Glossary · Portfolio construction

Commitment Pacing

Also called: pacing plan

Commitment pacing is the plan for how much new capital an investor commits to private funds each year, sized with a cash-flow model so that invested NAV reaches and holds the target allocation without straining liquidity.

Publisher: Altss LLCPublished 2026-01-06Content modified 2026-10-01

ALTSS-PORT-010

A private-fund commitment is a promise, not an investment. The manager calls the money over several years and returns it over several more, so at no point is the whole commitment invested. To keep, say, 15% of a portfolio in private equity, an investor has to keep committing every year, in amounts that depend on how fast its funds call and return capital. Pacing is the arithmetic and the discipline behind those annual commitment budgets.

Formulas

Takahashi–Alexander model: contributions

C_t = RC_t × (CC − PIC_t)

C_t: capital called in year t
RC_t: rate of contribution in year t (a share of the remaining unfunded commitment)
CC: capital commitment
PIC_t: paid-in capital before year t (sum of earlier contributions)

Annual steps with year-end flows. The published Yale base case for venture capital uses RC of 25% in year 1, 33.3% in year 2 and 50% thereafter.

Takahashi–Alexander model: distributions

D_t = RD_t × NAV_t−1 × (1 + G), where RD_t = max[Y, (t/L)^B]

D_t: distributions in year t
RD_t: rate of distribution in year t
Y: yield: minimum distribution rate (relevant for income-producing assets; zero for venture capital)
L: fund life in years; at t = L all remaining value is distributed
B: bow: shape of the distribution curve; a higher bow delays distributions
G: annual growth rate of NAV, net of fees

Deterministic: every fund with the same inputs produces the same path.

Takahashi–Alexander model: NAV

NAV_t = NAV_t−1 × (1 + G) + C_t − D_t

NAV_t: net asset value at the end of year t; NAV₀ = 0 for a new commitment

Because NAV compounds at G and is either distributed or carried forward, the IRR of the projected flows equals G by construction. Other pacing models simulate cash flows stochastically or fit them to historical vintage data.

Why pacing is needed

Commitments turn into exposure slowly and then decline as funds return capital. An investor that commits its full target in one year will be under-invested for several years and then over-exposed to a single vintage year. One that commits only when NAV is below target will commit in bursts. Pacing spreads commitments across years so that NAV holds near target, vintages are diversified, and capital calls can be met from expected distributions and the liquid portfolio.

What a pacing plan contains

A plan typically projects, year by year: calls, distributions and NAV for each existing fund (from its age, unfunded amount and strategy); planned new commitments by strategy; growth of the total portfolio, net of spending or benefit payments; and the resulting private allocation against target and range. Outputs include the commitment budget, the share of capacity reserved for re-ups versus new managers, the unfunded ratio, and net cash flow under a base case and stress cases (slower distributions, faster calls, a valuation shock, a public-market fall). Cash-flow projection is the engine (cash-flow forecasting); pacing is the commitment decision built on it.

The Takahashi–Alexander model: assumptions and limits

The Takahashi–Alexander model, published by two Yale Investments Office staff (Journal of Portfolio Management, 2002), is a simple, openly documented pacing model. It uses six inputs and three equations (see Formula) and can be re-based each year on a fund's actual calls, distributions and NAV.

Its limits follow from its simplicity:

It is deterministic. It projects one average path and says nothing about dispersion between funds or the chance of a bad outcome.
Growth is a constant input, so the projected IRR equals G by construction; returns do not respond to markets.
Calls and distributions are not linked to market conditions. In practice distributions slow and NAVs fall together in downturns while calls continue, which is when liquidity is tightest.
Contributions never fully stop; the authors note the residual amounts are insignificant at reasonable rates.
Recallable distributions, recycling, fee timing and subscription lines are not modelled separately.
Parameters are judgement calls; the authors describe their inputs as representative mathematical factors, not literal fund terms.

Practitioners therefore run the model with strategy-specific parameters, add stress scenarios, and replace parameters with actual fund data as funds age.

Overcommitment

Because no fund is fully invested at once and early funds return capital while later ones call it, reaching a NAV target requires total commitments outstanding above the target. Committing above target on purpose is overcommitment. The steady-state example above implies NAV plus unfunded of about 1.26 times target. The risk is the downturn case: if distributions stop while calls continue, the liquid portfolio must fund the gap.

Pacing through a downturn

When public markets fall, the private share of the portfolio rises (the denominator effect) and distributions slow. Common responses are to cut or defer the next year's budget, keep re-ups with core managers while pausing new relationships, or sell older interests in the secondary market at a discount to NAV. Stopping commitments entirely leaves a gap in vintage coverage that shows up years later as a dip in NAV and distributions.

Commitment pacing versus GP deployment pace

Commitment pacing is the LP's decision about how much to commit each year. Deployment pace is the GP's rate of investing a fund's capital. Faster GP deployment shortens the time to fund exposure and pulls the next fundraise forward, which pressures LP budgets.

Worked examples

Illustrative $100m commitment through the model

Using the Yale venture base case (growth G of 13%, fund life L of 12 years, rate of contribution (RC) of 25%, 33.3% and then 50%, bow B of 2.5, yield Y of 0%), a $100m commitment calls $25.0m in each of the first three years, $12.5m in year 4 and smaller amounts after that, $99.95m in total. NAV peaks at $106.8m in year 5. Net cash flow to the investor is negative in years 1 to 4 and positive from year 5. Distributions total $188.0m (1.88x the commitment), and the IRR of the projected flows is 13.0%, the growth input.

Sizing the annual budget for a 15% target in steady state

A $10bn plan targets 15% in private equity, or $1.5bn of NAV. Add up the year-end NAV of the single commitment above over its twelve years: $677.6m per $100m committed. If the plan commits the same amount every year, in steady state it holds one vintage of each age, so each $1 of annual commitment supports $6.78 of NAV. The annual budget is $1.5bn ÷ 6.78 ≈ $221m, about 2.2% of the plan a year. At that pace unfunded commitments settle near $387m, so NAV plus unfunded is about $1.89bn, 1.26 times the NAV target. The calculation assumes the plan does not grow; a growing plan needs larger commitments.

Sensitivity to the bow factor

Change only the bow. At B = 3.5 distributions come later, NAV peaks at $127.3m in year 6, total distributions rise to $214.0m and the NAV a $1 annual commitment supports rises to $8.77, so the budget for the same $1.5bn target falls to about $171m a year. At B = 1.5 distributions come earlier, the NAV supported falls to $4.33 per $1 and the budget rises to about $347m. A single shape parameter roughly doubles the answer, which is why pacing plans are run as ranges and stress cases.

Examples are illustrative; figures are not market data.

Not the same as

Deployment Pace: Deployment pace is how fast a GP invests a fund; commitment pacing is how much an LP commits across funds each year.
Overcommitment: Overcommitment is a choice within a pacing plan: committing more than the NAV target because commitments are never fully invested at once.
Private Markets Cash Flow Forecasting: Forecasting projects calls, distributions and NAV; pacing uses those projections to set commitment budgets.

Common mistakes

Sizing the annual commitment as target percentage times portfolio value. That ignores how much of each commitment is ever invested at one time.
Assuming distributions will fund calls in a downturn, when distributions are most likely to slow.
Using one set of model parameters for venture, buyout, credit and real assets.
Ignoring growth of the total portfolio and outflows such as benefit payments or spending, which move the target in dollars.
Stopping commitments after a market fall and creating a vintage gap.
Treating model output as a forecast with decimal precision rather than a central case with a range.

Edge cases

Secondary purchases add NAV immediately and return capital sooner, so they change the shape of the pacing curve.
Co-investments are usually funded at closing rather than called over years.
Recallable distributions can turn a past distribution back into a future call.
Evergreen funds invest subscriptions quickly but may restrict redemptions, so they need a separate liquidity assumption.

Questions

How much should an investor commit to private equity each year?

It depends on the NAV target, the portfolio's size and growth, and how fast the chosen strategies call and return capital. In the illustrative steady-state example, a 15% target in a $10 billion plan needs about $221 million a year, but plausible changes in one model parameter move that from about $171 million to $347 million.

What is the Takahashi–Alexander model?

A deterministic cash-flow model published by Yale Investments Office staff in 2002. It projects a fund's calls, distributions and NAV from six inputs and can be re-based each year on actual fund data.

Sources

Illiquid Alternative Asset Fund Modeling. Dean Takahashi; Seth Alexander, The Journal of Portfolio Management, Vol. 28(2), pp. 90-100, Winter 2002 (Crossref print date 2002-01-31); Yale University Investments Office working paper dated January 2001. Status: Published (journal paywalled) (checked 2026-10-01). Model inputs and equations; Yale venture base case; footnote on residual contributions; Exhibit 1 — supports: Model equations, inputs, base-case parameters and stated simplifications

8 terms

Concept record

Concept ID: ALTSS-PORT-010
Classification: Portfolio construction
Topics: Portfolio construction
Version: 2.0.0
Last reviewed: 2026-10-01
Structured data: JSON