A waterfall model answers the question every multifamily partnership eventually has to face: when cash comes out of the deal, who gets it, and in what order? The structure is tier based. Money flows through a structured sequence of distribution tiers, and each tier has to be satisfied before the next one opens. For general partners and limited partners alike, the model is the place where the economics of the deal stop being a handshake and become arithmetic.
That arithmetic is also where investment committee conversations tend to go sideways. A waterfall with annual periods and IRR or equity multiple hurdles produces a very different LP outcome than a structure with the same headline terms but a different tier order or a different hurdle definition. Reviewing the model carefully is the only reliable way to know which one you actually hold.
What a multifamily waterfall model calculates
A waterfall model outlines how cash flow from an investment is distributed to equity investors in the capital stack. In a real estate fund context, that means a tier based model for measuring the proper distribution of proceeds between the GP and the LPs. The model does not decide the terms. It applies them, period by period, to the cash the property or portfolio actually generates.
Investors need to see how a cash flow is being distributed among equity investors, because the answer changes as the deal performs. In a weak year, distributions may stop before any promote is earned. In a strong year, cash can clear several thresholds and split differently at each one. A waterfall model lays out those thresholds, usually expressed as IRR hurdles or equity multiple hurdles, and shows what each partner receives under each outcome. A waterfall distribution structure is also a practical way to set investment expectations before capital is called, since both sides can see the sequence they are agreeing to.
The tier structure, explained
Every waterfall is a sequence. Cash enters at the top, and each tier is tested in order until the distributable amount is exhausted. The specific tiers, the thresholds that trigger them, and the split percentages at each level all come from the partnership agreement. The model's job is to reproduce those terms faithfully rather than to improve on them.
Preferred return tier
The preferred return is the first economic threshold in most structures. Distributable cash is tested against it before any promote split applies, which is why the preferred return is often described as the LP's priority position. The rate, whether it compounds, and how it interacts with returned capital are all defined in the deal documents, not in the spreadsheet. If the model's preferred return differs from the operating agreement, every tier above it is wrong by extension.
Promote split at the hurdle
Above the preferred return, the model applies the promote split. This is where the GP begins sharing in distributions beyond a pro rata return, and where the split percentage typically changes at each hurdle level. A real estate equity waterfall model built with IRR or equity multiple hurdles will carry one split for the first threshold, another for the next, and so on. Because these splits compound across a deal's life, a small difference in the tier definitions produces a large difference in total GP and LP proceeds.
Additional hurdle levels
Larger deals frequently carry several hurdle levels rather than one. Each level adds a threshold and a new split, and each one has to be cleared before the next applies. When you review a model, count the tiers and match them one by one against the agreement. Missing or reordered tiers are among the most common and most expensive errors in a partnership model.
IRR hurdles versus equity multiple hurdles
Both hurdle types measure performance, but they measure different things and they are not interchangeable. The table below summarizes the practical difference for review purposes.
| Hurdle type | What it measures | What changes when it is met |
|---|---|---|
| IRR hurdle | Return measured across annual periods, so the timing of each distribution matters | The split shifts at each threshold based on the annual periods the model runs |
| Equity multiple hurdle | Total distributions relative to equity invested, without weighting for timing | The split shifts once cumulative distributions cross the multiple threshold |
A model can use either structure, and some use both. The choice affects how a deal behaves when cash arrives early versus late. An IRR hurdle rewards faster return of capital, while an equity multiple hurdle responds to total dollars returned. If your model and your partner's model use different hurdle types, you are not comparing the same deal.
Inputs that drive the waterfall
Waterfall logic sits on top of a cash flow forecast, so the quality of the output depends entirely on the inputs beneath it. The core inputs to check include:
- Distributable cash flow by period, which comes from property level revenue and expenses rather than from the waterfall itself.
- Equity contributions and the timing of each capital call, since IRR based tiers are sensitive to when money goes in.
- Tier order and the threshold that triggers each tier.
- Hurdle type, whether IRR or equity multiple, and the measurement period used.
- Promote split percentages at every tier, not just the first.
- Treatment of returned capital and any preferred return accrual conventions defined in the agreement.
Because the waterfall consumes the property level forecast, errors upstream propagate upward. A rent roll that double counts a unit or a trailing twelve month statement with a misclassified expense line changes net cash flow, which changes the period in which a hurdle is cleared, which changes the promote. This is the practical link between data quality and partnership economics.
How to review a multifamily waterfall model
A disciplined review follows the same sequence every time. Work through the model in this order and document each step for the investment committee file.
- Read the distribution section of the partnership agreement and list every tier, threshold, and split on paper before opening the model.
- Confirm the model's tier order matches that list exactly.
- Verify the hurdle type and the measurement period, whether annual periods, IRR based, or equity multiple based.
- Tie the equity contribution schedule to the actual capital calls and their dates.
- Trace distributable cash flow back to the property level forecast and confirm it reconciles.
- Test the tier mechanics with a simple scenario, then confirm the model produces the expected split at each threshold.
- Stress the model across a range of exit assumptions to see where the promote turns on and off.
- Document every assumption that was not specified in the agreement and flag it as a term to confirm.
The last step matters most. Where a model makes an assumption the documents do not address, that assumption is a negotiating item, not a modeling detail. Readers should confirm anything ambiguous with the deal's legal and tax advisors before relying on the output.
Where waterfall review breaks down in practice
Most review failures are not caused by sophisticated modeling mistakes. They come from mismatched sources. The person reviewing the waterfall is often working from a partner model built months earlier, while the property data has moved on. Cash flow assumptions embedded in the model may not match the current trailing twelve month statement, and the rent roll behind those assumptions may never be checked at all.
Timing assumptions are the second common failure point. IRR based tiers are sensitive to the annual periods used and to when capital is called and returned. Shifting a single distribution by a period can move a deal across a hurdle without any change in total dollars. Any review that does not explicitly test timing is incomplete.
The third failure point is opacity. When the waterfall lives in a heavily linked spreadsheet with hardcoded overrides, reviewers cannot trace a distribution back to a tier or an input. The model may be correct, but it is not verifiable, and an unverifiable model creates risk at the investment committee stage.
Connecting the waterfall to property level data
Waterfall accuracy starts with clean property data. Rent roll parsing and T12 financial statement analysis feed the cash flow forecast that drives every tier, and manual rekeying of those files is where most errors enter the chain. Coastwise Analytics automates the parsing of rent rolls and financial statements from property management systems including Yardi, RealPage, and Entrata, and standardizes the output into structured insights, performance memos, and underwriting proformas.
Because the underwriting proforma engine includes waterfalls alongside sensitivities, the tier logic and the property level forecast stay in one structured environment rather than in two disconnected files. That matters for review. When a reviewer changes an exit assumption, the same structured cash flow runs through the tiers, and the promote outcome is traceable back to the input that moved it. For teams preparing investment committee memos, that traceability is the difference between a defensible number and an asserted one.
What to put in front of the investment committee
A complete waterfall package includes more than the model itself. Bring the tier summary in plain language, the hurdle type and measurement period, the equity contribution schedule, the property level cash flow forecast with its source data, and a sensitivity view showing LP and GP proceeds across a range of outcomes. Note which assumptions came from the agreement and which were modeled judgment.
That package lets the committee evaluate partnership economics and property performance together instead of treating them as separate questions. It also makes the review repeatable, which is what allows a firm to compare one deal against the next on consistent terms.
Frequently Asked Questions
What is a waterfall model in real estate?
A waterfall model is a tier based structure that shows how cash flow from an investment is distributed among equity investors in the capital stack. In multifamily and fund deals, it measures the proper distribution of proceeds between the GP and the LPs, applying preferred returns, hurdle thresholds, and promote splits in a defined sequence. The model applies those terms rather than setting them.
How does a waterfall distribute cash between the GP and the LPs?
Cash enters at the top of the tier structure and is tested against each threshold in order. Distributions flow to the LPs according to the preferred return and upper tiers, while the GP receives the promote split once the applicable hurdles are met. The tier order, thresholds, and percentages are defined in the partnership agreement, so the model must match that document exactly.
What is the difference between an IRR hurdle and an equity multiple hurdle?
An IRR hurdle measures return across annual periods, so the timing of each distribution changes the result. An equity multiple hurdle compares total distributions to equity invested without weighting for timing. Both are common in real estate equity waterfalls, and some structures use them together. Comparing a model built on one hurdle type against a model built on the other is not a like for like comparison.
What are the downsides of a waterfall model?
The main weakness is sensitivity rather than logic. IRR based tiers can shift when a distribution moves by a single period, and a small change in tier definitions can produce a large change in total proceeds. Waterfalls also inherit every error in the property level forecast beneath them. Building the tier review on standardized rent roll and financial statement data reduces, but does not remove, that exposure.
How should a team review a waterfall model before closing?
List every tier, threshold, and split from the partnership agreement first, then confirm the model reproduces that list in order. Verify the hurdle type and measurement period, tie equity contributions to actual capital calls, and trace distributable cash back to the property forecast. Run sensitivities across exit assumptions, and flag any modeled assumption the documents do not address for advisors to confirm.
