Capital Gravity Index.
How tightly India's capital clusters, into a few funds, sectors, cities, and mega-deals.
The headline index
Capital is clustering, not spreading. Three cities took 79% of 2025 startup money; five sectors took 72%. The index sits at 115: gravity intensifying 6% over the year.
The 5 lenses
Latest reads anchor to official 2025 releases (cited per lense).
Share of startup capital landing in just three metros: the geographic pull of the ecosystem.
Source ↗ Inc42 / Business Standard, 2025Concentration of deal value in the five hottest sectors: how narrow the thesis has become.
Source ↗ Entrepreneur India / Bain VC, 2025Share of capital in $100m+ rounds: how much flows to a handful of category leaders.
Source ↗ Bain India VC / Kalviro, 2025Share of fresh fund capital captured by the ten largest vehicles: concentration at the LP layer.
Source ↗ Bain India PE/VC, 2025A Herfindahl read across the allocation map: the single-number measure of how lumpy capital has become.
Source ↗ Gravitywell (composite of above)Smoothed, rebased, winsorised, weighted.
How it's built
City, sector, deal-size, fund, and a Herfindahl roll-up: concentration measured at every layer of the capital stack, not just one.
Each pillar is a concentration share. We smooth on a 3-month average and rebase to 100 at mid-2024 so a rising index unambiguously means tightening.
City 25% · sector 20% · mega-deal 20% · top-10 fund 20% · HHI 15%. Geography and sector carry the most weight: they move the narrative.
Sub-indices are clamped to a [50, 180] band so no single lens (a one-quarter mega-deal spike) can overstate concentration.
Rising gravity with falling deal count = capital pooling into fewer hands. That is the signal allocators and founders both watch.
- · Mega-deal lumpiness: one jumbo round can spike a quarter; winsorisation and the 3-month average dampen it.
- · Definitional drift: 'sector' taxonomies shift; the universe and buckets are fixed at period start.
- · Share ≠ size: a rising share on a shrinking base still means concentration, and is flagged as such.
- · Survivorship: only funded entities appear; the unfunded long tail is invisible by construction.
Data vintage June 2026. Latest reads anchor to official 2025 releases: Inc42 (city), Entrepreneur India / Bain (sector & fund), and Bain India VC (mega-deals). The Herfindahl roll-up is Gravitywell's composite of the underlying shares; the within-period monthly distribution is a reconstruction that reconciles to the cited primary sources.
Methodology v3.2 (2026-07). Built to the OECD/JRC composite-indicator handbook and disclosed toward the IOSCO Principles for Financial Benchmarks: distance-to-reference normalisation, 3-month smoothing, a flagged contribution cap, weighted aggregation, plus a drop-one-pillar uncertainty band, an equal-weight robustness cross-check, and a data-coverage ratio (all shown above). Known limitation: the 24-month panel is too short for robust seasonal adjustment; India's March fiscal-year-end spikes are not yet removed. Series are point-in-time; published values are not silently restated.
Where capital pools, monthly.
The Capital Gravity Index and its city, sector, and fund breakdowns update every month.