Singapore's 90%+ Capture Rate of SEA Funding: Sustainable or Statistical Artifact?

Originally published: 2024-10 | Last verified: 2026-09-13 Capture-rate figures in this article are Tracxn’s domicile-based aggregation as reported by TechNode Global and The Independent, and each is linked inline to the report it comes from. Aggregators differ on this ratio for methodological reasons: which rounds each one counts, when it counts them, and which home market each deal is attributed to. For H1 2025, Tracxn’s first print put the region at about USD 2.0B with Singapore at 92%, while DealStreetAsia counted USD 1.85B with Singapore at USD 1.21B, almost two-thirds; Tracxn later restated H1 2025 to about USD 3.2B. So treat any single capture number as methodology-bound rather than as a fact about the region.

The number that gets quoted most often in Southeast Asia venture coverage is some version of “Singapore captures over 90% of SEA funding.” On Tracxn’s domicile-based aggregation, the reported share was 92% in H1 2025, 91% in H2 2025 and 94% in H1 2026, and 93% in Q1 2026.

What strikes me about that series is how little it moves. SEA funding volume itself is violently cyclical — the Q1 2026 total was up 110% year-on-year and 146% on the prior quarter — yet across these four readings Singapore’s slice barely budges, over periods in which the denominator more than doubled. A ratio that stable through that much volatility is not being set by funding volume quarter to quarter; my read is that it is set by which deals dominate the total and where they are booked — which mostly means where the holding companies are registered. One caution on how far to push that: the band is a 2025-26 feature. Tracxn’s December 2024 annual put Singapore at nearly 67% of 2024 funding, so the ratio can move — and its jump coincided with late-stage rounds, which tend to book through Singapore holding companies, coming back to dominate the totals.

The question I want to walk through is not whether the number is correct — Tracxn’s ratio has held within a few points in each release since H1 2025 (it has not republished the share on its restated totals), and a ratio that stable is worth taking seriously. The question is what the number actually measures. Because there are at least three different things it could mean, and they have very different implications for how an LP or a corporate strategist should be reading the SEA narrative.

The number, decomposed

The Tracxn capture rate is a simple-looking ratio: capital raised by Singapore-headquartered companies divided by total capital raised by SEA-headquartered companies. Both numerator and denominator are domiciliation-based. If a company is incorporated in Singapore, its funding rounds count as Singapore funding. If it’s incorporated in Indonesia or Vietnam, they count as that country’s funding.

Three layers sit beneath the headline ratio. No aggregator publishes this split — Tracxn and DealStreetAsia both report the ratio whole — so the weightings below are my own working estimates from deal-by-deal reading, not measured shares. Treat them as a way to structure the question, not as data.

Layer 1: Pure-play Singapore companies. Companies founded in Singapore, with operations primarily in Singapore, serving primarily the Singapore or global market. Think Trax (retail tech), Carousell (marketplace). Where you draw this layer’s boundary — is a Singapore-founded company selling only into Indonesia still pure-play? — moves its size more than any other judgement call in the exercise. This is the genuine Singapore innovation economy.

Layer 2: SEA-regional companies headquartered in Singapore. Companies whose business is meaningfully cross-border SEA — Indonesia, Vietnam, Philippines, Thailand operations — but whose holding entity sits in Singapore for legal, tax, and capital-raising reasons. Grab, Sea Group, Lazada, Ninja Van (regional logistics). This is, on my reading, the largest contributing category by some distance — it is what the headline ratio mostly consists of. This concentration is the geographic leg of the broader SEA funding reallocation.

Layer 3: Non-SEA companies routed through Singapore. Indian, Chinese, occasionally European companies whose primary business is outside SEA but whose Asia regional entity is incorporated in Singapore for fund-raising, regulatory, or treasury reasons. This category is structurally the smallest of the three, but on my reading it is growing, and it is where most of the methodology debates focus.

The headline 91-94% number is the sum of these three layers. The interesting analytical question is what it would look like decomposed and re-aggregated — because the three layers mean different things. Layer 1 is Singapore’s own economic activity. Layer 2 is real regional business booked in Singapore — and it is where the readings below split, since it can be counted either as hub activity or as geographic misattribution. Layer 3 is closer to pure routing. Nobody publishes that decomposition, which is precisely why the ratio gets quoted so confidently and understood so poorly.

Why this matters: three different reads of the same number

Depending on which layer you weight, the same 91-94% number tells three different stories.

The “Singapore is the regional hub” read. If you count Layer 2’s booking as genuine hub activity alongside Layer 1, the capture ratio reflects a genuine economic phenomenon: Singapore has accumulated the regulatory infrastructure, talent density, and capital-market depth that make it the natural domicile for any SEA-scale ambition. The high capture ratio is sustainable because it reflects underlying structural advantages that aren’t going to reverse. Indonesia, Vietnam, the Philippines have larger consumer markets, but the capital allocation function — the LP relationships, the fund admin infrastructure, the listing optionality — sits in Singapore and will continue to.

The “Singapore is a measurement artifact” read. If you treat Layer 2’s holding-company booking as geographic misattribution rather than hub activity, and add Layer 3’s routing on top, the capture ratio is partly an accounting illusion. A USD 100M Series C raised by an Indonesian fintech with a Singapore HoldCo — a Layer 2 case — gets booked as $100M to Singapore, not $100M to Indonesia. The economic activity, the customer base, the engineering team, and the eventual IPO listing might all sit in Indonesia. Under this read, the capture ratio overstates Singapore’s true ecosystem weight by a meaningful margin, and the policy implications for SEA economic development are different.

The “Singapore is a temporary equilibrium” read. Under this view, Layers 1 and 2 are real but the equilibrium is contingent on three things: (a) the absence of credible regional competitors for capital-allocation infrastructure (Hong Kong is the obvious candidate but is constrained by its current macro situation), (b) continued open-flow access to global LP capital through Singapore’s regulatory framework, and (c) tax-incentive maintenance under the 13O/13U regimes and the broader fund-vehicle ecosystem. If any of these shift — particularly if Indonesia builds a credible domestic capital-raising stack — the ratio could compress within five years.

I find the “temporary equilibrium” read the most useful framework. It accepts the data, doesn’t romanticize Singapore’s position, and treats the dominance as a structural fact whose half-life is measurable rather than infinite.

We don’t track Singapore exposure separately from SEA exposure. The two have collapsed into the same category for fundraising purposes, which is convenient, but I’d be lying if I said it didn’t worry me.— LP allocator, US-based fund-of-funds, Asia mandate

What the country-level breakdown actually looks like

The 91-94% number compresses out the country-level texture, which is where the structural dynamics are visible. For Q1 2026, Tracxn put the SEA total at USD 2.8B, a 110% year-on-year rise driven by a handful of large rounds — and a single deal, data-centre operator DayOne’s USD 2B Series C, was larger than the entire year-on-year increase of about USD 1.5B. The geography split it published alongside that total is worth reading closely:

Read against the population and GDP weights of these markets, that table is striking — and the most telling line in it is the one that is missing. Indonesia is roughly 41% of SEA’s population and about 35% of its GDP (World Bank, 2024), yet Tracxn’s Q1 release named no Indonesian city at all — nor did its H1 2026 release, whose next-largest cities after Singapore were Bangkok and Kuala Lumpur; a gap I unpack in the Indonesia discount. The same is true of Vietnam and the Philippines, both of which have populations many times Singapore’s; Vietnam’s deal quality in particular runs well ahead of its capital share.

This gap is the most important framing for an LP or strategist. The capital is concentrating in Singapore not because the underlying markets are concentrated there, but because the capital-allocation function has consolidated there. That’s a different kind of dominance — and a more fragile one.

What would compress the ratio

Three structural shifts could compress the capture rate over a five-year horizon, ordered from most to least likely.

Indonesian domestic capital-raising stack maturation. As Indonesian institutional capital — sovereign-adjacent funds, BUMN-affiliated investment vehicles, the second-tier domestic VC firms — increases its participation in Indonesian deals, the share of those deals that needs to route through Singapore for capital-raising purposes drops. It is not visible in the funding data yet — Indonesia-domiciled companies raised about 1% of Q1 2026 capital on DealStreetAsia’s monthly counts, as I work through in the Indonesia discount — but over five years it is the shift most likely to move the capture ratio materially, though I have not seen anyone put a defensible number on how far.

Hong Kong family office and capital infrastructure recovery. Hong Kong’s family-office count has been rising again under the FamilyOfficeHK program, and the listing optionality through HKEX remains attractive for China-adjacent businesses. If HK rebuilds the capital-allocation infrastructure for Asia mandates, the share of mainland-China-adjacent deals that route through Singapore versus Hong Kong rebalances. This is a slower shift, but the directional pressure is real.

MAS substance regime tightening. If MAS tightens the substance bar for Singapore-domiciled funds and corporate vehicles further — beyond the 2025 changes — some of the marginal Layer 3 routing-artifact volume becomes uneconomic and shifts to other domiciles. This is the smallest of the three effects but the easiest to model, since it’s policy-driven rather than market-driven.

The practitioner takeaway

For an LP or corporate strategist building exposure to SEA, the 90%-plus capture rate is the wrong number to navigate by. It tells you where the capital is being raised. It does not tell you where the economic activity is happening, where the operational risk is concentrated, or where the value is being created.

A better dashboard tracks (a) deal count by country of operation rather than country of HoldCo, (b) per-deal weighted-average customer geography for the financed companies, and (c) talent density in each operating market. Singapore continues to dominate by all the financial-routing metrics. By operational and talent metrics, the regional picture is likely to look far more distributed — which is exactly what the funding ratio cannot tell you.

The funding number is real. It just measures something narrower than the practitioner narrative usually implies. Read it for what it is, weight it accordingly, and the rest of the SEA story becomes easier to interpret.