How to calculate real estate portfolio-level IRR

What is portfolio-level IRR and why does it matter?

Portfolio-level Internal Rate of Return (IRR) is the annualised return an investor earns across an entire real estate portfolio, computed as the discount rate that sets the net present value of all portfolio cash flows to zero. In institutional real estate, IRR is the standard performance metric for closed-end funds, joint ventures, and separately-managed accounts because it captures both the magnitude and the timing of cash flows over the fund life.

The distinction between asset-level IRR and portfolio-level IRR matters because they answer different questions. Asset-level IRR tells you how a specific building performed. Portfolio-level IRR tells you how the fund's investment thesis performed as a whole, including capital pacing, disposition timing, reinvestment decisions, and the drag of assets that underperformed alongside those that outperformed.

Fund managers report portfolio-level IRR to LPs quarterly. LPs use it to compare funds, vintages, and managers. Underwriters use projected portfolio-level IRR to price commitments to future funds. Getting the calculation right is fundamental to institutional real estate operations.

For the broader context on how institutional asset managers track portfolio performance, see our real estate asset management trends blog and the REIT technology guide.

How is IRR calculated (the math)?

IRR is the discount rate r that solves this equation for a series of cash flows CF₀, CF₁, ..., CFₙ occurring at times 0, 1, ..., n:

0 = Σ (CFₜ / (1 + r)^t)  for t = 0 to n

No closed-form solution exists; the equation is solved numerically by iterating on r until the sum equals zero.

In practice, no institutional investor calculates IRR by hand. The workflow uses Excel or a specialist platform.

Excel formulas.

• =IRR(cashflow_range) for cash flows at regular annual intervals.

• =XIRR(cashflow_range, date_range) for cash flows on specific dates (the standard for real estate because acquisitions, distributions, and dispositions do not fall on year-end).

• =MIRR(cashflow_range, finance_rate, reinvest_rate) for a modified IRR that lets you specify reinvestment assumptions separately from the borrowing cost.

For most real estate portfolios, XIRR is the right function because cash flows arrive on irregular dates.

The one place investors go wrong on the math is the sign convention. In Excel, contributions (money going out to investments) are negative, and distributions (money coming back) are positive. Portfolio-level IRR should sum all contributions as negatives and all distributions as positives, then apply XIRR to the resulting time series.

What cash flows go into a real estate portfolio IRR?

A complete real estate portfolio IRR calculation captures five categories of cash flow across every asset over the hold period.

Equity contributions. Every dollar the investor deploys into an asset: initial acquisition equity, follow-on capex contributions, capital calls, and any additional funding rounds. All negative in the cash flow schedule.

Operating cash flow. Monthly or quarterly net operating income after debt service (for levered IRR) or before debt service (for unlevered IRR), less capex not funded through additional equity contributions. Typically positive after stabilisation.

Refinancing proceeds. When a property refinances at higher LTV, the excess proceeds distributed to investors are treated as a positive cash flow at the refinancing date. These are among the largest cash flows in a value-add strategy.

Interim distributions. Preferred return payments, quarterly distributions, and special distributions. Positive cash flows at each distribution date.

Disposition proceeds. The net sale price minus transaction costs and debt payoff (for levered IRR). The largest single cash flow in most real estate holds, occurring at the end of the hold period.

Portfolio-level IRR sums these five categories across every asset in the portfolio into a single chronological time series, then applies XIRR to solve for the annualised return.

For institutional investors, the platform tools handle this aggregation automatically. See our PropTech platform architecture guide for how portfolio-level data models are structured.

Levered vs unlevered IRR: which should you use?

Two versions of IRR are computed on the same underlying portfolio depending on whether cash flows include or exclude debt service.

Unlevered IRR (or "asset-level" or "property-level" IRR). Uses cash flows before debt service. The numerator is net operating income minus capex, without deducting interest or principal payments. The denominator is the total capital invested (equity plus debt) at acquisition. Measures the underlying real estate's economic performance regardless of financing structure.

Levered IRR (or "equity IRR"). Uses cash flows after debt service. The numerator is net operating income minus interest and principal payments and capex. The denominator is the equity invested. Measures what the equity investor actually earns.

Levered IRR is typically 300 to 600 basis points higher than unlevered IRR on the same asset, driven by the multiplier effect of debt on equity returns. Institutional fund benchmarks (Cambridge Associates, Preqin, NCREIF) are typically reported on a net levered basis because that is what LPs actually receive.

Both metrics matter and neither replaces the other. Underwriting a new acquisition typically starts with unlevered IRR (to isolate the real estate risk) and then layers in the financing structure to arrive at levered equity IRR (the equity investor's actual expected return).

Gross vs net IRR: how do fees affect the number?

Gross vs net IRR is the fund-level distinction that separates what the general partner reports internally from what limited partners actually receive.

Gross IRR. IRR computed on cash flows before deducting management fees and performance-based compensation (promote or carried interest). Typically 200 to 400 basis points higher than net IRR on a value-add or opportunistic strategy.

Net IRR. IRR computed on cash flows after deducting management fees, promote, and fund-level expenses. What LPs actually earn on their invested capital.

For institutional benchmarks, industry data sources (Preqin, Cambridge Associates, NCREIF) typically report net IRR. When comparing fund performance across managers, always confirm whether the reported number is gross or net; comparing gross to net produces misleading conclusions.

The wedge between gross and net IRR is driven by the fund's fee structure. Typical institutional real estate fund economics in 2026: 1.5 to 2 percent management fee on committed capital during the investment period, 1 to 1.5 percent on invested capital thereafter, 20 percent promote on returns above an 8 to 10 percent preferred hurdle. Some funds carry a 6 to 8 percent hurdle and lower promote; others carry higher hurdles and higher promote. All materially affect the gross-to-net wedge.

What tools do institutional investors use to calculate portfolio-level IRR?

Institutional real estate investors use a stack of tools rather than a single platform.

Excel. Still the default for underwriting new acquisitions, scenario modelling, and manual portfolio analysis. Every institutional analyst is fluent in XIRR and cash-flow modelling in Excel. Excel-based models are usually where a new acquisition IRR is first calculated before it flows into a portfolio system.

Argus Enterprise. The industry-standard commercial real estate cash-flow modelling platform. Handles lease-by-lease modelling, capex scheduling, refinancing scenarios, and portfolio aggregation. Owned by Altus Group. Widely used at institutional operators, appraisers, and lenders. Typical cost $2K to $5K per user per year plus enterprise contracts.

Yardi Investment Manager. Portfolio-level investment tracking, IRR calculation, LP reporting, and capital-account management inside the Yardi ecosystem. Best fit for operators already running Yardi Voyager on the property side.

RealPage IMS (Investor Management Services). Similar function to Yardi Investment Manager, from RealPage. Handles LP reporting, capital calls, distributions, and portfolio IRR tracking. Widely used in US multifamily fund and syndication management.

Juniper Square. Cloud-native fund administration and investor management platform. Growing quickly in institutional real estate through 2020 to 2026. Handles LP portal, capital calls, distributions, K-1 delivery, and portfolio IRR reporting.

Custom data warehouses. Larger institutional operators (Blackstone, Brookfield, larger REITs) increasingly build custom portfolio data warehouses on top of Snowflake, Databricks, or similar cloud platforms, with IRR computation and LP reporting layered on top through Tableau, Power BI, or specialist dashboards.

For scoped platform selection support, see our CRM implementation service and the property management software blog.

What is a good IRR for a real estate portfolio?

Institutional benchmarks vary by strategy, vintage, and cycle. Broad 2026 ranges published by Cambridge Associates, Preqin, and ULI for net IRR (after fees and promote) on institutional US real estate funds.

Core strategies (stabilised, income-focused). Target net IRR 6 to 9 percent. Actual delivery for top-quartile funds typically 7 to 10 percent. Bottom-quartile 3 to 5 percent.

Core-plus strategies (light value-add on stabilised assets). Target net IRR 8 to 12 percent. Actual top-quartile 10 to 14 percent.

Value-add strategies (major repositioning, lease-up, or development). Target net IRR 12 to 18 percent. Actual top-quartile 15 to 20 percent. Bottom-quartile can be zero or negative in bad vintages.

Opportunistic strategies (development, distressed, ground-up). Target net IRR 15 to 25 percent. Actual top-quartile 18 to 30 percent. Bottom-quartile can be materially negative.

Two important caveats. First, IRR is not the whole picture; multiple on invested capital (MoIC), cash-on-cash yield, and risk-adjusted metrics like the modified Sharpe ratio matter alongside. Second, vintage matters enormously; funds raised in 2007 to 2008 or 2020 to 2022 experienced very different market cycles than 2013 to 2015 vintages, and comparing across vintages without adjusting for macro conditions produces misleading conclusions.

For historical context on REIT returns as a benchmark, see our REIT technology guide. The FTSE Nareit All Equity REITs Index has produced an average annual total return of around 9.4 percent over the last 20 years (per Nareit), providing a public-market reference point for institutional private real estate performance.

What are the common mistakes in calculating portfolio-level IRR?

Seven mistakes explain most of the wrong IRR numbers institutional investors report.

Wrong sign convention in Excel. Contributions should be negative, distributions positive. Getting this backwards produces obviously wrong numbers (usually enormously negative), which analysts sometimes "fix" by taking the absolute value. Always double-check the sign convention.

Mixing gross and net cash flows in the same series. If some cash flows are net of fees and others are gross, the resulting IRR is meaningless. Pick one convention and stick with it.

Missing capex or follow-on equity contributions. Every capital call, capex draw, and follow-on funding round is a negative cash flow in the schedule. Leaving them out inflates IRR materially.

Using annual instead of dated cash flows. For real estate, most cash flows do not fall on year-end. IRR produces wrong answers on irregular timing; XIRR handles dated cash flows correctly. Always use XIRR for real estate.

Treating unrealised gains as cash flows. Only actual cash movements count in IRR. Mark-to-market appraisal gains that have not been realised through a sale or refinancing do not belong in the IRR calculation, even though they appear in NAV.

Confusing time-weighted return with money-weighted return. IRR is money-weighted (dollar-weighted). Time-weighted return (TWR) is a different metric that neutralises the impact of contribution timing. Public market benchmarks are typically TWR; private real estate benchmarks are typically IRR. Comparing across the two produces misleading results.

Cherry-picking hold periods. Reporting IRR on a truncated hold period (e.g., 3 years into a 10-year hold, using a mark-to-market NAV as the terminal value) is common but can materially misstate final performance. Always disclose whether reported IRR is realised, unrealised, or blended.

Ready to build the data infrastructure that supports accurate portfolio-level IRR?

Book a working session with the Noseberry Digitals team. We will audit your current portfolio reporting stack, identify the platform and data-layer investments that would materially improve IRR calculation reliability, and hand back a scoped roadmap. For personalised guidance on fund structuring, fee waterfalls, or specific investment decisions, we can point you to qualified real estate CFAs and CPAs; for institutional data infrastructure and platform selection work, we run the sessions in-house.

Book a portfolio infrastructure working session →