DCF for Real Estate: Property-Level vs REIT-Level

Introduction

Real estate uses two structurally distinct DCFs that share the name but answer different questions. The property-level DCF values a single building by projecting NOI minus capex, tenant improvements, and leasing commissions over a 5 to 10 year hold, capping the year-N+1 NOI at an exit cap rate for the terminal value, and discounting the whole stream at a property-level required return. The REIT-level DCF values an entire publicly traded REIT by projecting either AFFO or dividend cash flows over a forecast horizon, applying a terminal value via Gordon Growth or a forward P/AFFO multiple, and discounting at the REIT's cost of equity. The two methods use different inputs, different discount rates, and answer different questions. Confusing them is one of the most common technical-interview errors at RE IB superdays.

Both DCFs are right when applied to the right question. A sponsor underwriting a value-add office acquisition runs the property-level DCF; a public-equity analyst valuing a listed multifamily REIT runs the REIT-level DCF. A banker pitching a take-private uses both: property-level DCF on the underlying assets to support NAV, REIT-level DCF on the entity to cross-check NAV against a public-equity valuation framework.

The property-level DCF is the standard sponsor underwriting tool and the second of the four real estate valuation methods at the property layer. The mechanic projects unlevered or levered cash flow over a defined hold (typically 5 to 7 years for value-add and core-plus, 10+ years for core), with the terminal value computed as the forward NOI capitalized at an assumed exit cap rate. The standard formula:

Where NCF in each year is property-level cash flow (NOI minus capex, TIs, LCs, and debt service for the levered version, or just NOI minus capex, TIs, LCs for the unlevered version), r is the discount rate, and the terminal value applies the exit cap to the forward NOI immediately after the hold ends.

Property-Level DCF

A valuation method that projects a single property's annual cash flows (typically NOI minus recurring capex, tenant improvements, and leasing commissions) over a 5 to 10 year hold period, discounts each year's cash flow at a property-level required return, and adds a terminal value computed as the forward-year NOI divided by an assumed exit cap rate. Used by sponsors for acquisition underwriting and by lenders for credit analysis. The dominant inputs are the NOI growth assumptions, the exit cap rate, and the discount rate.

The discount rate for a property-level DCF is the property's required return at the cash flow definition used. For an unlevered DCF (NCF before debt service), the discount rate is the unlevered required return: typically 7-9% for core stabilized property, 10-12% for value-add, 12-15% for opportunistic. For a levered DCF (NCF after debt service), the discount rate is the levered required return, which is the unlevered return plus a leverage premium that compensates for the financial-leverage risk added.

The Exit Cap Sensitivity

The single most sensitive input in a property-level DCF is the exit cap rate. Because a meaningful share of the property's value sits in the terminal value (often 60-70% of total present value on a 7-year hold), a small movement in the exit cap assumption produces a large movement in the implied property value. Standard underwriting assumes the exit cap is 25 to 50 basis points wider than the going-in cap rate to account for the property's age at sale. Aggressive underwriting sometimes assumes flat or compressing exit caps; lender stress tests routinely widen the exit cap by an additional 50-100 basis points as a downside scenario.

Isolated from the discounting, the terminal value itself is just the forward NOI capitalized at the exit cap rate, the same arithmetic as the direct capitalization method applied at the end of the hold:

This standalone figure is then discounted back N years inside the property-level DCF above. Because the exit cap sits in the denominator, the terminal value moves inversely with it, which is the mechanical source of the sensitivity described here.

The discount rate is the second most sensitive input, but its impact concentrates on the cash flow years rather than the terminal value, so its effect is more linear than the exit cap's. The two inputs together drive almost the entire variance in any property-level DCF output.

The REIT-level DCF values an entire publicly traded REIT by projecting either AFFO per share or dividend per share over a forecast horizon (typically 5 to 10 years), applying a terminal value, and discounting at the REIT's cost of equity. Two variants are common.

The DDM is the cleanest theoretical method for REITs because REITs are required to distribute at least 90% of taxable income as dividends, which makes dividend cash flows highly representative of the underlying economics. A two-stage DDM projects explicit dividends for 5 to 10 years, then applies a Gordon Growth terminal: Terminal Value = D_N+1 / (r - g), where g is the long-run sustainable dividend growth rate.

The DDM and AFFO variants both discount equity cash flows at cost of equity, but the third variant in the table above, the less common FCFF DCF, discounts free cash flow to the firm and therefore uses the blended WACC rather than cost of equity. WACC weights the cost of equity and the after-tax cost of debt by their shares of total capital:

Where E and D are the market values of equity and debt, V is E plus D, and t is the marginal tax rate. The FCFF variant is rare for REITs precisely because the entity-level tax shield is muted: REITs pay little corporate tax when they distribute their required income, so the after-tax debt adjustment carries less weight than it does for an ordinary corporate DCF.

Dividend Discount Model (DDM) for REITs

A REIT valuation method that projects per-share dividends over an explicit forecast horizon (5 to 10 years), applies a Gordon Growth perpetuity terminal value, and discounts the whole stream at the REIT's cost of equity. The formula for the Gordon Growth terminal is D_N+1 divided by (r minus g), where r is the cost of equity and g is the long-run dividend growth rate. Most effective for mature REITs with stable distribution policies; less reliable for high-growth REITs where the growth phase has not yet stabilized.

The AFFO-based DCF is the more common bank-side method because AFFO is the recurring-cash-flow metric most analysts use for REIT comparable trading multiples. A forecast of AFFO per share over 5 to 10 years, discounted at cost of equity, with a terminal value computed as the year-N+1 AFFO multiplied by a forward P/AFFO multiple, produces a present-value-per-share estimate that can be compared directly to the listed market price.

When the analyst opts for the Gordon Growth terminal named above rather than a forward multiple, the perpetuity-growth form of the terminal value is the final-year cash flow grown one period and capitalized at the spread between the discount rate and the long-run growth rate:

Where CF_N is the final explicit-forecast cash flow (the dividend in a DDM or AFFO per share in an AFFO-based DCF), g is the long-run sustainable growth rate, and r is the cost of equity. This is the same Gordon Growth perpetuity referenced for the DDM above, written in its general grown-cash-flow form.

The Cost of Equity Input

The REIT-level discount rate is the cost of equity, calibrated either through CAPM (risk-free rate plus equity beta times equity risk premium) or through a build-up that adds property-type and capital-structure premia to a base required return. In its standard CAPM form:

Where R_f is the risk-free rate, beta is the REIT's equity beta, and (R_m minus R_f) is the equity risk premium. Typical US listed REIT cost of equity sits in the 7-10% range for stable IG-rated names and 9-12% for higher-growth or higher-leverage names. The exact figure varies meaningfully with the rate environment and with sub-sector dynamics.

The critical constraint on the Gordon Growth terminal is that g must be meaningfully less than r. As g approaches r, the terminal value approaches infinity, which is structurally nonsensical. The standard discipline is to cap g at the long-run nominal GDP growth rate (roughly 4-5% in steady state) for any terminal-value calculation. The same Gordon Growth versus exit multiple tradeoff that governs a standard corporate DCF applies here, with the forward P/AFFO multiple playing the role the exit EBITDA multiple plays elsewhere.

The two DCFs answer different questions and rarely substitute for each other.

The recurring technical-screen mistake is applying property-level DCF mechanics to a REIT valuation question or applying REIT-level DCF mechanics to a single-property question. They are not interchangeable. A property-level DCF with exit cap rates and unlevered hold-period IRR is the right answer when the question is "how should the sponsor underwrite this acquisition." A REIT-level DCF with cost of equity, AFFO projections, and Gordon Growth or terminal multiple is the right answer when the question is "what is the per-share intrinsic value of this listed REIT." Mixing the methods (running a cost-of-equity discount on property NOI; applying an exit cap to a multi-property REIT enterprise value) produces numbers that look defensible at first glance but fall apart under any rigorous diligence. The two DCFs share a name and a discounting mechanic; the inputs and the question they answer differ structurally.