Committing to a real estate AIF: reading the sponsor's register record first

Limited partners diligence real estate funds the way funds diligenced developers a decade ago: on presented track record, references and reputation. The manager's deck shows realised IRRs and marquee exits; the portfolio's failures appear, if at all, as footnotes. Since 2017, the underlying asset class has kept a public conduct record that LP diligence rarely touches. For a family office writing a Rs 5 or 25 crore commitment, an afternoon on the register is the cheapest underwriting test available. Here is the checklist, and the calibration that has to come before it.

Why the register belongs in LP diligence#

The scale first. Real estate has long been the largest single sector for AIF investment, and SEBI's data put it at a record Rs 1.29 lakh crore invested by March 2026, inside an industry whose commitments reached Rs 17.53 lakh crore by June 2026, with Rs 7.58 lakh crore raised and Rs 7.10 lakh crore invested (SEBI data via Business Standard, September 2026). The June figure for real estate itself came in at Rs 87,969 crore, the first fall on record, which industry commentary reads as a reclassification in SEBI's sector data rather than as funds leaving. Either way the sector still leads.

A growing share of Indian private capital's real estate exposure now flows through managers whose skill claims are, at bottom, claims about picking developers and micro-markets. Both claims are testable. Developers keep sworn public records: delivery against promises, extensions, litigation, money certificates, the whole conduct file this series built into an underwriting checklist. Micro-markets keep computable records: supply, bookings, distress share, the catchment mapping method. LP diligence that stops at the fund's own numbers takes the manager's word on precisely the two things the public record can verify.

The state's own vehicle shows the method works at scale. SWAMIH Investment Fund I is a SEBI-registered Category II AIF, sponsored by the Ministry of Finance and managed by SBICAP Ventures, and its eligibility rules start from the register: a project must be RERA-registered, stalled for want of funds, in the affordable or mid-income bracket, and net-worth positive, meaning the money still due from buyers plus the unsold stock must exceed the cost to finish. By 31 January 2026 it had committed to 148 projects covering 1,01,443 homes and delivered 63,200 of them, the Finance Minister told the Lok Sabha in March 2026, and a second Rs 15,000 crore fund announced in the 2025 Union Budget was then being launched. Underwriting from the filing is not a boutique idea.

Calibrate before you read: what a normal record looks like#

One trap waits for every LP who reads the register for the first time. A large developer almost always has a dead registration somewhere, so finding one proves little on its own. Counted by exact filed promoter name across the 55,995 published MahaRERA registrations, as updated on 29 September 2026:

Read the chart with the number it hides. The chance that a promoter carries at least one dead registration rises from 29 percent for a single-project name to 76 percent at ten or more, while the share of registrations that end that way falls, from 29.4 percent to 20.8 percent. So a large sponsor with one lapsed project is the norm, and the diligence question is its rate against that peer set, not whether a lapse exists. Lapsed is also a status the regulator records rather than a finding about the building, and the filing does not say why a registration lapsed, so read each one as a question, not a verdict.

The count is by filed name for a reason worth stating. MahaRERA records each SPV separately and no parent company, so a group trading through five SPVs appears here as five promoters. Assembling the group from the fund's disclosures and the company filings is the LP's job, and it is the first step of the checklist.

The three-part checklist#

1. The sponsor and manager's own real estate footprint. Where the fund's sponsor group develops or co-invests, build the group from the fund documents, the company registry and the promoter names on each registration, then read its record exactly as a lender would: delivery slips, extension frequency, filing discipline, enforcement flags, each against the calibration above. A manager whose affiliated projects file late and extend serially is asking you to believe underwriting standards they do not apply at home.

2. The pipeline's filing health. PPMs and DDQs disclose completed deals and, often, a pipeline. Map every named or inferable counterparty to the register: current QPRs, certified progress against time elapsed, litigation, declared charges. Then run the five early-warning signals across the disclosed portfolio. Filing lapses are common enough to matter: MahaRERA issued show-cause notices to 8,212 projects in May 2026 for missing the April deadline for the January to March quarter. You are not re-underwriting each deal; you are sampling the manager's claimed discipline against sworn data. Two stressed names presented as performing tells you more than any reference call.

3. Catchment concentration. Aggregate the portfolio's projects by pincode and attach unsold stock and the booking pace each filing implies, units booked per month on the register. Neither is a filed field; both come from fields that are, and a catchment's months of inventory at its filed pace is one division away. Funds concentrate where their networks are, so two funds can each be internally diversified while stacking an LP's total exposure into one corridor. This is the connected-exposure logic banks apply to sponsor groups, applied to buyer pools.

Worked example: one commitment, two register reads#

Illustrative numbers. A family office evaluates Rs 10 crore commitments to two Category II real estate credit funds, both targeting mid-market Maharashtra residential, both quoting mid-teens gross returns, both with presentable realised track records.

The register pass took one afternoon per fund. Fund A's eight identifiable portfolio developers show a median delivery slip of 3 months across 31 completed projects; one live borrower carries an extension; every QPR in the sample is current; and the portfolio spans nine catchments, the slowest carrying 17 months of unsold stock at its filed pace. Fund B's seven identifiable developers show a median slip of 9 months; two borrowers appear in the May 2026 show-cause list of 8,212 projects; two carry lapse rates well above the norm for their size; and four of the seven sit in two Pune west catchments whose unsold stock would take more than 28 months to clear at the pace their filings imply.

Same decks, same references, different books. The office committed to Fund A and put a written question to Fund B, whose answer, when it came, contradicted the register in two particulars. That letter cost nothing, and it settled the decision.

Five DDQ additions that operationalise this#

For offices that run a standing due diligence questionnaire, the register pass converts to five questions worth adding verbatim.

1. "List the RERA registration numbers of all current portfolio projects." The answer enables everything else; reluctance to answer is itself an answer. Registrations are public, so the request costs the manager nothing but control of the narrative.
2. "Describe your quarterly filing-review process for portfolio projects." A real process names artifacts: QPRs, Forms 1 to 3, extension records. A vague answer ("our asset management team monitors sites closely") means monitoring runs on site visits and the borrower's own MIS, which the lender series opener shows to be the weaker half of the record.
3. "Has any portfolio project missed a quarterly filing deadline in the past eight quarters? What followed?" The latest filing date is on the register, so the recent part of the answer can be checked independently, which is the point of asking.
4. "What is your policy when a borrower's other projects, outside our deal, show extensions or enforcement flags?" Group contagion is the most common surprise in developer credit; a manager without a stated policy discovers it in real time, with your capital.
5. "Map current portfolio projects by catchment and share the overlap with your prior funds." Concentration disclosure at buyer-pool level, asked as a question rather than reconstructed.

One structural note completes the checklist. Most real estate credit strategies sit in Category II, where SEBI's framework leaves portfolio construction to the manager and disclosure to the fund documents, which is precisely why LP-side verification carries the weight it does: no regulator grades the deals. The fee arithmetic sharpens the incentive. On a Rs 10 crore commitment, a 2 percent management fee costs Rs 20 lakh a year regardless of underwriting quality; the register pass costs an afternoon and at most some tens of thousands of rupees of report spend, and it is the only line in the diligence budget that tests the thing the fee is supposed to buy.

The questionnaire's job is not to catch managers lying; most do not. It is to distinguish managers whose underwriting edge is real and filings-literate from managers whose edge is deal flow and relationships, because the two books behave differently in a down cycle, and the fee is the same.

The analogy from allocation practice: nobody commits to an equity manager without checking the claimed alpha against a benchmark and the holdings against the mandate. Real estate AIFs have had no benchmarkable holdings data, so LP diligence substituted trust. The register is the holdings feed: not prices, but conduct, which in credit is the variable that matters.

The Bhandarkar family office (illustrative, as our stories always are) now writes the register pass into its DDQ for every real estate commitment: the manager is asked to list portfolio projects' registration numbers, and the office reads the filings itself. One manager called the request unusual. The other called it overdue. Both answers were informative.

The assembly work is what ReraGenie sells. The free developer pages list every registration under a filed promoter name with its status; the Rs 2,999 project analysis sets one portfolio project against its micro-market, its operators and a risk register; and the area market report, a flat Rs 2,999 per pincode, carries a catchment's supply, status ledger and sales cross-section for the concentration check. None of them resolves a group for you, because the register does not record one, so that step stays with the LP.

The one-line summary#

The fund's deck shows the record the manager chose; the register shows the one the market wrote. Calibrate against the peer set, read the sponsor's conduct, sample the pipeline's filings, map the catchments, and let commitments follow the book that survives the public record.

Methodology and sources#

• AIF industry and sector figures: SEBI data as reported by Business Standard (June and September 2026).
• SWAMIH: fund structure and eligibility from SBICAP Ventures and the Press Information Bureau (March 2026); portfolio figures to 31 January 2026 from the Finance Minister's statement in the Lok Sabha, March 2026.
• Show-cause notices for missed quarterly filings: MahaRERA, as reported in May 2026.
• Register figures: ReraGenie analysis of the 55,995 published MahaRERA registrations, as updated on 29 September 2026. Promoters are counted by exact filed name; lapsed, revoked and de-registered are the statuses as filed.