2025 |
| ABP (Dutch civil servants' pension) | Exposure rising from 6% to 8% in a new strategic mix | Announced February 2025 |
ABP's change is driven by structure rather than markets: it linked its new mix to the Netherlands' move to a new pension system, in which the strict buffer requirements of the old system disappear. The same logic governs pension real estate allocations.
From Fixed Targets to a Total Portfolio Approach
CalPERS shows both ways to loosen the constraint. In March 2024 it raised its private equity target from 13% to 17%, keeping a band of five points either side. Then in November 2025 its board adopted a total portfolio approach (TPA), which from July 1, 2026 replaces the strategic asset allocation with a single reference portfolio of 75% equities and 25% bonds and no specific asset-class targets. Without a fixed private equity target, there is no fixed ceiling to breach.
Commitment Pacing: Why LPs Commit More Than Their Target
The Gap Between Commitments and Invested Capital
A commitment is never fully invested at one time. Capital is called over the investment period, and exits begin returning money before the last calls arrive, so a buyout fund's NAV tends to peak around its fifth year, as the fund lifecycle article sets out. Commonfund observed in a 2019 note on building a private capital allocation that investors rarely have more than two-thirds of their commitment at work at any one time, so reaching a 10% allocation may take commitments closer to 15%.
- Commitment Pacing
The plan an institutional investor uses to decide how much to commit to private funds each year, and in which vintages, so that projected NAV reaches and then holds its target allocation. It rests on forecasts of capital calls, distributions, and fund returns, and is usually revised annually.
The Takahashi-Alexander Model
The classic pacing framework came from the Yale University Investments Office, where Dean Takahashi and Seth Alexander described it in a January 2001 paper, "Illiquid Alternative Asset Fund Modeling," later published in the Journal of Portfolio Management. It replaced a venture capital rule of thumb, roughly 50 cents in each of a series of funds per dollar of ongoing exposure, with a few assumptions per fund:
- Contributions as a rate applied to the commitment not yet called.
- Distributions as a share of NAV that rises with fund age, shaped by a fund life and a "bow" parameter controlling how quickly realizations build.
- Growth, an assumed net return applied to NAV each year.
Summing existing funds and planned commitments projects the LP's future NAV, calls, and distributions; the pacing decision is the annual commitment that keeps projected NAV near target. The model's value lies less in precision than in showing what happens when distributions slow.
Over-Commitment Ratios
That forecast justifies committing more than the target, and the size of the excess is expressed as a ratio.
- Over-Commitment Ratio
A measure of how far an LP's private fund commitments exceed its target allocation, commonly calculated as NAV plus unfunded commitments divided by the target in dollars. A ratio of 1.5x means total exposure, invested and promised, is one and a half times the intended holding. Definitions vary by institution.
Take an illustrative $100 billion pension with a 20% target: $20 billion of private equity NAV plus $10 billion of unfunded commitments gives total exposure of $30 billion, a ratio of 1.5x. The extra third keeps NAV at target as older funds distribute. Every capital call must still be paid on the GP's schedule, a mechanism covered in capital calls, distributions, and the LP cash flow problem.
The Denominator Effect: A Higher Share Without a New Commitment
The same pension can move over target without any transaction at all.
- Denominator Effect
The rise in a private asset class's share of an investor's portfolio when the value of the rest of the portfolio falls faster than private marks. The private holdings have not grown; the total they are divided by has shrunk.
Take it through a sell-off in which its public and other assets lose 24% while private equity marks fall 4%:
| Illustrative pension | Before | After the sell-off |
|---|---|---|
| Public and other assets | $80bn | $60.8bn (down 24%) |
| Private equity NAV | $20bn | $19.2bn (down 4%) |
| Total portfolio | $100bn | $80bn |
| Private equity share | 20% | 24% |
| 20% target in dollars | $20bn | $16bn |
| Unfunded commitments | $10bn | $10bn |
| Over-commitment ratio | 1.5x | about 1.83x |
The pension has signed nothing new, yet it sits four points and $3.2 billion above target: inside a band of five points either side, outside a band of three. Its unfunded commitments are unchanged, so each call pushes the share higher unless distributions offset it, and the ratio against the smaller target rises to about 1.83x. Part of the gap reflects lagging private marks, the smoothing examined in how fund NAV is set, and closes on its own as marks catch down.
2022: The Textbook Episode
Coller Capital's Global Private Equity Barometer for Winter 2022-23, published in December 2022, found that 42% of LPs expected the denominator effect to slow the pace of their private equity commitments over the next one to two years, and two-thirds of large LPs (over $20 billion of assets) and public pension funds said it was already a factor in their slower pace.
2025: Over Target From the Numerator
After two strong years for public equities, overallocation persisted into 2025. S&P Global Market Intelligence data on 298 pension funds found 174, about 58%, overallocated to private equity in the first quarter of 2025, as reported by Banking Exchange. CalSTRS itself held 14.6% in private equity at the end of 2025 against its 14% target, in a year when its performance summary recorded public equity markets up 22.8%. Here much of the pressure came from the numerator: slow exits kept NAV invested and distributions thin.
How LPs Respond: Slower Pacing, Wider Ranges, Secondary Sales
An overweight LP has three levers, and they differ in speed and cost:
- Slowing new commitments: the cheapest response, since it involves no sale and reverses easily, but it leaves gaps in later vintage years and squeezes managers raising funds.
- Raising or widening the target: a board decision that gives the program more room, as CalPERS's higher target and later move away from fixed targets did, although CalPERS attributed its 2024 increase to strong private equity returns.
- Selling fund interests: the only lever that cuts NAV and unfunded commitments at once, because the buyer takes both, at the cost of accepting a price that is usually below NAV.
The first lever's cost lands on managers, a pressure traced in first-time funds and the 2025 fundraising squeeze. The third lever is where allocation arithmetic becomes a PCA mandate, though overallocation is only one reason LPs sell; why LPs and GPs need liquidity sets out the others.
For an advisor, the trigger for an LP sale is often written in a board document rather than set by the market. A wide band, an annual pacing plan, or a total portfolio approach with no private equity ceiling can absorb an overweight that a narrow band turns into a sale process. Two pensions hit by the same shock can therefore behave differently, and the reason is usually written in advance in each one's investment policy statement.
