In this model a GP that puts in 10.0% of the equity, $8,945,990, receives 31.4% of the deal’s $139,111,216 profit. The three LP classes put in the other 90.0% and earn a 19.6% net IRR. The GP earns 38.8%, and the deal earns 22.7% before any promote is taken out.
You work in acquisitions or asset management at an institutional real estate investor, and a joint-venture term sheet has reached your desk. A real estate equity waterfall turns one stream of deal cash flow into what each party keeps. We asked Compound to build one on a real property’s cash flows from public data. It ran on November 15, 2026 and returned one Excel workbook and a summary.
The joint venture in this workbook is hypothetical. The investor classes, their commitments, the preferred returns, the hurdles and the splits are assumptions, some set by the request and some by Compound as analyst assumptions, laid over one real hotel’s filed cash flows. Nothing on this page describes who holds the hotel’s equity or on what terms. The purchase price is also an assumption that the request set: the latest appraised value in the filing, from an appraisal dated December 21, 2017. The returns are the output of those assumptions. They are neither a forecast for this property nor advice to buy or sell it.
The returns also lean on one financing assumption. A refinance at the end of year 1 pays out $37.5 million, which is 54.4% of the equity (our arithmetic). The section on the deal’s cash flows shows how that loan is sized.
What the workbook shows:
- The GP’s share of profit is 31.4%, above the 30% line the request asked to flag. The promote is $33,356,840, or 24.0% of profit. The rest is the profit on the GP’s own co-investment.
- The LP classes earn 19.3%, 19.8% and 20.3%. The class with the highest preferred return earns the most.
- The 50/50 tier pays nothing. The LP classes finish $3,251,040 short of the 20% IRR that opens it.
- No clawback is due. All of the promote is paid at the sale in year 7, after every class has its capital and preferred return.
- The first hurdle does not matter between 4% and 12%. With a 100% catch-up, LP net IRR and the promote are the same at every value in that range.
Each input on the Assumptions sheet carries a citation, the word “Request”, or the words “Analyst assumption” with a reason. The property, its four years of income and its loan come from CMBS loan-level filings and a prospectus on SEC EDGAR. Revenue growth is set from the 2026 guidance that five hotel REITs gave on their third-quarter earnings calls. The refinance is priced off the Daily Treasury Par Yield Curve, with a switch to Compound’s SOFR forward curve, which is built from the New York Fed’s SOFR and from swap trades.
“I am negotiating joint-venture terms with a sponsor and want a real estate equity waterfall model that shows what each investor class earns, run on a real property's cash flows. From the CMBS loan-level filings, pick one U.S. hotel property whose filings give revenue, operating expenses and net operating income for at least three reporting periods, along with appraised value and the loan balance, rate and maturity; hotel cash flow moves enough from year to year to test every tier. Use its latest appraised value as the purchase price and its existing loan as the debt. Build the deal cash flows: a year-0 capital call for the equity, then seven years of annual cash flow after debt service starting from the reported NOI, grown at a rate set from the revenue-per-available-room trends public hotel REITs report in their latest 10-K or earnings call, with one down year, and a sale in year 7 at an exit cap rate; if the loan matures inside the hold, refinance it off today's 10-year Treasury yield and the SOFR forward curve. Then build the waterfall: three LP classes and a GP, each with an active switch, a commitment and a preferred return rate; a whole-fund waterfall that returns capital, pays an annually compounding preferred return per class, runs a GP catch-up to a target share, then splits 80/20 above an 8% LP IRR, 70/30 above 15% and 50/50 above 20%, all as inputs; and a clawback test at exit, capped at the promote the GP has received. Lay it out as inputs, cash flows, returns by class and the clawback test. Report IRR and equity multiple for each class and for the deal, GP IRR, the promote in dollars and as a share of profit, the dollars paid in each tier each year, and whether a clawback applies and how much. Add checks that the class shares sum to 100%, that the hurdles rise, and that distributions equal the deal cash flow every year, and a sensitivity table of LP net IRR and GP promote to the exit cap rate and the first hurdle rate. Flag a clawback, any class that does not reach its preferred return, and a GP share of profit above 30%. Use the stored public data wherever it has the figure, and web search only for what it lacks. Put every input on an Assumptions sheet with one column that says where it came from: a cited source, or "analyst assumption" for a figure I would set myself, with the reason for the value you chose. Every output must be a live formula that traces back to the Assumptions sheet. Add a Sources sheet listing each data source, what it gave and its date. Deliver one Excel workbook and a short summary of the result, the flags, and what I would need from the seller to firm it up. Cite every sourced figure.”
That is the request as we sent it, without its closing line, which asks that no name or email be put in the workbook.
Which hotel’s cash flows does the waterfall run on?
Compound chose the Hilton Clearwater Beach Resort & Spa, a 418-room hotel at 400 Mandalay Avenue in Clearwater Beach, Florida, as a CMBS filing names it. The filing gives four fiscal years of revenue, expenses and net operating income, an appraised value of $199,000,000, and a 5.13% fixed-rate loan that matures on March 6, 2028.
| FY2022 | FY2023 | FY2024 | FY2025 | |
|---|---|---|---|---|
| Revenue | $60,925,230 | $58,323,487 | $52,846,879 | $52,523,688 |
| Operating expenses | $33,292,129 | $33,934,436 | $31,838,360 | $31,555,952 |
| Net operating income | $27,633,101 | $24,389,051 | $21,008,519 | $20,967,736 |
| Net cash flow | $25,196,091 | $22,056,111 | $18,894,644 | $18,866,788 |
| NOI margin | 45.4% | 41.8% | 39.8% | 39.9% |
| Revenue growth | (4.3%) | (9.4%) | (0.6%) |
Source: Compound workbook, Assumptions sheet, citing the CSAIL 2018-CX11 Form ABS-EE, Exhibit 102 reports of April 2023 to April 2026. The margin and growth rows are formulas in the workbook.
Net operating income fell in each of the three years, and by our arithmetic FY2025 is 24.1% below FY2022. The model starts from the FY2025 figures.
The loan terms come from the same filing: a 360-month amortization term and debt service on the whole loan of $8,694,923 a year, which the Assumptions sheet notes is a 2.41x DSCR on FY2025 net operating income.
The loan is split into pari passu notes in more than one trust, and Compound sized the whole loan from one note: its October 2026 balance of $8,581,655 in the CSAIL 2018-C14 filing, scaled by the note split in that trust’s prospectus of November 29, 2018. The whole loan works out to $114,136,016 in October 2026 and to $113,661,714 at the assumed close on December 31, 2026. A second estimate from the debt-service ratio gives $113,899,374, and the Checks sheet passes the two at 0.21% apart.
The price is where a reader should be most careful. At $199,000,000 the FY2025 net operating income is a 10.5% going-in cap rate, and the assumed loan is 57.1% of the price. An appraisal from 2017 is a starting point that a buyer replaces with a bid. Compound’s summary says so itself: it calls the going-in cap rate one that “flatters returns” and puts a current appraisal first on its list for the seller.
What are the joint venture’s terms, and who set them?
The request fixed the outline: three LP classes and a GP, splits of 80/20, 70/30 and 50/50 above LP IRRs of 8%, 15% and 20%, and a clawback capped at the promote the GP has received. Compound set the rest as analyst assumptions with a reason for each: the commitment shares, the preferred return rates and a 100% catch-up to a 20% share.
Equity is what the sources and uses leave. The price, closing costs of $2,985,000 and a loan assumption fee of $1,136,617 make total uses of $203,121,617. The assumed loan covers $113,661,714, and the equity required is $89,459,903.
Class A takes 50.0% of that, Class B 25.0%, Class C 15.0% and the GP co-invest 10.0%. The reason on the sheet is a 90/10 split between LPs and GP, “inside the 90/10 to 95/5 market range”. The preferred returns are 8.0%, 9.0% and 10.0% for the three LP classes and 8.0% for the GP’s co-invest, compounding annually.
Three points of structure decide how the tiers behave:
- The waterfall is whole-fund. Every class gets its capital back and its preferred return before any promote is paid.
- The GP’s co-invest ranks with the LP classes for capital and preferred return. The IRR hurdles are tested on the three LP classes together, without the co-invest.
- The catch-up sends 100% of the cash after the preferred return to the GP until its promote equals 20% of the profit paid out so far.
| B | C | D | E | F | H | |
|---|---|---|---|---|---|---|
| 83 | EQUITY CLASSES | |||||
| 84 | Class A LP | Class B LP | Class C LP | GP co-invest | ||
| 85 | Active (1 = on, 0 = off) | 1 | 1 | 1 | 1 | Switch per Request; analyst assumption: all classes on in the base case |
| 86 | Commitment share of equity | 50.0% | 25.0% | 15.0% | 10.0% | Analyst assumption: 90/10 LP/GP split, inside the 90/10 to 95/5 market range |
| 87 | Preferred return (annual, compounding) | 8.0% | 9.0% | 10.0% | 8.0% | Analyst assumption: 8% base, rising for junior-priced classes to test classes that miss their pref |
| 88 | LP class (1) or GP (0) | 1 | 1 | 1 | 0 | Structure: GP co-invest ranks pari passu with LPs for capital and pref, outside the LP IRR hurdles |
| 89 | Commitment | 44,729,952 | 22,364,976 | 13,418,985 | 8,945,990 | Computed: share x equity required at close |
| 91 | PROMOTE STRUCTURE (WHOLE-FUND) | |||||
| 92 | GP catch-up target share of profit | 20.0% | Analyst assumption: 20%, equal to the first promote split | |||
| 93 | Catch-up rate (share of tier cash to GP) | 100.0% | Analyst assumption: full (100%) catch-up | |||
| 94 | Hurdle 1: LP IRR | 8.0% | Request | |||
| 95 | Hurdle 2: LP IRR | 15.0% | Request | |||
| 96 | Hurdle 3: LP IRR | 20.0% | Request | |||
| 97 | GP split above hurdle 1 (80/20) | 20.0% | Request | |||
| 98 | GP split above hurdle 2 (70/30) | 30.0% | Request | |||
| 99 | GP split above hurdle 3 (50/50) | 50.0% | Request | |||
| 100 | Flag: GP share of profit above | 30.0% | Request | |||
| 101 | Clawback cap | Promote received | Request | |||
| 102 | Check and flag tolerance | 1 | Analyst assumption: $1 rounding tolerance for checks and flags |
This is a property-level joint venture with investor classes. A buyout fund’s carry uses the same tiers at fund level, across many deals, and the private equity returns page covers that case. In the bridge-to-perm multifamily model a waterfall is one tab of a deal model. Here it is the whole workbook.
Where does the cash come from?
Equity of $89,459,903 receives $228,571,119 over seven years, a 22.7% levered IRR and a 2.56x multiple. Two payments carry it: $48,675,737 in year 1, most of it from a refinance, and $149,032,508 in year 7, most of it from the sale.
Revenue grows 4.18% a year. Compound took the 2026 RevPAR guidance of Host Hotels & Resorts, DiamondRock Hospitality, Park Hotels & Resorts, Pebblebrook Hotel Trust and Apple Hospitality REIT from their third-quarter 2026 earnings calls, averaged the midpoints to 6.18%, and took off a haircut of 2.00%. The haircut is an analyst assumption, and the reason given is that 2026 guidance carries one-off World Cup demand and that this property’s revenue fell from FY2022 to FY2025. Expenses grow at 3.0% inflation, with 50.0% of them moving with revenue, and an FF&E reserve of 4.0% of revenue comes out below net operating income.
The request asked for one down year. Compound put it in year 3, after the refinance, “to stress debt service and tiers”, and sized it at the property’s worst reported year: revenue falls 9.4%, as it did from FY2023 to FY2024. Net operating income drops 18.5% that year to $18,873,075.
Net operating income: four reported years, then seven model years
Reported years from the CMBS filings. Model years grow revenue 4.18% a year with a 9.4% fall in year 3.
Compound workbook, Assumptions and Cash Flows sheets. Reported figures: CSAIL 2018-CX11 Form ABS-EE. The growth rate and the down year are analyst assumptions.
The existing loan matures in March 2028, inside the hold, so the model refinances it at the end of year 1. The new loan is 55.00% of a value set by capitalizing year-2 net operating income of $23,144,101 at 8.50%, which gives $149,755,946. Its rate is 7.40%: the 10-year Treasury par yield of 5.15% on November 13, 2026 plus a 2.25% spread, an analyst assumption. A switch on the Assumptions sheet reprices it as a floating loan at the 12-month SOFR forward for each year plus 3.00%.
The new loan repays the old balance of $110,729,327 and costs $1,497,559 to close. By our arithmetic, the refinance values the hotel at $272.3 million against the $199,000,000 price, leaves $37.5 million after the old loan and costs, and makes the year-1 distribution 54.4% of the equity. Debt service rises from $8,694,923 to $12,556,799 a year. The workbook computes the new payment on annual periods, which comes out a little above twelve monthly payments on the same terms. The DSCR is 2.53x in year 1, 1.84x in year 2 and 1.50x in the down year.
The sale is at the end of year 7 at an 8.50% cap rate on year-8 net operating income of $24,287,683. That is a gross price of $285,737,446 and, after 2.00% of costs, net proceeds of $280,022,697, from which the new loan’s $139,098,939 balance is repaid. The exit cap rate is an analyst assumption. Without debt the property returns 13.2%.
| B | C | D | E | F | G | H | I | J | |
|---|---|---|---|---|---|---|---|---|---|
| 3 | Hold year | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
| 4 | Period end | Dec-26 | Dec-27 | Dec-28 | Dec-29 | Dec-30 | Dec-31 | Dec-32 | Dec-33 |
| 11 | Net operating income ($) | 20,967,736 | 22,030,367 | 23,144,101 | 18,873,075 | 19,855,374 | 20,885,677 | 21,966,238 | 23,099,418 |
| 27 | New loan funding | 149,755,946 | - | - | - | - | - | - | |
| 36 | DSCR (NOI / total debt service) | 2.53x | 1.84x | 1.50x | 1.58x | 1.66x | 1.75x | 1.84x | |
| 56 | Net cash flow before debt service | 19,841,600 | 20,863,843 | 16,806,935 | 17,702,870 | 18,643,197 | 19,630,023 | 20,665,549 | |
| 57 | Existing loan debt service | (8,694,923) | - | - | - | - | - | - | |
| 58 | New loan debt service | - | (12,556,799) | (12,556,799) | (12,556,799) | (12,556,799) | (12,556,799) | (12,556,799) | |
| 59 | Existing loan repayment | (110,729,327) | - | - | - | - | - | - | |
| 60 | New loan funding, net of costs | 148,258,387 | - | - | - | - | - | - | |
| 61 | Net sale proceeds before debt | - | - | - | - | - | - | 280,022,697 | |
| 62 | New loan repayment at sale | - | - | - | - | - | - | (139,098,939) | |
| 63 | Levered cash flow to equity ($) | (89,459,903) | 48,675,737 | 8,307,044 | 4,250,136 | 5,146,071 | 6,086,398 | 7,073,224 | 149,032,508 |
| 66 | Levered IRR (deal) | 22.7% | |||||||
| 67 | Levered equity multiple (deal) | 2.56x | |||||||
| 68 | Unlevered IRR (property) | 13.2% |
How does the waterfall pay each tier, year by year?
Every dollar in years 1 to 6 returns capital, so no preferred return and no promote is paid before the sale. In year 7 the $149,032,508 runs through the tiers in one period: $9,921,292 of remaining capital, $29,553,779 of preferred return, a $7,388,445 catch-up, then $46,823,030 at 80/20 and $55,345,963 at 70/30.
Tier 1 returns capital to all four investors in proportion to what each is still owed.
Tier 2 pays the preferred return, which has compounded on the unreturned balance since the close. Each class has its own account. At the sale Class A is paid $13,473,061, Class B $7,898,655, Class C $5,487,450 and the GP’s co-invest $2,694,612.
Tier 3 exists for a class whose preferred return is below the first hurdle. It tops the LP classes up to an 8% IRR before the GP shares in anything. Here it pays nothing, because every class’s preferred return is at or above 8%.
Tier 4 is the catch-up. The Waterfall sheet computes the amount needed in one cell:
=MAX(0,$C$13*J79-J80)/(Assumptions!$C$93-$C$13*(1-Assumptions!$C$93))
The factor in cell C13 is the target share divided by one minus the target share, which is 0.25 at a 20% target. The catch-up is a quarter of the $29,553,779 preferred return, and after it the GP holds 20% of the profit paid so far.
Tier 5 splits cash 80/20 until the LP classes reach a 15% IRR. Investors receive $37,458,424 and the GP $9,364,606. Tier 6 splits what is left 70/30: $38,742,174 to investors and $16,603,789 to the GP. In the workbook “investors” includes the GP’s co-invest, so the LP classes receive 90.0% of each investor amount.
Tier 7, the 50/50 split, never opens. The Waterfall sheet keeps an account of what the LP classes still need to reach each hurdle, and the 20% account closes at $3,251,040. Compound’s summary adds that the tier starts paying at exit cap rates of about 8.0% or lower.
| B | D | E | F | G | H | I | J | K | |
|---|---|---|---|---|---|---|---|---|---|
| 3 | Hold year | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
| 100 | DISTRIBUTIONS BY TIER ($) | Total | |||||||
| 101 | Tier 1: return of capital (investors) | 48,675,737 | 8,307,044 | 4,250,136 | 5,146,071 | 6,086,398 | 7,073,224 | 9,921,292 | 89,459,903 |
| 102 | Tier 2: preferred return (investors) | - | - | - | - | - | - | 29,553,779 | 29,553,779 |
| 103 | Tier 3: top-up to hurdle 1 (investors) | - | - | - | - | - | - | - | - |
| 104 | Tier 4: catch-up to GP | - | - | - | - | - | - | 7,388,445 | 7,388,445 |
| 105 | Tier 4: catch-up to investors | - | - | - | - | - | - | - | - |
| 106 | Tier 5: 80/20 to investors | - | - | - | - | - | - | 37,458,424 | 37,458,424 |
| 107 | Tier 5: 80/20 to GP | - | - | - | - | - | - | 9,364,606 | 9,364,606 |
| 108 | Tier 6: 70/30 to investors | - | - | - | - | - | - | 38,742,174 | 38,742,174 |
| 109 | Tier 6: 70/30 to GP | - | - | - | - | - | - | 16,603,789 | 16,603,789 |
| 110 | Tier 7: 50/50 to investors | - | - | - | - | - | - | - | - |
| 111 | Tier 7: 50/50 to GP | - | - | - | - | - | - | - | - |
| 112 | Total distributions ($) | 48,675,737 | 8,307,044 | 4,250,136 | 5,146,071 | 6,086,398 | 7,073,224 | 149,032,508 | 228,571,119 |
| 113 | of which LP classes | 43,808,164 | 7,476,340 | 3,825,123 | 4,631,464 | 5,477,759 | 6,365,902 | 104,368,867 | 175,953,618 |
| 114 | of which GP (co-invest and promote) | 4,867,574 | 830,704 | 425,014 | 514,607 | 608,640 | 707,322 | 44,663,641 | 52,617,502 |
Dollars paid in each tier over the hold
Return of capital is paid in every year. Every other tier is paid at the sale in year 7.
Compound workbook, Waterfall sheet, column K. Investor amounts include the GP's 10% co-invest.
By our arithmetic, year 7 carries 65.2% of all the dollars distributed. A waterfall that pays this late puts the whole promote on one number, the sale price.
What does each class earn, and what does the GP keep?
The LP classes earn 19.3%, 19.8% and 20.3% in the order of their preferred returns, and 19.6% together at a 2.19x multiple. The GP earns 38.8% and 5.88x on its $8,945,990: a promote of $33,356,840 and $10,314,672 of profit on its co-invest.
The three LP classes share the tiers above the preferred return in proportion to their commitments, so the gap between them comes from the preferred return alone. Class C puts in $13,418,985 and takes out $30,336,525, a 2.26x multiple. Class A puts in $44,729,952 and takes out $96,303,312, a 2.15x multiple. By our arithmetic Class C earns 1.0 point of IRR more than Class A for the same place in the order of payment.
The GP’s co-invest has an 8.0% preferred return and earns the same 19.3% as Class A. Everything above that in the GP’s return is promote.
| B | C | D | E | F | G | H | I | |
|---|---|---|---|---|---|---|---|---|
| 121 | RETURNS BY CLASS | |||||||
| 122 | Class A LP | Class B LP | Class C LP | GP co-invest | LP classes | GP total | Deal | |
| 123 | Capital contributed | 44,729,952 | 22,364,976 | 13,418,985 | 8,945,990 | 80,513,913 | 8,945,990 | 89,459,903 |
| 124 | Distributions | 96,303,312 | 49,313,780 | 30,336,525 | 19,260,662 | 175,953,618 | 52,617,502 | 228,571,119 |
| 125 | Profit ($) | 51,573,360 | 26,948,804 | 16,917,540 | 10,314,672 | 95,439,705 | 43,671,512 | 139,111,216 |
| 126 | IRR | 19.3% | 19.8% | 20.3% | 19.3% | 19.6% | 38.8% | 22.7% |
| 127 | Equity multiple | 2.15x | 2.20x | 2.26x | 2.15x | 2.19x | 5.88x | 2.56x |
| 128 | Preferred return rate | 8.0% | 9.0% | 10.0% | 8.0% | |||
| 129 | Reached preferred return? | Yes | Yes | Yes | Yes | |||
| 131 | LP net IRR (Class A + B + C) | 19.6% | ||||||
| 132 | GP IRR (co-invest + promote) | 38.8% | ||||||
| 133 | GP promote ($) | 33,356,840 | ||||||
| 134 | Promote as share of deal profit | 24.0% | ||||||
| 135 | GP share of deal profit (co-invest + promote) | 31.4% | ||||||
| 136 | Flag: GP share of profit above threshold | FLAG | ||||||
| 137 | Classes short of preferred return | 0 | ||||||
| 138 | Clawback due at exit ($) | - | ||||||
| 139 | Flag: clawback applies | OK |
The request asked for a flag when the GP’s share of profit is above 30%. The flag is raised at 31.4%. The cell divides the GP’s whole profit, promote and co-invest together, by deal profit. The promote alone is 24.0%, which would pass. By our arithmetic the co-invest profit is 7.4% of deal profit. A term sheet that caps “the GP’s share” needs to say which of the two it means, and the workbook shows both.
IRR by investor
Each class from its own cash flows on the Waterfall sheet. The deal figure is before the promote.
Compound workbook, Waterfall sheet, returns by class. The classes and their terms are assumptions; see the Assumptions sheet.
By our arithmetic the promote costs the LP classes 3.1 points of IRR, the gap between the deal’s 22.7% and their 19.6%.
Does a clawback apply?
No clawback is due. The GP receives its whole $33,356,840 promote in year 7, after every class has its capital and preferred return, so the test at exit finds nothing to give back.
The Clawback sheet adds up the capital and preferred return that any class is still owed at exit and any shortfall of the LP classes against the first hurdle. It then checks whether the second and third hurdles were met at exit and, if one was not, counts the promote paid at that hurdle’s split as excess. Here the second hurdle is met and the third is not, and no promote was paid at 50/50, so the excess is nothing. The amount due is the smaller of the shortfall and the cap, which is the promote received.
| B | C | |
|---|---|---|
| 4 | CLAWBACK TEST AT EXIT | |
| 5 | Exit year | 7 |
| 6 | Promote received through exit | 33,356,840 |
| 7 | Unreturned capital at exit, all classes | - |
| 8 | Unpaid preferred return at exit, all classes | - |
| 9 | Class capital and pref shortfall | - |
| 10 | LP shortfall to hurdle 1 IRR at exit | - |
| 11 | Hurdle 1 shortfall, grossed up to all investors | - |
| 12 | Investor shortfall to restore | - |
| 13 | Hurdle 2 met at exit (1 = yes) | 1 |
| 14 | Hurdle 3 met at exit (1 = yes) | 0 |
| 15 | Excess promote paid at the 70/30 split | - |
| 16 | Excess promote paid at the 50/50 split | - |
| 17 | Clawback before cap | - |
| 18 | Cap: promote received | 33,356,840 |
| 19 | Clawback due ($) | - |
| 20 | Clawback applies? | No clawback |
| 21 | Promote net of clawback ($) | 33,356,840 |
| 22 | Net promote as share of deal profit | 24.0% |
A whole-fund waterfall with all of its promote paid at the sale leaves little for a clawback to do. The test earns its place when promote is paid before the end, for example when a large refinance returns all capital and preferred return early and later years fall short. Compound’s summary says it stress-tested the logic and that the clawback “does trigger and cap correctly when promote is paid early”.
On the Checks sheet, class shares sum to 100.0%, the hurdles rise, sources equal uses, and distributions equal the deal cash flow in every year. No check fails. One flag is raised, the GP’s share of profit. The other two that the request asked for, a clawback and a class short of its preferred return, are clear.
How much do the exit cap rate and the first hurdle move the result?
The exit cap rate moves everything: from 5.50% to 11.50%, LP net IRR runs from 26.2% to 12.9% and the promote from $108,694,590 to $13,212,363. The first hurdle moves nothing: at 4%, 6%, 8%, 10% and 12% the LP IRR and the promote are the same to the dollar.
Both tables are live Excel data tables on the Assumptions sheet, so they recalculate when an input changes. The tables change the exit cap rate at the sale. The cap rate that values the hotel for the refinance is a separate input and stays at 8.50%.
| B | C | D | E | F | G | |
|---|---|---|---|---|---|---|
| 105 | LP net IRR: exit cap rate (down) vs hurdle 1 (across) | |||||
| 106 | 19.6% | 4.0% | 6.0% | 8.0% | 10.0% | 12.0% |
| 107 | 5.50% | 26.2% | 26.2% | 26.2% | 26.2% | 26.2% |
| 108 | 6.25% | 24.3% | 24.3% | 24.3% | 24.3% | 24.3% |
| 109 | 7.00% | 22.6% | 22.6% | 22.6% | 22.6% | 22.6% |
| 110 | 7.75% | 21.1% | 21.1% | 21.1% | 21.1% | 21.1% |
| 111 | 8.50% | 19.6% | 19.6% | 19.6% | 19.6% | 19.6% |
| 112 | 9.25% | 17.9% | 17.9% | 17.9% | 17.9% | 17.9% |
| 113 | 10.00% | 16.2% | 16.2% | 16.2% | 16.2% | 16.2% |
| 114 | 10.75% | 14.6% | 14.6% | 14.6% | 14.6% | 14.6% |
| 115 | 11.50% | 12.9% | 12.9% | 12.9% | 12.9% | 12.9% |
| 117 | GP promote ($): exit cap rate (down) vs hurdle 1 (across) | |||||
| 118 | 33,356,840 | 4.0% | 6.0% | 8.0% | 10.0% | 12.0% |
| 119 | 5.50% | 108,694,590 | 108,694,590 | 108,694,590 | 108,694,590 | 108,694,590 |
| 120 | 6.25% | 82,728,849 | 82,728,849 | 82,728,849 | 82,728,849 | 82,728,849 |
| 121 | 7.00% | 62,327,195 | 62,327,195 | 62,327,195 | 62,327,195 | 62,327,195 |
| 122 | 7.75% | 45,874,249 | 45,874,249 | 45,874,249 | 45,874,249 | 45,874,249 |
| 123 | 8.50% | 33,356,840 | 33,356,840 | 33,356,840 | 33,356,840 | 33,356,840 |
| 124 | 9.25% | 26,545,477 | 26,545,477 | 26,545,477 | 26,545,477 | 26,545,477 |
| 125 | 10.00% | 20,755,818 | 20,755,818 | 20,755,818 | 20,755,818 | 20,755,818 |
| 126 | 10.75% | 16,100,363 | 16,100,363 | 16,100,363 | 16,100,363 | 16,100,363 |
| 127 | 11.50% | 13,212,363 | 13,212,363 | 13,212,363 | 13,212,363 | 13,212,363 |
The first hurdle result is the one a negotiator should stop on. Compound’s summary gives the reason: “With a 100% catch-up the first hurdle only binds when the LP return lands close to it”. Once the catch-up is complete the GP has 20% of all profit, whatever hurdle came before it. The catch-up rate is an input on the same sheet, and it is the term to test next.
GP promote by exit cap rate
Promote in dollars at the 8% first hurdle. Each label gives the LP net IRR at the same exit cap rate; the base case is 8.50%.
Compound workbook, Assumptions sheet, sensitivity data tables. The exit cap rate is an analyst assumption.
The going-in cap rate is 10.5%. An exit at about that yield sits between the 10.00% and 10.75% rows, where LP net IRR is 16.2% to 14.6% and the promote is $20,755,818 to $16,100,363. By our arithmetic the promote at a 10.75% exit is $17.3 million below the base case. At 11.50%, the highest rate in the table, the LP classes still earn 12.9%, above every class’s preferred return.
What can public filings not tell you about this deal?
The filings give annual totals for the property and the terms of the loan. A price, the operating detail and the documents come from the seller, and each one replaces an assumption in the workbook.
Compound’s summary ends with what a buyer would get from the seller to firm the model up:
- A current appraisal or price guidance. The $199,000,000 is from December 2017, and every return follows from it.
- Monthly profit and loss for the last 12 months, the reports on occupancy, average daily rate and RevPAR against the competitive set, and the 2026 budget.
- The ground lease. Compound’s summary cites the CSAIL 2018-C14 offering documents for the loan being secured by a leasehold interest. Its term, rent, resets and consent rights are in the lease.
- The loan documents. They give the exact whole-loan balance and note split, the prepayment terms, which the model assumes allow a payoff without cost in December 2027, and the assumption terms and fee.
- Any franchise and management agreements, a required renovation plan, FF&E reserve balances and the capital spending history. The 4.0% reserve is the reported gap between net operating income and net cash flow.
- The management fees and the property tax reassessment expected after a sale.
The joint-venture terms are a different kind of input. They come from your negotiation, and the workbook marks each one “Request” or “Analyst assumption” so that you can replace it with the term sheet’s figure.
Our view
We would negotiate two terms on this sheet before any other: what the 30% test counts, and the 100% catch-up. The LP return of 19.6% depends more on the 2017 price and the year-1 refinance than on any term in the waterfall.
The model shows where the terms have force. The first hurdle has none across the tested range. The catch-up and the two upper splits decide the promote, and the 70/30 tier alone pays the GP $16,603,789. The third hurdle sits just out of reach at the base exit cap rate, so a small change in the sale price opens the 50/50 tier.
The other side is fair to the GP. It is paid no promote until every class has its capital and preferred return, and the clawback test finds nothing owed. At the highest exit cap rate in the table the LP classes still clear every preferred return.
Three things would change our view. A current price above the 2017 appraisal would lower the going-in cap rate and every return with it. A lender that values the hotel on trailing net operating income, or after the down year, would shrink the year-1 distribution, which is more than half of the equity in this model. And a floating-rate refinance, which the switch on the Assumptions sheet prices off the forward curve, would change the debt service that the down year has to cover. The SOFR forward curve page shows that curve, and the real estate pro forma page covers the property-level model that would sit under a waterfall like this one.
Real estate equity waterfall model (Excel)Compound’s workbook from this run: the Assumptions sheet with a source or a reason for every input and two live sensitivity tables; the Cash Flows sheet with the loan, the refinance and the sale; the Waterfall sheet with each class’s account, the seven tiers by year and returns by class; the Clawback test; the Checks; and the Sources list.Download · FreeFrequently asked questions
What is a real estate equity waterfall?
It is the set of rules that divides a deal’s cash between its investors and the GP, in order. In this workbook the order is return of capital, a preferred return for each class, a top-up to the first hurdle, a GP catch-up, and three splits above LP IRRs of 8%, 15% and 20%.
How do you build an equity waterfall in Excel?
Start with one row of deal cash flow by year, then give each tier a capacity row and a payment row, where the payment is the smaller of the cash remaining and the capacity. Compound’s workbook keeps an account for each class and each hurdle, and checks that the tiers add back to the deal cash flow in every year.
What is a GP catch-up in a real estate waterfall?
After the preferred return is paid, the catch-up sends cash to the GP until its promote reaches a target share of the profit paid so far. In this model the rate is 100% and the target is 20%, so the catch-up is $7,388,445, a quarter of the $29,553,779 preferred return.
Is the preferred return the same as the first IRR hurdle?
They can differ, and this model keeps them apart. Each class has its own preferred return, from 8.0% to 10.0%, while the first hurdle is an 8% IRR on the LP classes together.
How does a clawback work in a real estate joint venture?
At exit the test asks whether the GP was paid promote that the final result does not support: capital or preferred return still owed, or promote paid at a split whose hurdle was not met in the end. The GP returns the excess, capped here at the promote it received. In this model the amount due is nothing, because all promote is paid at the sale.
Which input moves LP returns most in an equity waterfall?
In this model it is the exit cap rate. Between 5.50% and 11.50%, LP net IRR runs from 26.2% to 12.9%. The first hurdle, tested from 4% to 12%, changes neither figure when the catch-up is 100%.





