Southeast Asia’s startup ecosystem is not
suffering a capital crisis — it is suffering an imagination crisis. The Business Times reported on 19th
October 2026 that debt funding for Southeast Asian startups has grown as equity
options shrink. Debt funding in the
region rose from US$487.8 million in 2023 to US$3.1 billion in 2024. It nearly doubled again to US$6.1 billion in
2025. Some called this a maturing
market. Others claim it is a global
trend. Both positions are somewhat correct. Neither says what it means.
Debt is what companies reach for when the
equity case has collapsed, but the lights need to stay on. Calling that maturity is the financial
equivalent of calling bankruptcy proceedings a liquidity management
exercise. The shift from equity to debt
across Southeast Asian startups is not a sign of a sophisticated ecosystem
finding its footing. It is a sign that
institutional equity investors have looked at what is on offer and decided the
risk-return ratio does not work. They
are right. The question worth asking is
why.
The Capital Exists; the Conviction Does Not
Singapore closed 472 venture deals worth
US$4.6 billion in 2025 — a 34 per cent drop in value from the year before,
according to EnterpriseSG’s Singapore Venture Funding Landscape report. That figure was the third consecutive year of
decline from the pandemic-era peak of US$10.6 billion across 1,058 deals in
2022. Seed funding collapsed from
US$810.9 million in 2023 to US$443.2 million in 2025. Singapore led the regional contraction. It recorded one final fund close in the
venture space in 2025. One. Against ten in the prior year and 21 in 2023.
The money did not go anywhere. GIC manages US$936 billion. Temasek Holdings reports a net portfolio
value of S$518 billion as of March 2026.
Singapore-domiciled family offices, per MAS estimates, hold hundreds of
billions more in deployable capital. 70
per cent of LP commitments in 2024 and 2025 went to existing GP relationships,
up from 60 per cent five years earlier, per Preqin data. Of every US$100 of institutional PE capital
available, US$10 to US$15 was accessible to a manager without an established LP
base. The capital has retreated to
familiar names funding familiar ideas.
This is not selectivity. It is
institutional risk aversion with a new vocabulary.
Late-stage deals accounted for 33.3 per
cent of total deal volume in Singapore in 2025, up from 26.9 per cent in
2024. The GPs are funding companies that
are already large rather than the ones that could become large. Locking capital into illiquid assets at
compressed multiples in a market that produced four PE-backed IPOs in an entire
region in a calendar year is not conservatism.
It is a strategy with no exit.
The Embarrassing Comparison
The United States deployed US$193.2 billion
into venture-backed startups in 2025, according to CB Insights data. Asia, including China, India, and Southeast
Asia combined, raised US$56.4 billion.
Southeast Asia alone raised US$2.2 billion in 2024 — a figure that would
not fund a single mid-tier US tech company’s Series C. The gap is not merely financial. It is structural.
Silicon Valley’s top-tier firms operate
under a specific thesis: back founders who are trying to change the world, not
fix a broken supply chain in one of ten markets with different regulatory
regimes. Andreessen Horowitz manages over
US$90 billion in assets under management.
Sequoia raised US$950 million across two vehicles in October 2025
alone. Lightspeed closed over US$9
billion across six vehicles in December 2025.
These are not funds. They are
sovereign-scale capital deployment machines attached to networks of founders,
engineers, and operators that Southeast Asia cannot replicate.
The exit infrastructure explains the
rest. US startups have access to NASDAQ
and the NYSE — the deepest technology equity markets in human history, with a
combined daily trading volume that dwarfs Southeast Asia’s entire annual
venture funding. Google, Microsoft,
Meta, Apple, and Amazon collectively acquire dozens of startups annually,
providing the M&A liquidity that sustains the entire VC ecosystem. Southeast Asia has the Indonesia Stock
Exchange, the SGX, and the Vietnam HOSE.
The region logged 31 public listings and 81 acquisitions across the
first seven months of 2026. The entire
region. In seven months.
There is also the talent drain. Since 2025, venture firm Antler alone has
helped more than 30 Asian founding teams relocate to the United States. The region’s most ambitious founders — the
ones who want to build globally rather than compete for market share in a
single Southeast Asian country — have calculated that they cannot raise serious
capital from Singapore if they are still in Singapore. They are correct. They are leaving.
The B2B software category tells the story
most starkly. One prominent investor
noted that there is “very little B2B software in Southeast Asia — almost none
of it — and if I look at our US portfolio, all of our big winners are in that
category.” The absence is not
accidental. Building enterprise software
requires a customer base willing to pay for it, a sales motion sophisticated
enough to close enterprise deals, and an exit market with strategic acquirers. Southeast Asia has none of these at
scale. Silicon Valley has all three.
The Incumbency Problem
The second failure is not about
capital. It is about competitive
reality. Most of what is being funded in
Southeast Asia is not a whitespace opportunity.
It is an attempt to occupy a niche that larger, better-capitalised
entities will take over the moment the thesis is proven. Grab Holdings holds over 70 per cent of
Southeast Asia’s ride-hailing market. In
Malaysia and the Philippines, it exceeds 90 per cent. Grab posted full-year 2025 revenue of US$3.37
billion and its first net profit of US$200 million — the first in its fourteen-year
history — after burning through over US$12 billion in cumulative capital. Any startup that enters ride-hailing, food
delivery, or digital financial services in Southeast Asia today is not
competing with other startups. It is
competing with Grab, GoTo Financial, SeaMoney, ShopeePay, and the digital
banking arms of the three largest state banks in every market it operates.
The lesson from Vietnam is
instructive. Gojek — having burned
billions — exited Vietnam in 2024 after losing market share to two opponents it
should have been able to outlast. Xanh
SM, a subsidiary of Vietnamese conglomerate Vingroup, entered ride-hailing with
artificially low prices and government-connected advantages that no startup
could replicate regardless of its funding level. It captured 32 per cent of the Vietnamese
market. Be Group, a local operator with
a fraction of Gojek’s capitalisation, captured 24 per cent. Gojek had 22 per cent and was falling before
it gave up. Capital does not beat
structural advantage. It never did.
This is what Singapore’s VCs and their LPs
persistently fail to account for. The
large domestic conglomerates, the government-linked entities, and the incoming
wave of well-capitalised foreign operators will enter any market the moment a Southeast
Asian startup proves it is worth entering.
The startup will have spent three to five years and US$50 to US$200
million demonstrating the thesis. The
incumbent will arrive with ten times the resources and capture the market in
eighteen months.
Exhibit A: Temasek Holdings
No institution better illustrates the
failure of imagination and due diligence in Southeast Asian venture investing
than Temasek Holdings — Singapore’s own sovereign wealth fund, the institution
that is supposed to represent the gold standard of institutional investing in
the region. Temasek Holdings’
early-stage investment plummeted from US$4.4 billion in 2021 to US$509 million
in 2024, and a mere US$70 million in the first months of 2025, according to
Tracxn data. First-round investments in
unlisted companies fell from 82 in 2021 to 11 in 2024. The world’s fourteenth-largest sovereign
wealth fund has effectively exited the early-stage venture game — not because
of strategic recalibration, but because of a sequence of high-profile failures
that exposed the depth of its due diligence problem.
Case One: FTX — US$275 Million into a Lie
In October 2021, Temasek invested US$210
million for 1 per cent of FTX International.
In January 2022, it invested a further US$65 million for 1.5 per cent of
FTX US. Total exposure: US$275 million. FTX filed for bankruptcy in November
2022. Its founder, Samuel Benjamin
Bankman-Fried, is currently serving a 25-year prison sentence for fraud. Temasek Holdings wrote down the entire US$275
million. Sequoia Capital and SoftBank
did the same. Temasek Holdings’ defence
was that its “extensive due diligence” had relied on audited financial
statements showing FTX to be profitable, and that the fraud was “intentionally
hidden from investors.” Both statements
are true. Neither is exculpatory. Sophisticated institutional investors are
supposed to identify what audited statements cannot capture — the character of
management, the control environment, the governance structure, the alignment of
incentives. Holdings held no board
seat. It conducted due diligence over
eight months and concluded that a company run by a 29-year-old who slept on
bean bags in the Bahamas and played video games during investor calls was a
sound institutional investment.
Then-Finance Minister Lawrence Wong Shyun
Tsai stated publicly that the FTX loss caused reputational damage. Temasek Holdings cut the compensation of the
investment team and senior management responsible. Then it moved on. The lesson apparently learned: reduce
early-stage exposure. The lesson that
should have been learned: do not confuse narrative charisma for investment
substance.
Case Two: Zilingo — S$297.5 Million into a
Management Catastrophe
Zilingo was a Singapore-based fashion
e-commerce platform founded in 2015 by Ankiti Bose and Dhruv Kapoor. It raised US$308 million in total from
investors including Temasek Holdings, Sequoia Capital India, Burda Principal
Investments, and EDBI. Temasek Holdings
was among its top five shareholders, holding 8.3 per cent of its shares as of
June 2022. Zilingo had not filed a
financial statement since 2019.
Investors, including Temasek Holdings and EDBI, continued funding the
company through 2020 despite this elementary governance failure. In March 2022, whistleblowers approached the
Zilingo board reporting unexplained payments approved by Bose to companies with
no operational connection to Zilingo.
Bose was suspended on 31st March 2022 and fired on 20th
May 2022 after a forensic investigation.
The company filed for liquidation in January 2023.
The red flags were visible long before the
collapse. Other potential investors had
declined to back Zilingo after finding evidence of merchant fraud in Indonesia
— which accounted for more than half of Zilingo’s GMV in FY2021. Temasek Holdings and EDBI continued ploughing
money in. The company’s governance was
sufficiently disordered that it had not filed a statutory account in three
years. Temasek Holdings, which is bound
by the same Companies Act obligations as every other Singapore-incorporated
entity, apparently did not treat this as a material concern.
To be precise: as of October 2026, no court
has convicted Ankiti Bose or her co-founders of any offence connected with
Zilingo, and the full facts remain disputed.
What is not disputed is that Temasek Holdings funded a company that had
not filed accounts since 2019, which then collapsed into liquidation with total
investor losses exceeding US$308 million.
Case Three: eFishery — US$300 Million into
Fictitious Fish
eFishery was an Indonesian aquaculture
technology company that deployed AI-powered feeders for fish and shrimp
farmers. It was valued at over US$1
billion — a unicorn — before the fraud allegations emerged. Temasek Holdings invested alongside SoftBank
and Sequoia Capital India across multiple rounds from 2021 onward. FTI Consulting’s investigation found that
eFishery had falsified its accounting since at least 2021. For the first nine months of 2024, the
revenue of three eFishery subsidiaries was inflated fivefold. US$596 million of revenue was
fabricated. The company reported US$752
million in first-nine-months 2024 revenues.
It actually earned US$157 million.
FTI estimated that as of February 2025, eFishery had approximately US$50
million in cash and was not commercially viable in its current form. Investors, including Temasek Holdings, were
told they could expect to recover less than ten cents for every dollar
invested.
The due diligence process was
instructive. The firms conducting
verification used eFishery’s own database of farms to identify which farms they
planned to visit — allowing management to prepare the sites in advance. Only a handful of random spot inspections
were conducted. Mak Yuen Teen, Professor
of Accounting at the National University of Singapore, noted on LinkedIn with
commendable understatement: “Temasek didn’t think this may be fishy?”
The wider list of Temasek Holdings’ failed
investments includes gene therapy company Locanabio, which ceased operations;
Pear Therapeutics, a digital health company that filed for bankruptcy in the
United States; and Tessa Therapeutics, a Singapore-based biotech that ceased
development.
The Pattern
FTX.
Zilingo. eFishery. Locanabio.
Pear Therapeutics. Tessa
Therapeutics.
These are not isolated bad bets in a
diversified portfolio. They are a
pattern. The pattern is: an institution
of enormous resources and theoretical sophistication, deploying capital at the
peak of a valuation cycle, into companies with inadequate governance,
insufficient financial controls, and in some cases active fraud, based
primarily on narrative momentum and the endorsement of other prominent
co-investors rather than independent analytical rigour.
Temasek Holdings’ response to this pattern
has been to cut early-stage exposure from US$4.4 billion to US$70 million and
cap early-stage investments at 6 per cent of the portfolio. This is the institutional equivalent of
burning your hand on a stove and deciding never to cook again. The problem was never early-stage
investing. The problem was the quality
of the due diligence and the susceptibility of the investment team to the
narrative that every prominent Southeast Asian founder with a compelling story
and co-investors with recognisable names was worthy of sovereign wealth fund
capital.
Silicon Valley Does It Differently
The contrast with the American venture
ecosystem is not primarily about capital quantum, though the quantum
matters. It is about institutional
architecture. Andreessen Horowitz built
a portfolio support infrastructure — legal, marketing, recruiting, technical —
that functions as a management consultancy for its portfolio companies. Sequoia’s Scout programme identifies pre-seed
opportunities through a network of over 100 scouts embedded across the
technology industry. Founders Fund’s due
diligence on Anduril — into which it wrote US$1 billion in June 2025 — rested
on years of relationship, deep technical assessment, and a thesis about the
defence technology market that was developed over a decade before the cheque
was written.
These firms do not fund companies because
they are already famous, because the co-investors are reputable, or because the
founder gives a compelling presentation.
They fund companies because the underlying thesis is independently
verified, the market is sized rigorously, the management has demonstrated
relevant capability, and the exit pathway is credible.
B2B software — the category that has
produced most of Silicon Valley’s large exits — requires this discipline. The customer is a business, not a
consumer. The sale requires a proof of
concept, a pilot, a procurement process, and a multi-year contract. The discipline of selling to businesses that
scrutinise every line item is the best filter for investment quality that the
venture industry has ever produced.
Southeast Asia does not build B2B software because the institutional
infrastructure to sell, buy, and exit it does not exist at scale. Until it does, the region will continue
building consumer products and marketplace models that consolidate into the
hands of incumbents the moment the market is proven.
What Should Change
The Business Times article frames the
debt-equity shift as a consequence of macroeconomic conditions and a maturing
market. It is not. It is a consequence of years of capital deployed
into the wrong companies and industries by institutions that lacked the
imagination to identify what was genuinely new and the analytical rigour to
avoid what was merely fashionable. The
LP community in Singapore needs to fund managers who have built real sector
expertise — not generalist funds with pan-regional mandates and fintech in the
portfolio because fintech is what everyone else has in their portfolio. The GP community needs to stop treating
co-investor reputation as a substitute for independent due diligence — FTX had
BlackRock and Sequoia on the cap table, which apparently satisfied the
investment committee at Temasek that the investment was sound. It was not.
The founders need to stop building for the
exit and start building for the customer.
The companies that have generated real returns in Southeast Asia — Grab,
after fourteen years and US$12 billion; Sea, after nearly a decade of losses
and a brutal restructuring — did so by building something that a large number
of people needed. The companies that
failed did so by building a narrative that institutional investors wanted to
hear.
The debt funding increase is the ecosystem
telling you something. It is telling you
that the equity case has failed. The
question is not how to solve the financing problem. The question is why the ideas on offer do not
merit the financing being sought. That
question has an answer. The ecosystem
does not want to hear it.
Terence Nunis | Executive Chairman, Equinox Zenith | Author,
The 1% Playbook: The Billionaire Cheat Code
No comments:
Post a Comment
Thank you for taking the time to share our thoughts. Once approved, your comments will be poster.