Asia-Pacific Investment Report 2025 — Different Bifurcations

The 2025 global report describes a market that split in two: abundant capital for one segment, scarcity for everything else. That bifurcation appeared across Asia-Pacific too, but along different lines in each market — and the differences are more informative than the similarity.

In the US, the axis was artificial intelligence. Companies with a credible AI position experienced abundant capital; others did not.

In several Asia-Pacific markets, the axis was different:

• In some, it was profitability. Following the exit constraint and the funding contraction, capital concentrated toward companies with demonstrated unit economics regardless of sector. The distinction was not what a company did but whether it made money.
• In others, it was domestic versus export orientation, as trade and technology policy made export-dependent businesses harder to underwrite.
• In others, it was alignment with state priorities, where policy direction determined capital availability more reliably than commercial characteristics.
• In the manufacturing economies, it was position in the AI supply chain, per the 2024 report — a hardware axis rather than a software one.

The general observation is the useful one: the axis of a bifurcation reveals what a market's capital is selecting for. A market that bifurcates on AI is selecting for thematic exposure. One that bifurcates on profitability is selecting for capital efficiency. One that bifurcates on policy alignment is selecting for regulatory safety. These are different markets with different risk characteristics, and a regional aggregate blends all of them.

A structural development underlay all of it: domestic capital grew as a share of regional funding. Local institutional investors, corporate venture arms and sovereign vehicles became more significant relative to foreign capital — which changes both the stability of funding and what gets funded.

Reading the axis

The idea that the axis of a bifurcation is diagnostic deserves setting out, because it turns an observation into an analytical tool.

A bifurcated market has capital concentrating on one side of some dividing line. The line is chosen — implicitly, by the aggregate of investor decisions — and it reveals what investors believe is the most important distinction between companies.

What each axis implies:

• A thematic axis (AI) implies investors believe sector exposure matters more than company quality. This is a bet on category, and it produces high dispersion within the favoured category, since capital flows to the theme rather than to the best companies in it.
• A profitability axis implies investors have been disciplined by an exit constraint and are underwriting to cash generation rather than to a future sale. This produces lower dispersion and lower returns, and it is the healthier configuration if less exciting.
• A policy alignment axis implies regulatory risk dominates commercial risk in investors' assessments. This produces capital allocation that tracks policy rather than opportunity, with the associated risk that policy changes.
• An export orientation axis implies investors are pricing trade and geopolitical risk as a primary variable.

Each axis has different implications for the returns available:

• A thematic axis produces the highest dispersion and the greatest risk of overpaying, since price is set by category demand rather than by company assessment.
• A profitability axis produces the most durable businesses and the least exciting returns.
• A policy axis produces returns that depend on a variable investors cannot forecast well.

The interesting question about a bifurcated market is not which side is winning. It is what the dividing line is, because that tells you what capital in that market is actually underwriting.

For an allocator with regional exposure, this means the segment matters more than the country. A profitable, domestically-oriented, policy-aligned company in a difficult market may face better conditions than an unprofitable, export-dependent company in a favoured one.

Count versus value, inverted

A structural difference between the region's AI exposure and the US pattern appeared clearly in 2025 and is worth stating precisely.

The US pattern, per the 2025 US venture report: AI took a larger share of deal value than of deal count. Capital concentrated into fewer, much larger rounds — reflecting the capital intensity of infrastructure and model development.

The pattern in several Asia-Pacific markets was inverted. AI accounted for a larger share of deal count than of deal value — many companies raised AI-related rounds, but the rounds were smaller.

What that inversion means:

• The regional AI activity was concentrated at the application layer, which requires far less capital than model or infrastructure development.
• Frontier model development was largely not happening at venture scale in most of the region, being either absent or funded by corporate and state capital outside the venture market.
• The exposure was to AI adoption rather than AI creation — building on top of models rather than building them.

Whether this is favourable is genuinely unresolved, and it depends on the value accrual question the 2025 AI report frames:

• If value accrues to the application layer, the regional position is favourable — exposure to the winning layer at much lower capital intensity.
• If value accrues to models or infrastructure, the regional position missed it.

The capital efficiency point is separately worth noting. Regional AI companies, funded at smaller sizes, need smaller outcomes to produce good returns. A company that raised a modest round and reaches a modest exit can be an excellent venture outcome — while a company that raised at multi-billion valuations needs an exceptional one. The bar for success scales with the capital raised, which means the regional pattern produces a wider range of acceptable outcomes.

The Singapore Venture Capital Report 2025 documents this pattern with specific figures and is the archive's best-sourced illustration of it.

Domestic capital and what it funds

The growth of domestic capital as a share of regional funding is a structural change whose consequences extend beyond stability.

The sources that grew:

• Domestic institutional investors — pension funds, insurers, sovereign vehicles — increasing allocations to private markets in their own regions.
• Corporate venture arms of the region's large technology and industrial companies.
• Family offices and domestic wealth, growing with regional wealth accumulation.
• Government-backed funds, established explicitly to develop domestic ecosystems.

Why this changes more than the funding source:

• Stability improves. Domestic capital is less likely to withdraw during a global risk-off episode, because its liabilities are local. This directly addresses the vulnerability the 2015 and 2022 reports describe.
• Sector preferences differ. Domestic capital frequently favours categories serving local markets over those aiming at global scale, since local knowledge is an advantage in the first and not the second.
• Return expectations differ. State-backed and corporate capital may accept lower financial returns for strategic or developmental objectives, which affects pricing — the same dynamic the 2016 US venture report describes for corporate venture at scale.
• Governance expectations differ, which affects how companies are built and what standards they are held to.

The consequence for foreign investors is a competitive change. Domestic capital with local knowledge, local networks and potentially lower return requirements is a formidable competitor for the same opportunities. Foreign capital's historical advantages — capital availability and experience from more developed markets — matter less as domestic ecosystems mature.

The practical implication is that foreign participation increasingly requires a differentiated offer rather than capital alone: cross-border expertise, access to international markets, or specialist sector knowledge. Capital by itself is no longer scarce in most of the region's major markets, which is a genuine change from the situation described in the 2015 report.

India and Japan diverge

Two markets diverged from the region's cycle for structural reasons, continuing the pattern the 2023 report describes.

India continued its structural case — demographics, domestic demand, digital infrastructure, deepening exit markets. The divergence from the regional cycle reflects that its drivers are largely domestic. An economy whose growth depends on internal demand is less affected by global trade conditions and by the technology supply chain politics dominating other markets.

The qualification from the 2023 report remains. The structural case was substantially priced. Strong fundamentals at high valuations produce lower forward returns than the same fundamentals at lower ones, and by 2025 the entry point question was more pressing than the allocation question.

Japan continued its governance reform trajectory, which is similarly disconnected from regional conditions. A return source based on improving capital efficiency does not depend on growth, trade or technology cycles.

The general observation is that both markets offered something the rest of the region did not: a return driver uncorrelated with the dominant regional variables. In a region where most markets are exposed to trade, technology supply chains and the dollar cycle, exposure to Indian domestic demand or Japanese capital efficiency is genuinely diversifying.

That is the strongest argument for the unbundled allocation approach the 2023 report describes. A regional allocation averages these away; separate allocations preserve them — and preserving genuinely uncorrelated return drivers is the entire point of diversification.

What domestic capital changes beyond stability

The growth of domestic capital is usually discussed in terms of resilience. Its effects on what gets funded are less examined and are arguably more consequential for the region's development.

Sector preferences shift toward domestically-oriented businesses. A domestic investor has an informational advantage in assessing companies serving local customers — they understand the customer, the regulatory environment and the competitive set without a learning cost. Foreign capital, applying imported frameworks, systematically favoured categories recognisable from other markets. The categories that domestic capital funds are frequently ones foreign capital had underweighted, and some of them are large.

Return expectations differ by source. State-backed and corporate capital may accept lower financial returns for strategic or developmental objectives. That supports businesses a purely financial investor would decline and affects pricing in the categories where such capital is significant — the same effect the 2016 US venture report describes for corporate venture and the 2022 Gulf report describes for strategic capital.

Time horizons lengthen. Family office and corporate capital is not bound by a ten-year fund life. That permits backing businesses whose path to scale is longer than a fund structure tolerates, which is a genuine expansion of what is fundable rather than merely a change in who funds it.

Governance expectations change. Domestic investors bring different norms about reporting, board composition and founder control. Whether this is an improvement varies; that it is a change is not in doubt.

Exit route preferences shift toward domestic listing. An investor who is themselves domestic, whose LPs are domestic, and who understands the domestic exchange will favour a domestic exit — which reinforces the development of that route, as the 2021 and 2023 India reports describe. The public and private markets develop together.

Domestic capital does not merely replace foreign capital in the same deals. It funds different deals, on different terms, toward different exits — and that reshapes an ecosystem rather than just stabilising it.

What an allocator could act on

Read the bifurcation axis as diagnostic. A market bifurcating on theme is selecting for category exposure; one bifurcating on profitability is selecting for capital efficiency; one bifurcating on policy alignment is selecting for regulatory safety. These are different markets with different risk characteristics, and a regional aggregate blends them.

Compare AI's share of deal count with its share of value. In the US, value share exceeded count share — capital concentrating into fewer, larger rounds at the infrastructure and model layers. In several regional markets the inverse held, indicating application-layer activity at much lower capital intensity. Whether that is favourable depends on the value accrual question, which is unresolved.

Note that the bar for success scales with capital raised. A company funded at a modest size needs a modest outcome to produce a good return; one funded at multi-billion valuations needs an exceptional one. The regional pattern produces a wider range of acceptable outcomes, which is a genuine structural advantage that headline funding comparisons obscure.

Recognise that capital alone is no longer a differentiated offer. In most of the region's major markets, foreign participation now requires cross-border expertise, access to international markets or specialist sector knowledge. Domestic capital with local knowledge and potentially lower return requirements is a formidable competitor for the same opportunities.

Value uncorrelated return drivers explicitly. India's domestic demand orientation and Japan's capital efficiency case are driven by variables unrelated to the region's dominant ones — trade, technology supply chains, the dollar cycle. A regional allocation averages those away, which is the strongest practical argument for the unbundled approach and one that is easy to state and hard to implement.

Check the axis before comparing two markets' funding data. Two markets can report similar totals while selecting for entirely different things — one for thematic exposure, one for demonstrated profitability. The totals look comparable and the risk characteristics are not, which makes a cross-market comparison on funding alone actively misleading.

What 2025 established for Asia-Pacific

• The bifurcation axis was shown to be diagnostic of what a market's capital is selecting for.
• The count-versus-value inversion revealed the region's AI exposure to be at the application layer, at much lower capital intensity.
• Domestic capital growth changed competitive conditions, requiring foreign investors to offer more than capital.
• India and Japan provided genuinely uncorrelated return drivers, which is the strongest case for unbundled allocation.
• The bar for success scales with capital raised, making the region's smaller rounds a wider-tolerance proposition.