How to read the IRR and ROE of a closed project using a public example

IRR is the annual rate at which the present value of all the project's cash flows equals zero, and ROE is the return on equity; in this example it is stated for the whole term of the project. A look at a public example: on 22 June 2026 Fellow Funders reported through Capital-Riesgo.es the Sant Joan Despí Fase II round: 23 homes with parking, 5,4 M€ in total, of which 4,32 M€ from the platform's investors and 1,08 M€ (20%) from the developer SHEVAT. An estimated IRR of 15,02% a year, an ROE of 34,29% and a term of about 24 months are stated. It is a publication based on platform data, and the figures are estimates. I did not find the project page itself, so I checked against the text of the publication.

How do the 15,02% IRR and the 34,29% ROE relate?

They measure different things: ROE accumulates over the whole term, while IRR is expressed per year. To compare them, convert the ROE to an annual figure with the formula published in the DNPI methodology: (1 + result)^(12 / months) − 1. For 34,29% over 24 months this gives 15,88% a year. That is close to 15,02%, but not equal.

The gap can be explained with simple arithmetic. If the money goes in as one sum and comes back as one sum, a multiple of 1,3429 at an IRR of 15,02% corresponds to a term of about 25,3 months, not 24. Either "about 24 months" is rounded, or the flows came at different times: staged payments and partial returns change the IRR. For reading, this is the main lesson: IRR depends on the timing of flows, ROE does not.

A worked calculation on the publication's figures, not a forecast: the same multiple of 1,3429 gives a different annual result at a different term.

A half-year delay at the same profit eats 3,4 points of annual return (from 15,9% to 12,5%). That is why an ROE without a term and an IRR without flows say nothing. The limitation of IRR was described by Ludovic Phalippou on the blog of the CFA Institute Research and Policy Center (8 November 2024): IRR is not the same as ROI, and the rate for an asset with interim flows requires assumptions about reinvestment. In detail: IRR in plain language and equity multiple versus IRR.

What does the publication lack for a conclusion?

In the text I read, the following are not stated:

• the base of the ROE (all capital of 5,4 M€, only the investors' capital, or net profit after distribution);
• whether the 5,4 M€ includes bank financing or the project is funded entirely by participants' own and raised funds;
• platform fees and the investor's taxes: whether the 15,02% IRR was estimated before or after them is unknown;
• the scenario: the word "estimada" refers to the base calculation, and no conservative scenario is shown beside it;
• the difference between the developer's 20% of capital and its share of profit: the text says the developer takes part on the same economic terms as the others, but its construction fee, if any, is not disclosed.

The platform itself writes that the figures are estimates, do not guarantee future results, and that an investor may lose capital in full or in part. The comparison with deposits (2–3%), conservative funds (4–6%) and SOCIMI (Spanish listed property companies, 5–8%) is given in the publication without a source, and I do not use it in the calculation.

What does the developer's 20% of own capital provide?

It is a signal of aligned interests, not a measure of protection. The developer put in 1,08 M€ of the 5,4 M€, so in the event of a loss it loses alongside the others. But the effect depends on the structure: whether it takes a construction fee regardless of the result, and whether its money sits in the payout queue in the same position as the investors' money. A separate analysis: why an investor wants the operator to put in its own capital.

How can this analysis be applied to another project?

Ask five questions of any closed offer: to what term and to which flows the IRR is tied; what the base of the ROE is; before or after which costs and taxes it was calculated; which scenario is shown; and how much of its own money the operator put in, and on what terms.

In the DNPI methodology, project return and annual return are calculated separately: ROI shows the result for the whole project, the annual figure brings it to a comparable period, and the investor's result is set by the contract. You can estimate your own scenario in the calculator, and the difference between project and investor figures is explained in the article on project return and investor return.

Who is this analysis not for?

Anyone who wants to assess a specific platform or decide whether to invest in Sant Joan Despí: for that there are the project's own documents, not this article. Anyone looking for a return forecast in Barcelona: the public example shows how to read the figures and does not derive a market norm from them. One project with estimated figures from a press release does not make a sample.