How firms get scored.
How the GW GP Score is built: five weighted factors on public evidence, the G1-G6 ladder shared with Gravitywell Ratings, a credit-lens variant for debt platforms, vintage-cycle context, and a point-in-time revision policy.
Five factors · two lenses
Every firm is scored 0-100 as a weighted sum of five factor scores, each set on public evidence and labelled a GW estimate (E). Equity investors and credit platforms are read through different evidence for the same factors: realisation means OFS proceeds for a fund and repayment records for a lender. Weights are identical across lenses so the desks remain comparable.
| Factor | Weight | Equity lens: evidence | Credit lens: evidence |
|---|---|---|---|
| Portfolio quality | 25% | Unicorn/soonicorn density, externally-led up-rounds, loss ratio in the tracked book, quality of the IPO queue. | Borrower quality mix, unicorn-borrower count, visible stressed names in the loan book. |
| Exit realisation | 30% | Realised OFS/block/secondary proceeds and multiples, documented DPI, distribution cadence: the heaviest weight because paper is not performance. | Published loss rates / GNPA (rating-agency data ranks above self-claims), repayment record, recovery on stressed names, realised equity-kicker gains. |
| Sector positioning | 15% | Alignment of the current book and latest cheques with GW sector outlooks and the capital cycle. | Product fit to the cycle (e.g. counter-cyclical lending into a down-round-averse market), borrower-sector concentration. |
| Capital velocity | 15% | Fund cadence, close-vs-target outcomes, deployment pace against the cycle. | Origination pace, liability-side quality (bank lines, DFI capital, ratings), fund closes vs guidance. |
| Franchise stability | 15% | Partner continuity, spin-out risk, succession evidence, LP re-up behaviour. | Same, plus balance-sheet governance (NBFC ratings, board structure) where applicable. |
The ladder: shared with Gravitywell Ratings
Confidence: how sure we are, stated separately
High: audited or regulator-published evidence dominates (listed-company disclosures, rating-agency data, IPO filings).
Medium: tier-1 press and firm statements cross-verified; key figures self-reported but consistent.
Low: young or opaque vehicles: undisclosed fund sizes, unverified claims, pre-deployment firms. Scores here are placeholders that re-rate on facts.
Confidence is shown per firm in the league tables and dossiers. A high score with Low confidence is a hypothesis, not a verdict.
Vintage-cycle context: the macro layer applied to funds
A fund's entry conditions are set the year it deploys, not the year it exits. Every fund vintage in the dossiers carries a cycle-era tag: the same macro read that drives our Capital Cycle Clock and Formation Index, compressed into eras. A 39.5x from a 2009 entry and a markdown from a 2021 entry are the same skill measured in different weather; the tags keep that visible. Editorial mapping, E.
| Vintage | Era | Read |
|---|---|---|
| ≤2013 | Foundation era | Pre-unicorn pricing; discovery-cost entries |
| 2014-16 | First boom | Unicorn discovery; pricing still forming |
| 2017-19 | Expansion | Rational growth vintages |
| 2020 | Covid dislocation | Fear-priced entries: strong vintage in hindsight |
| 2021 | Peak frenzy | Cycle-top entry marks; the vintage still being digested |
| 2022 | Correction onset | Repricing began mid-deployment |
| 2023 | Trough window | Best entry conditions of the cycle |
| 2024-25 | Recovery / IPO window | Exit-led repricing; entry discipline decisive |
| 2026+ | AI repricing | Barbell market: AI premium vs everything else |
Point-in-time policy
Scores never move silently. Every change publishes as a dated score action on the firm's dossier with the reason stated. The current vintage is July 2026; coverage is refreshed monthly.
Watch items resolve in public. Each dossier lists the unresolved, dated questions we could not verify at the vintage. On refresh they are confirmed, corrected or dropped, as dated resolutions, never deletions.
Position books are reconstructions. Built from public disclosures only: representative, not exhaustive. Ownership stakes are omitted because they are not reliably public. Credit exposures are tracked separately from equity positions and say so.
Contested claims stay contested. Where reporting conflicts (a denied partner exit, a disputed fund size), the dossier carries both sides with sources rather than resolving by fiat.
Sourcing. Every figure is sourced and dated. We tier provenance: Primary (official, regulatory, exchange or company filings), Secondary (tier-1 industry research and reputable media), and GW estimate (our own reconstruction or opinion, labelled, never presented as external fact). We prefer primary where it exists, reconcile divergent prints to cited ranges, and hold every number point-in-time: dated, and never silently restated; revisions publish as dated changes.
Fact vs opinion. Facts vs opinion: market sizes, official prints, prices, named deals and agency ratings are sourced facts (Primary/Secondary). Scores, grades, purity weights, scenario paths and indicative sparkline points are Gravitywell's analytical opinion (GW estimate): labelled, not presented as external data.
GW GP Scores are research opinions, not investment advice, not a solicitation, and not an assessment under any SEBI regulation. Dossiers are compiled from public sources believed reliable; firms named did not participate and figures marked E are Gravitywell estimates. Point-in-time: dated to the vintage shown and never silently restated.