Credit desks love a public grade: it is free, it is a third party's work, and it fits in a sanction note. So when MahaRERA announced in August 2023 that it would grade projects every six months, the temptation was to plan for a mini-rating. The grades never came. The first were due from April 2024, none were published, and in November 2024 the scheme was paused. What did come is more useful than it sounds: a regulator's published rubric for which filed inputs matter. Its inputs are exactly a monitoring desk's inputs, and it was built to answer a regulator's question, not a lender's. This piece maps one onto the other, dimension by dimension, and shows how a desk computes for itself the score MahaRERA did not publish.
Key takeaways
- MahaRERA's Circular No. 46/2023 of 28 August 2023 designed half-yearly grades for projects registered after January 2023, generated through its MahaCRITI system, with a first window of October 2023 to March 2024.
- No grade was published. In November 2024 a senior MahaRERA official said the scheme would not proceed, because the RERA Act does not empower the authority to grade projects.
- The matrix's four snapshots, overview, technical, financial and legal, overlap heavily with a credit desk's monitoring set, so the design survives as a free specification a desk can compute every quarter.
- Its blind spots even as designed, sponsor-group conduct, pricing, catchment demand, forward pipeline and older vintages, are precisely where credit losses live, so a desk-built score supplements underwriting and never substitutes for it.
What the matrix was designed to measure
The design, from MahaRERA's Circular No. 46/2023: every project registered after January 2023 was to be assessed each half-year, with grades generated through the regulator's MahaCRITI system from the project's own filings. Four dimensions were to feed the score. The project overview covers promoter details and project type. The technical snapshot scores the discipline and content of quarterly updates: percentage of work completed against percentage of time elapsed, booking levels, quality certification, extensions, and the housekeeping obligations of conveyance and society formation. The financial snapshot reads financial progress, the CA's audit report and any qualifications in it, financial encumbrances and penalties. The legal snapshot counts complaints against the project, warrants issued, and court or NCLT proceedings.
Read that list with a monitoring officer's eye and the overlap is striking: work-versus-time is the progress gap from the early-warning framework; update discipline is the filing-silence signal; encumbrances and audit qualifications are the money side of the certificate reconciliation; warrants and NCLT proceedings are the enforcement flags. The regulator, in effect, published a scoring rubric for the same artifacts this series has been turning into desk procedure.
What happened to the scheme
The circular set the first assessment window as 1 October 2023 to 31 March 2024, with grades to follow from April 2024. They did not follow. The first publication was put back, and in November 2024 a senior MahaRERA official told the Times of India that the grading would not proceed, because the Real Estate (Regulation and Development) Act, 2016 does not give the authority the power to grade projects; the decision followed opposition from developers. No project grade has been published since, and no project's public filing carries one.
For a credit desk, the pause changes less than it seems. A grade would have been a compression of public inputs, published twice a year by a third party. The inputs are still public, still filed quarterly, and the circular still names them. What is lost is the third party's work and a shared label; what remains is the regulator's own list of what matters, which is the more useful half.
The mapping, and the gaps
| Credit desk question | Did the matrix cover it? | Where the answer actually lives |
|---|---|---|
| Is this project executing to plan? | Partly: work-vs-time and update discipline were to be scored | The underlying QPR series and Forms 1/2, read as a trend |
| Is the money behaving? | Partly: audit qualifications, encumbrances, penalties | Form 3 reconciliation against disbursement, quarterly |
| Is the sponsor sound? | No: the grade was to be project-level | Group-wide register record: every entity, every project |
| Will the market absorb the inventory? | No: bookings were to be scored, demand was not | Catchment supply, absorption and months-of-inventory |
| Is my vintage exposure covered? | Only registrations after January 2023 were in scope | Direct filing analysis for older projects |
| Is trouble coming next quarter? | No: grades were to be six-monthly and backward-looking | The five-signal quarterly score on live filings |
The pattern in the right-hand column is consistent: every question the matrix would have half-answered is fully answerable from the same public filings it was to be computed from. That is the practical conclusion, and the pause only sharpens it. The grade was always going to be a compression of data you can read uncompressed, on your own cadence, with your own thresholds; now the uncompressed reading is the only form in which it exists.
The worked example: the matrix as checklist, the inputs as monitoring
Illustrative numbers. A Rs 30 crore facility to a Pune project registered in 2024, squarely inside the scheme's intended scope. With no grade to read, the desk uses the matrix three ways. As a screen: the desk computes the matrix's inputs at origination and files them in the annexure beside the six-dimension sponsor checklist; strong project inputs on a weak sponsor group are caught by the group check, which is the point of running both. As a covenant: the facility documents an event of review on either of two filed triggers, a missed quarterly progress report or a work-versus-time gap past a stated threshold, both cheap, externally observable and impossible for a borrower to negotiate down as subjective. As monitoring calibration: the desk computes the work-versus-time ratio every quarter from the filed data. When the project's certified work sits at 31 percent against 52 percent of time elapsed, a 21-point gap, the desk's own threshold of 20 points trips, and the review starts that quarter rather than at the next six-monthly cycle a grade would have waited for. The matrix cost nothing and contributed a checklist, two covenant triggers and a baseline; the desk's own reading of the inputs contributed the early move.
The wider numbers frame why screening layers matter at all. Maharashtra's register carries more than 55,000 projects (MahaRERA counted 50,162 by May 2025). Enforcement sweeps run four-digit lists: 8,212 show-cause notices under RERA section 7 in May 2026, for one missed quarterly deadline. Extensions are routine rather than rare: of the 10,379 approvals MahaRERA granted in 2025-26, 3,687 were timeline extensions, against 4,204 new registrations, which is the base rate the matrix's extension input would have been scored against. And bank credit to commercial real estate grew 16.2 percent year on year in January 2026 (RBI, sectoral deployment of bank credit). Books are growing faster than monitoring headcount; free, structured screens are how the coverage gap closes.
Movements over levels, computed by the desk
Because no grade has been published, the movement series is one a desk builds: compute the matrix's work-versus-time and filing-discipline inputs each quarter and log the changes. Read deltas, not levels. A project whose inputs hold steady across three quarters is a stable mediocrity, priceable; one whose work-versus-time gap just widened by ten points is an event, because something in the filed inputs moved, and the inputs are public, so you can find out what within an hour. The deltas also aggregate at sponsor level: a promoter whose five projects all drifted the same way in one quarter is a group trend wearing five project labels, exactly the pattern the group-level underwriting view exists to catch. And cohort context matters: the same gap means something different in a stressed catchment and in a booming one, which is one more reason the catchment data belongs beside the inputs in every review.
None of this is exotic. It is what desks already do with special mention accounts under the RBI's Prudential Framework of 7 June 2019, where a loan moves through SMA-0, SMA-1 and SMA-2 as its overdues pass 30 and 60 days, applied to a free series of filed inputs that can move before any repayment does. The series is new, which cuts both ways: migration statistics do not exist to calibrate against, so the desks that log input movements against their own outcomes from now on will own the only validation data that matters.
The matrix's deepest limitation, as designed and as a desk computes it, is inheritance: every input is the promoter's own filing. A promoter disciplined enough to file fully and honestly tends to score well, which is real information; but a score can be no better than the disclosure feeding it. False filing can draw a penalty of up to 5 percent of the estimated project cost under section 60 or 61 of the RERA Act, and revocation under section 7, which makes it a risk to price, not an impossibility. The desk's protection is the same reconciliation it should run anyway: certified numbers against your LIE, your disbursement, and the sponsor's group record.
The analogy is the credit bureau score. No sane desk lends on a CIBIL number alone, and no sane desk ignores it: the score screens, the file decides. MahaRERA designed a bureau score for projects and did not publish it, but its entire input data is public, so a desk can compute the score itself and still read past it to the record, which is what the developer-facing view of the same filings urges promoters to assume everyone will do.
Kavya Menon's desk (illustrative, as ever) settled on a one-line policy when the grades failed to appear: the matrix is the checklist, the inputs are the evidence. Every Maharashtra sanction note carries the matrix's inputs as the desk computed them, and no decision cites a score without citing the filed numbers behind it.
ReraGenie serves the inputs: the same filed artifacts the matrix named, assembled per project in the Rs 2,999 project analysis with the filing gaps, extensions and complaints set against the district. The quarter-on-quarter series and the group map stay the desk's own work, because the register keeps only the current filing and records each entity separately. The project pages show each project's construction progress beside its possession timeline, free.
The one-line summary
MahaRERA designed a project grade and paused it before publishing one, which leaves a desk the better half of the idea: the regulator's own list of decision-relevant filed inputs, free to compute every quarter with your own thresholds, as a screen and a covenant trigger that never substitutes for the quarterly reconciliation that catches trouble first.
Methodology and sources
- The matrix: MahaRERA Circular No. 46/2023, 28 August 2023, introducing the MahaRERA Grading Matrix for projects registered after January 2023.
- The pause: a senior MahaRERA official to the Times of India, 6 November 2024, as reported by PropNewsTime.
- Approvals: MahaRERA's figures for 2025-26, as reported in 2026: 10,379 approvals, including 4,204 new registrations and 3,687 extensions.
- Enforcement: MahaRERA show-cause notices to 8,212 projects, May 2026.
- Credit growth: Reserve Bank of India, sectoral deployment of bank credit, January 2026.
- Early-warning classes: Reserve Bank of India, Prudential Framework for Resolution of Stressed Assets, 7 June 2019.
- Register size: MahaRERA's count of 50,162 projects, May 2025; ReraGenie's count of 55,995 published filings, as updated on 29 September 2026.
Underwriting or monitoring Maharashtra real estate exposure?
The ReraGenie project analysis (Rs 2,999 per project) and the area market report (a flat Rs 2,999 per pincode) assemble the filing record for one project or one pincode. For portfolio and underwriting views, write to alerts@reragenie.com and we will shape them with you.
See the reports