The Distribution WaterfallNil Masferrer JiménezAugust 26, 2026

A Worked 8% Pref / 70-30 / 50-50 Waterfall, Line by Line

Descriptions of waterfalls are hard to hold in your head. Arithmetic is not. This is one deal, computed twice, with nothing skipped.

Descriptions of waterfalls are difficult to hold in the head. Arithmetic is not. What follows is one hypothetical deal, computed through five tiers, with every intermediate figure shown.

The deal¶

Total distributable cash: $150,000 × 5 + $3,600,000 = $4,350,000.

The accrued preferred return over five years, compounding at 8% on $2,500,000, is $2,500,000 × (1.08⁵ − 1) = $1,173,320.

The five tiers in order

Paid in order. Each tier fills completely before the next receives anything.

1Return of capitalUntil the $2,500,000 of LP equity is repaid100% LP
2Preferred returnUntil the $1,173,320 of accrued preferred is paid100% LP
3GP catch-upUntil the GP holds 30% of profit distributed so far100% GP
4First splitUntil the LPs reach a 15% internal rate of return70 / 30
5Residual splitEverything above that50 / 50

Hypothetical. Each tier fills completely before the next receives anything.

Running it¶

Tier 1 — return of capital. $4,350,000 available. $2,500,000 goes to the limited partners. Remaining: $1,850,000.

Tier 2 — preferred return. $1,173,320 goes to the limited partners. Remaining: $676,680. Profit distributed so far: $1,173,320, all of it to the limited partners.

Tier 3 — catch-up. For the sponsor to hold 30% of profit distributed, total profit must reach $1,173,320 / 0.70 = $1,676,172, of which the sponsor's share is $502,852. That is the size of the tier. $502,852 goes to the sponsor. Remaining: $173,828.

Tier 4 — 70/30 to a 15% IRR. A 15% internal rate of return over five years implies an equity multiple of 1.15⁵ = 2.01136, so the limited partner target is $2,500,000 × 2.01136 = $5,028,393. The limited partners hold $3,673,320 so far, leaving $1,355,073 of target. At a 70% share, the tier would need $1,935,818 to close that gap — far more than the $173,828 remaining. So the whole remainder runs through this tier. Limited partners take 70%: $121,680. Sponsor takes 30%: $52,148. Remaining: $0.

Tier 5 — residual. Nothing reaches it.

Tier	What it pays	Split	To LPs	To GP
1	Return of capital	100% LP	$2,500,000	$0
2	Preferred return, 8% compounding	100% LP	$1,173,320	$0
3	GP catch-up	100% GP	$0	$502,852
4	Split to a 15% LP IRR	70 / 30	$121,680	$52,148
5	Residual split	50 / 50	$0	$0
Total			$3,795,000	$555,000

Hypothetical. The limited partners receive a 1.52x multiple on their $2,500,000. The sponsor receives $555,000, which is 30% of the $1,850,000 of profit — exactly what the agreement promised, arrived at through four tiers rather than one split.

The same deal, sold two years earlier¶

Now change one thing. The property sells in year three for the same $3,600,000 of distributable proceeds, with three years of operating distributions instead of five.

Total distributable: $150,000 × 3 + $3,600,000 = $4,050,000. The accrued preferred is smaller: $2,500,000 × (1.08³ − 1) = $649,280.

Tier	Split	To LPs	To GP
1 — Return of capital	100% LP	$2,500,000	$0
2 — Preferred return	100% LP	$649,280	$0
3 — GP catch-up	100% GP	$0	$278,263
4 — Split to a 15% LP IRR	70 / 30	$435,720	$186,737
5 — Residual split	50 / 50	$0	$0
Total		$3,585,000	$465,000

Hypothetical. Same property, same sale price, two fewer years. The limited partner multiple falls from 1.52x to 1.43x, but the internal rate of return rises because the money came back sooner.

What changes if the preferred return is not cumulative¶

The example above assumes a cumulative, compounding preferred return, which is the most investor-favorable of the common forms. Changing that one assumption changes the result substantially, and it is a one-word change in an agreement.

Hypothetical example

The five-year case with a non-cumulative preferred

Same deal. $2,500,000 of LP equity, $150,000 of operating distributions each year, $3,600,000 at sale, a full catch-up to 30%, then 70/30.

With an 8% preferred return, the annual accrual is $200,000. Operating distributions of $150,000 fall short by $50,000 every year.

Cumulative and compounding, as computed above: those shortfalls accrue and compound, and by the exit the accrued preferred obligation is $1,173,320. The sponsor's total promote is $555,000.

Non-cumulative: each year's $50,000 shortfall is forfeited when the year closes. At the exit, the preferred return has been satisfied by the operating distributions already paid, so no accrued balance stands between the sponsor and the split.

The $1,173,320 tier disappears. The catch-up recalculates against a much smaller base of distributed profit. The consequence is that a substantially larger share of the sale proceeds reaches the split tiers, so the sponsor's promote rises while the limited partners' total falls — from one word in a definition, on a deal whose marketing described an 8% preferred return either way.

Illustrative arithmetic on invented figures. Not a projection, not an offering, and not a representation about any real transaction.

That comparison is the practical reason the definition matters more than the rate, and it is why this site keeps returning to the definitions article of an agreement rather than to its summary page.

What the comparison shows¶

Three things, none of which is visible from the sentence "8% pref, 70/30 above it."

The catch-up is the largest single payment to the sponsor in both runs. In the five-year case it is $502,852 out of $555,000 — over ninety percent of the promote. That tier is absent from every short description of a waterfall.

Total dollars and rate of return move in opposite directions. The shorter hold returns less money and a better annualized rate. Which one the sponsor is paid on is set by the hurdle measure, and it determines what the sponsor will do when an offer arrives in year three.

The compounding preferred return does real work. Over five years it accrues $1,173,320 rather than the $1,000,000 that simple accrual would produce. That $173,320 is money that reaches the limited partners before the sponsor participates. See cumulative and compounding preferred returns.

The worked model page runs the same waterfall across five different exits, including two where the sponsor earns nothing at all, with each tier printed.

Everything above is arithmetic on invented figures. Its purpose is not to suggest what a deal should return but to make the mechanism visible, so that the same tiers can be recognized in an agreement that describes them in five pages of defined terms.

Primary sources

Every factual claim above is traceable to a filing, a rule or an agency publication. These are the ones this article relies on.

SEC, private placements under Rule 506(b)sec.gov
Investor.gov, private placements explainedinvestor.gov
Legal Information Institute, 26 US Code 704 on a partner's distributive sharelaw.cornell.edu
IRS, Partner's Instructions for Schedule K-1 (Form 1065)irs.gov

Questions readers ask

Are these numbers from a real deal?

No. Every figure here is invented and chosen to make the arithmetic legible. It is not a projection, not an offering, and not a representation about any real transaction or sponsor.

Why compute the same deal twice?

Because the comparison is the point. Two runs differing only in the exit year show how much of a sponsor's promote comes from timing rather than from the property.

Is the 15% IRR tier computed exactly?

It is approximated by the equity multiple a 15% internal rate of return implies over the stated hold, which avoids an iterative solve while producing the right shape. A real agreement would require the iterative calculation, and a real model would perform it.

What is the most important line in these tables?

The catch-up row. It is the tier that pays the sponsor while the limited partners receive nothing, and it is the one absent from every one-sentence description of a waterfall.

Where can I change the assumptions?

The interactive version is on the worked model page, which runs the same waterfall across five exit scenarios with the same arithmetic printed.