Saint Clair Global · 마켓 인텔리전스 | 2026년 8월

유럽 우선(Europe First) 시리즈 — 2/4

세계에서 로봇을 가장 많이 쓰는 나라는 한국이다. 제조업 노동자 1만 명당 산업용 로봇 1,220대. 유럽연합(EU) 평균은 231대에 그친다. 다섯 배가 넘는 이 격차가 곧 시장의 크기다. 인구가 줄어드는 유럽은 더 적은 사람으로 더 많이 만들어야 하고, 그 가장 직접적인 해법인 산업 자동화야말로 한국이 20년간 앞서 키워 온 분야다.

1만 명당 1,220대, 세계 1위 한국

국제로봇연맹(IFR)이 쓰는 ‘로봇 밀도’는 제조업 취업자 1만 명당 가동 중인 산업용 로봇 대수로, 규모가 다른 경제도 곧바로 비교할 수 있는 지표다. 2026년 4월 공개된 2024년 통계에서 한국은 1만 명당 1,220대로 세계 1위에 올랐다. 싱가포르가 818대, 독일이 449대, 일본이 446대로 뒤를 이었고, 미국은 307대, EU 27개국 평균은 231대였다.

한국은 유럽 평균의 다섯 배가 넘는 강도로 산업 현장을 자동화한다. 유럽 최강의 제조 강국 독일조차 한국의 3분의 1을 겨우 넘는 수준이다. 이 격차는 근소한 차이가 아니다. 로봇과 제어, 시스템 통합, 그리고 이를 설치하고 유지하는 숙련 인력에 20년간 쏟아부은 투자가 국가적 역량으로 쌓인 결과다.

이 역량이 지금, 자동화 말고는 선택지가 없는 대륙과 마주하고 있다. 유럽은 생산가능인구가 줄고 있어, 국민소득을 유지하려면 노동자 한 명이 내는 산출을 끌어올려야 한다. 제조와 물류에서 그 가장 직접적인 수단이 자동화다. 그렇게 보면 231이라는 숫자의 의미가 분명해진다. 이는 유럽이 이미 설치한 로봇 밀도일 뿐, 앞으로 필요한 수준에는 한참 못 미친다. 231과 그 필요 수준 사이의 격차, 그것이 곧 유럽이 채워야 할 자동화 수요다.

한국 기업의 기회

1만 명당 1,220대에 이르려면 부품과 비전 시스템, 모션 컨트롤, 그리고 이를 조율하는 소프트웨어까지 탄탄한 공급망이 뒷받침돼야 한다. 그 주역의 상당수는 이름난 대기업이 아니라 중견·신생 기업이다. 그리고 이들 앞에는 지금, 드라기 보고서를 통해 ‘자동화 외에는 길이 없다’는 진단을 받은, 그것도 낮은 출발선에서 시작하는 유럽 시장이 열려 있다.

한국의 강점과 유럽의 수요는 막연히 겹치는 게 아니라 구체적으로 맞아떨어진다. 유럽에서 인력난이 가장 심한 현장이, 공교롭게도 한국 로봇이 가장 많은 실적을 쌓은 분야와 정확히 겹치기 때문이다. 자동차·전자 조립, 창고와 물류, 식품 생산, 그리고 고령화로 일손을 구하기 어려워진 서비스 업무가 그렇다. 한국이 이 밀도를 쌓는 과정에서 축적한 기술은 다른 산업 경쟁력보다 국경을 넘어 이전되기 쉽다.

다만 한 가지는 짚고 넘어가야 한다. 로봇 밀도는 이미 설치된 로봇 대수일 뿐, 앞으로의 주문이 어디로 갈지까지 알려 주지는 않는다. 유럽의 자동화 투자는 현지 부품 비율 요건과 표준, 그리고 단순 수출보다 현지에 자리 잡은 기업을 우대하는 현지 파트너의 기대에 좌우될 것이다. 이 격차를 매출로 바꾸는 기업은 인천에서 하드웨어를 실어 보내고 기다리는 곳이 아니라, 유럽에 법인과 통합 파트너십을 세우는 곳이다. 그 구조를 어떻게 설계하느냐는 이어지는 편에서 다룬다.

유럽은 자신의 문제를 이미 진단했다. 로봇 밀도라는 숫자는 그 문제가 어디에 있는지를 정확히 가리킨다. 바로 공장의 자동화 격차다. 세계에서 가장 앞선 산업 자동화 역량은 한국에 있고, 그 역량을 가장 절실히 필요로 하는 시장은 지금 유럽이다. 격차가 클수록, 먼저 들어가 자리 잡는 기업의 몫도 커진다.

출처:

• Robot Density Surges in Europe, Asia, and Americas — 2024년 통계 (국제로봇연맹 IFR, 2026년 4월 8일): https://ifr.org/ifr-press-releases/news/robot-density-surges-in-europe-asia-and-americas

• World Robotics — Industrial Robots (국제로봇연맹 IFR): https://ifr.org/wr-industrial-robots

• 유럽 경쟁력에 관한 드라기 보고서 (유럽집행위원회): https://commission.europa.eu/topics/competitiveness/draghi-report_en

Disclaimer: 본 글은 정보 제공만을 목적으로 한다. 투자 또는 사업 자문에 해당하지 않는다. 모든 의사결정은 독자적 조사와 자격을 갖춘 자문가의 상담을 기반으로 이뤄져야 한다.

About Saint Clair: 세인트 클레어는 유럽과 아시아를 잇는 국경 간 투자 회사다. 자본이 국경을 넘나드는 인프라까지 함께 구축하는 기관 투자자다. Saint Clair Global은 기업을 국제 기관 표준에 맞춰 구축한다. 사업이 구하는 자본 앞에서 성공적이고 매력적인 투자 기회로 읽히게 만드는, 그 실천의 기업 측면이다. 2016년 이래로.

Saint Clair Pte. Ltd. · 싱가포르 · saintclair.global

Saint Clair Global · Market Intelligence | August 2026

Europe First — Part 2 of 4

The most robot-dense country on earth is Korea: 1,220 industrial robots for every 10,000 manufacturing workers, against an EU-27 average of 231. That fivefold gap is the size of a market. Europe’s workforce is shrinking, so it must produce more with fewer people — and industrial automation, the most direct way to do that, is the capability Korea has spent two decades building.

The Gap on the Factory Floor

The International Federation of Robotics (IFR) measures robot density as operational industrial robots per 10,000 manufacturing workers, a figure that lets economies of different sizes be compared directly. On its 2024 data, published in April 2026, Korea ranks first in the world at 1,220. Singapore follows at 818, Germany at 449, Japan at 446 and the United States at 307. The EU-27 average is 231.

Korea automates its industrial base at more than five times the European average, and Germany — Europe’s strongest manufacturing economy — sits at barely a third of Korea’s level. This is not a narrow difference. It is twenty years of sustained investment in robotics, controls, systems integration and the skilled workforce that installs and maintains them, compounded into a national capability.

That capability now meets a continent with no choice but to automate. Europe’s working-age population is shrinking, and holding national income steady means raising output per worker — for which, in manufacturing and logistics, automation is the most direct instrument. That is what turns the 231 into a demand signal. It is only what Europe has already installed, and it sits far below what a shrinking workforce will require. The gap between that 231 and the level Europe must reach is the automation it has yet to buy.

Korea’s Opening

A country does not reach 1,220 robots per 10,000 workers without deep supplier networks in components, vision systems, motion control and the software that coordinates them — many of them mid-sized and younger firms, not the large conglomerate names. They face a European market that the Draghi report has told, in plain terms, that it must automate, and that starts from a low base.

The overlap is specific, not general. Europe’s sharpest labour shortages sit exactly where Korean robotics has the deepest record: automotive and electronics assembly, warehousing and fulfilment, food production, and the service work an ageing population makes hard to staff. The engineering that produced Korea’s lead travels across borders more readily than most industrial advantages.

One caveat matters. Density counts the installed base, not where future orders land. Europe’s automation budgets will favour local content, standards and partners already on the ground over hardware shipped in and hoped for. The firms that turn this gap into revenue will build European operations and integration partnerships rather than export from Incheon — the structuring question a later part of this series takes up.

Europe has diagnosed itself, and the density data shows exactly where the problem sits: on the factory floor. The world’s deepest reservoir of applied robotics capability is in Korea; the largest unmet demand for it among wealthy economies is in Europe. The wider the gap, the larger the prize for the company that gets in first.

Sources:

• Robot Density Surges in Europe, Asia, and Americas — 2024 data (International Federation of Robotics, 8 April 2026): https://ifr.org/ifr-press-releases/news/robot-density-surges-in-europe-asia-and-americas

• World Robotics — Industrial Robots (International Federation of Robotics): https://ifr.org/wr-industrial-robots

• The Draghi report on EU competitiveness (European Commission): https://commission.europa.eu/topics/competitiveness/draghi-report_en

Disclaimer: This article is for informational purposes only and does not constitute investment or business advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Saint Clair Global builds companies to international institutional standards: the company side of the practice, where a business is made to read as a successful, attractive investment opportunity to the capital it seeks. Since 2016.

Saint Clair Pte. Ltd. · Singapore · saintclair.global

Saint Clair Global · Market Intelligence | August 2026

Europe First — Part 1 of 4

Europe has published an unusually frank account of its own competitiveness problem, and to a company weighing entry, the Draghi report reads as a list of unmet demand. Between 2008 and 2021, 147 companies later worth more than a billion dollars were founded in Europe; 40 moved their headquarters abroad, most to the United States. More than 60 per cent of EU companies now name regulation as an obstacle to investment. The continent has the capital, the customers and the public purpose. What it lacks is the firms that turn research into products at scale — and that shortfall is the strongest reason for Korean companies to look to Europe, because it is a gap Korean industry is built to fill.

The Weakness Europe Wrote Down

The report commissioned from Mario Draghi, the former European Central Bank president, set out to explain why the European economy has fallen behind the United States and China. Its findings are blunt. Only four of the world’s top fifty technology companies are European. In 2021, European firms spent €270 billion less on research and innovation than their United States counterparts. And for the past twenty years the largest corporate research investors in Europe have been dominated by automotive companies; in the United States, where autos and pharmaceuticals led in the early 2000s, the top three are now all in technology. Europe kept funding the industries of its past.

The talent record is the sharpest figure in the document. Of the 147 European unicorns founded between 2008 and 2021, 40 relocated their headquarters overseas, most of them to the United States, in search of the capital and scale their home market could not supply. Europe invents, then exports its winners. The report frames this as a scaling failure: the science is world-class, and the machinery that turns science into companies lags well behind it.

Regulation compounds the drag. The report records that more than 60 per cent of EU firms see regulation as a barrier to investment, and that 55 per cent of small and medium-sized enterprises rank regulatory and administrative burden as their single greatest challenge. Above that sits a fragmented rulebook: around 100 technology-focused laws and more than 270 regulators active in digital networks across the member states. Europe governs its digital economy thoroughly. It has made that economy hard to build in.

The clearest shortfall is in the digital and deep-technology sectors that now drive productivity growth. Where the United States has scaled cloud, chips and artificial intelligence, Europe has funded research programmes and watched the resulting companies leave.

The Response Is Already Funded

Implementation has been slow even where the diagnosis is accepted. The European Policy Innovation Council’s Draghi Observatory, which tracks the report’s 383 recommendations, put full implementation at 15.7 per cent in its July 2026 update, and 41.3 per cent counting partial progress. Energy and digitalisation are the least advanced.

Mario Draghi delivered the report in September 2024, at the European Commission’s own request, and few governments commission a review of their own weaknesses and then publish it in full. Since then, the Commission’s Competitiveness Compass has, by its own account, mobilised more than €1 trillion towards the problems Draghi identified. The largest lines include roughly €200 billion for artificial intelligence, of which €20 billion is set aside for a planned programme of AI gigafactories; €150 billion for defence through Security Action for Europe (SAFE); more than €100 billion for decarbonisation through the Clean Industrial Deal; and €70 billion for start-ups and scale-ups through TechEU.

The four figures are not of one kind. The artificial-intelligence and Clean Industrial Deal sums are mobilisation targets combining public and private money; SAFE is a lending ceiling for member states rather than a subsidy to industry; TechEU is the European Investment Bank’s financing envelope for 2025 to 2027. None of it is money already spent.

The money funds the effort to close the gap, on a timeline Europe controls. A continent committing more than €1 trillion to build capability it does not yet have will keep buying that capability from outside while it builds its own. For now, demand exceeds domestic supply.

Why This Is Korea’s Opportunity

Korea’s industrial character is the mirror image of Europe’s weakness. Where Europe struggles to scale research into products, Korea excels at exactly that conversion, moving from laboratory to manufactured output and then to global markets with unusual speed. Ugo Astuto, the European Union’s ambassador to Korea, put it plainly at the NextRise 2026 forum in Seoul in June: Europe is “at the cutting edge in several sectors of research”, while Korea is “particularly strong in transforming research into industrial output and scaling innovation”.

The proof already exists in the field. Since 2023, a Korean medical-AI company has worked as an independent reader inside a Swedish breast-screening centre, taking one of the two slots in a double-reading protocol that European systems could describe but not staff. The recall decision still rests with a radiologists’ consensus panel; what the system replaces is a read, not a judgement. The same logic extends across the sectors the Draghi report flags: applied artificial intelligence, semiconductors and advanced hardware, clean technology, and the industrial software European incumbents have been slow to build. In each, the European demand is documented and the domestic supply is thin. Korean firms sell into that space from a standing start.

Europe is a difficult market. The regulatory burden the report names falls on every entrant, and the fragmentation across member states defeats companies that treat the continent as a single customer. But compliance costs that slow the first entrant also slow the second. A Korean firm that clears the European standard holds that position longer than it would in a lighter-touch market.

For a founder choosing between regions, the United States offers scale and capital, and domestic competitors who already occupy every gap. Europe offers a market that has published, in detail, the places where it needs what it does not have. For the Korean companies able to fill the gap, Europe is the most open market they will find.

Sources:

• The Draghi report on EU competitiveness — European Commission: https://commission.europa.eu/topics/competitiveness/draghi-report_en

• The Future of European Competitiveness, Part A — European Commission (PDF): https://commission.europa.eu/document/download/97e481fd-2dc3-412d-be4c-f152a8232961_en

• The Future of European Competitiveness, Part B — European Commission (PDF): https://commission.europa.eu/document/download/ec1409c1-d4b4-4882-8bdd-3519f86bbb92_en

• The Draghi report: one year on — European Commission: https://commission.europa.eu/topics/competitiveness/draghi-report/one-year-after_en

• Draghi Observatory & Implementation Index, July 2026 update — European Policy Innovation Council: https://draghiwatch.eu/

• InvestAI (IP/25/467) — European Commission: https://ec.europa.eu/commission/presscorner/detail/en/ip_25_467

• SAFE (Security Action for Europe) — Council of the European Union: https://www.consilium.europa.eu/en/policies/safe/

• TechEU Platform — European Investment Bank: https://www.eib.org/en/press/all/2025-314-europe-s-innovative-companies-get-boost-as-eib-group-launches-techeu-platform-to-simplify-financing

• Europe and Korea seek scale-up partnership at NextRise 2026 — Aju Press: https://www.ajupress.com/view/20260618171516456

• Lunit and Capio S:t Göran Hospital collaborate to address radiologist shortage, June 2023: https://www.prnewswire.com/news-releases/lunit-and-capio-st-goran-hospital-collaborate-to-address-radiologist-shortage-with-ai-powered-mammography-analysis-301842062.html

Disclaimer: This article is for informational purposes only and does not constitute investment or business advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Saint Clair Global builds companies to international institutional standards: the company side of the practice, where a business is made to read as a successful, attractive investment opportunity to the capital it seeks. Since 2016.

Saint Clair Pte. Ltd. · Singapore · saintclair.global

Saint Clair Global · 마켓 인텔리전스 | 2026년 8월

유럽 우선(Europe First) 시리즈 — 1/4

유럽이 자국의 경쟁력 문제를 이토록 솔직하게 기록한 문서는 드물다. 드라기 보고서 이야기다. 유럽에 진출하려는 기업의 눈으로 보면, 이 보고서는 유럽이 아직 스스로 메우지 못한 수요의 목록이다. 2008년부터 2021년까지 유럽에서 기업가치 10억 달러를 넘긴 기업 147곳이 태어났지만, 그중 40곳이 본사를 해외로, 대부분 미국으로 옮겼다. EU 기업의 60% 이상은 규제를 투자의 걸림돌로 꼽는다. 자본도 고객도 정책적 의지도 유럽엔 있다. 없는 것은 연구 성과를 사업으로, 그것도 큰 규모로 키워 내는 기업이다. 바로 이 빈틈이 한국 기업에게는 가장 큰 기회다.

유럽이 스스로 쓴 진단서

이 보고서는 마리오 드라기 전 유럽중앙은행(ECB) 총재가 유럽 경제가 왜 미국과 중국에 뒤처졌는지 설명하기 위해 작성했다. 진단은 냉정하다. 세계 50대 기술기업 가운데 유럽 기업은 단 넷뿐이고, 유럽 기업이 연구·혁신에 쓰는 돈은 2021년 기준 미국 기업보다 2,700억 유로가량 적었다. 특히 상징적인 대목은 연구·개발(R&D)이다. 지난 20년간 유럽의 R&D 투자 상위권은 자동차 기업이 줄곧 주도해 왔다. 미국도 2000년대에는 자동차와 제약이 그 자리를 차지했지만, 이후 소프트웨어·하드웨어를 거쳐 지금은 상위 3곳이 모두 디지털 기업으로 바뀌었다. 결국 유럽은 과거의 산업에 계속 돈을 댄 셈이다.

가장 뼈아픈 것은 인재 통계다. 2008~2021년 유럽에서 태어난 유니콘 147곳 중 40곳이 본사를 해외로, 대부분 미국으로 옮겼다. 자국 시장이 대주지 못한 자본과 성장 무대를 찾아 떠난 것이다. 유럽은 발명은 하되 키운 승자는 수출한다. 보고서가 이를 연구의 실패가 아니라 스케일업의 실패로 규정한 이유다. 과학은 세계 최고 수준이지만, 그 과학을 기업으로 키우는 역량은 한참 뒤처져 있다.

규제도 부담을 키운다. EU 기업의 60% 이상이 규제를 투자의 장벽으로 여기고, 중소기업(SME)의 55%는 규제·행정 부담을 가장 큰 애로로 꼽는다. 규제 자체도 파편화돼 있어, 회원국 전역의 디지털 분야에 적용되는 기술 관련 법령만 100개에 이르고 규제 당국도 270곳을 넘는다. 유럽은 디지털 경제를 촘촘히 규율한 대신, 그 경제를 키우기 어려운 곳으로 만들었다.

빈틈이 가장 큰 곳은 생산성을 좌우하는 디지털·첨단기술 분야다. 미국이 클라우드와 반도체, AI를 키우는 사이, 유럽은 연구 프로그램에 돈을 댔고 그렇게 키운 기업이 떠나는 것을 지켜봤다.

유럽은 이미 1조 유로로 답하고 있다

물론 유럽이 문제를 인정한다고 실행까지 빠른 것은 아니다. 유럽정책혁신위원회(EPIC) 집계를 보면, 383개 권고 가운데 완전히 이행된 것은 15.7%에 그쳤고, 부분 이행까지 더해도 41.3%에 머문다. 특히 에너지와 디지털화 분야가 뒤처졌다.

그렇다고 유럽이 손을 놓고 있는 것은 아니다. 드라기 보고서는 2024년 9월, EU 집행위원회가 직접 요청해 나온 문서다. 한 정부가 자국의 약점을 스스로 진단하게 하고 그 결과를 그대로 공개하는 일은 흔치 않다. 보고서가 나온 뒤 집행위는 ‘경쟁력 나침반(Competitiveness Compass)’을 통해, 스스로 밝힌 바에 따르면 드라기가 지목한 과제에 1조 유로가 넘는 자금을 동원했다. 대표적인 항목만 봐도 인공지능(AI)에 약 2,000억 유로(이 가운데 200억 유로는 ‘AI 기가팩토리’), 방위산업에 SAFE(Security Action for Europe) 프로그램으로 1,500억 유로, 탈탄소화에 ‘클린 인더스트리얼 딜’로 1,000억 유로 이상, 스타트업·스케일업에 ‘테크EU(TechEU)’로 700억 유로다.

네 항목은 성격이 저마다 다르다. AI 2,000억 유로와 클린 인더스트리얼 딜 1,000억 유로 이상은 민관 자금을 합쳐 끌어내겠다는 ‘동원 목표’다. SAFE 1,500억 유로는 산업계에 주는 보조금이 아니라 회원국을 위한 대출 한도이고, 테크EU 700억 유로는 유럽투자은행(EIB)이 2025~27년에 집행하는 금융 지원 규모다.

다만 이 돈이 격차를 곧바로 메우지는 못한다. 격차를 메우려는 노력에 자금을 댈 뿐이고, 그 속도는 유럽이 정한다. 아직 갖추지 못한 역량을 1조 유로를 들여 키우는 동안에도, 유럽은 그 역량을 밖에서 계속 사들일 수밖에 없다. 적어도 지금은 수요가 자국 공급을 앞선다.

한국 스타트업에는 기회다

한국 산업의 성격은 유럽의 약점을 정확히 뒤집어 놓은 모양이다. 유럽이 어려워하는 바로 그 지점, 곧 연구를 사업으로 연결하는 일에서 한국은 속도가 빠르다. 실험실에서 양산으로, 다시 세계 시장으로 넘어가는 데 걸리는 시간이 유독 짧다. 우고 아스투토 주한 EU 대사도 지난 6월 서울에서 열린 넥스트라이즈(NextRise) 2026 포럼에서 같은 취지로 말했다. 유럽은 여러 연구 분야에서 앞서 있고, 한국은 그 연구를 산업 성과로 바꿔 규모를 키우는 데 특히 강하다는 것이다.

증거는 이미 나와 있다. 한 한국 의료 인공지능 기업은 2023년부터 스웨덴의 한 유방암 검진 센터에서 ‘독립 판독자’로 일하고 있다. 유럽 검진이 쓰는 이중 판독, 곧 두 명이 각 영상을 보는 방식에서 한 자리를 이 인공지능이 맡는다. 재검사를 부를지 여부의 최종 판단은 여전히 영상의학과 의사들의 합의 몫이다. 유럽이 진단은 해도 인력으로는 채우지 못한 의사 부족을 이렇게 현장에서 메우고 있는 셈이다. 이 사례는 이어지는 편에서 자세히 다룬다. 같은 공식은 드라기 보고서가 지목한 다른 분야에도 그대로 적용된다. 응용 AI, 반도체와 첨단 하드웨어, 청정기술, 그리고 유럽 기존 기업이 더디게 만들어 온 산업용 소프트웨어다. 어느 쪽이든 유럽의 수요는 뚜렷한데, 정작 유럽 내부의 공급은 부족하다.

물론 유럽이 쉬운 시장은 아니다. 규제는 외국 기업에도 똑같이 적용되고, 회원국마다 규정이 달라 유럽을 하나의 시장으로 뭉뚱그린 기업은 고전한다. 그러나 그 높은 문턱은 일단 넘고 나면 경쟁자를 막아 주는 진입 장벽이 된다. 기준이 까다로울수록, 그것을 통과한 기업이 그 시장에서 더 오래 살아남는다.

어느 지역에 진출할지 고민하는 한국 창업자에게, 미국은 규모와 자본을 주지만 모든 빈틈을 이미 차지한 현지 경쟁자도 함께 안겨 준다. 반면 유럽은 자신에게 무엇이 없는지를 스스로, 그것도 상세히 공개한 시장이다. 결국 유럽의 약점은 유럽만의 문제가 아니다. 그 빈틈을 채울 역량을 가진 한국 기업에게, 유럽은 지금 가장 열려 있는 시장이다.

출처:

• 유럽 경쟁력에 관한 드라기 보고서 (유럽집행위원회): https://commission.europa.eu/topics/competitiveness/draghi-report_en

• The Future of European Competitiveness — Part A (유럽집행위원회, PDF): https://commission.europa.eu/document/download/97e481fd-2dc3-412d-be4c-f152a8232961_en

• The Future of European Competitiveness — Part B (유럽집행위원회, PDF): https://commission.europa.eu/document/download/ec1409c1-d4b4-4882-8bdd-3519f86bbb92_en

• The Draghi report: one year on (유럽집행위원회): https://commission.europa.eu/topics/competitiveness/draghi-report/one-year-after_en

• Draghi Observatory & Implementation Index, 2026년 7월 갱신 (유럽정책혁신위원회, EPIC): https://draghiwatch.eu/

• InvestAI (IP/25/467, 유럽집행위원회): https://ec.europa.eu/commission/presscorner/detail/en/ip_25_467

• SAFE (Security Action for Europe), Council of the European Union: https://www.consilium.europa.eu/en/policies/safe/

• TechEU Platform (유럽투자은행, EIB): https://www.eib.org/en/press/all/2025-314-europe-s-innovative-companies-get-boost-as-eib-group-launches-techeu-platform-to-simplify-financing

• EU and Korea seek scale-up partnership at NextRise 2026 (AJU Press): https://www.ajupress.com/view/20260618171516456

• Lunit and Capio S:t Göran Hospital collaborate to address radiologist shortage, 2023년 6월: https://www.prnewswire.com/news-releases/lunit-and-capio-st-goran-hospital-collaborate-to-address-radiologist-shortage-with-ai-powered-mammography-analysis-301842062.html

Disclaimer: 본 글은 정보 제공만을 목적으로 한다. 투자 또는 사업 자문에 해당하지 않는다. 모든 의사결정은 독자적 조사와 자격을 갖춘 자문가의 상담을 기반으로 이뤄져야 한다.

About Saint Clair: 세인트 클레어는 유럽과 아시아를 잇는 국경 간 투자 회사다. 자본이 국경을 넘나드는 인프라까지 함께 구축하는 기관 투자자다. Saint Clair Global은 기업을 국제 기관 표준에 맞춰 구축한다. 사업이 구하는 자본 앞에서 성공적이고 매력적인 투자 기회로 읽히게 만드는, 그 실천의 기업 측면이다. 2016년 이래로.

Saint Clair Pte. Ltd. · 싱가포르 · saintclair.global

Bangladesh’s startup funding is reported at USD 124 million across twelve deals in 2025, roughly triple the year before. One transaction, a USD 110 million merger, accounts for eighty-nine per cent of it. Strip it out and the remaining eleven deals total approximately USD 14 million, against USD 42 million across forty-one deals in 2024: a market that lost two-thirds of its capital and three-quarters of its deal count while its headline tripled.

Ground Truth: Bangladesh | Saint Clair Market Intelligence | 10 August 2026

Based on published data and reporting from LightCastle Partners, Bangladesh Bank, The Business Standard, The Daily Star, Wamda and ExitStack. Saint Clair’s analysis follows.

Bangladeshi startups raised USD 124 million across twelve deals in 2025, roughly three times the USD 42 million raised across forty-one deals in 2024. Read quickly, the headline says recovery.

One Deal in Twelve

One of those twelve deals is worth USD 110 million: the April merger of ShopUp, a Bangladeshi commerce and financial-services company, with Sary, a Saudi business-to-business trading platform, to form SILQ Group. LightCastle Partners, the analytics house whose annual report is the market’s closest count of record, states on page five of its own document that excluding SILQ, 2025 funding was approximately USD 14 million. The report states it plainly. It is the headline that travels without it.

Financial services accounted for USD 110 million of the year’s total, eighty-nine per cent, from two deals rather than one. The top three deals made up approximately ninety-five per cent of capital deployed. Average ticket size across the year was USD 10 million; excluding SILQ, closer to USD 1 million. Late-stage financing and mergers and acquisitions (M&A) together made up ninety-two per cent of funding value, in a year with zero Series A rounds and zero debt rounds (Series A alone raised USD 21 million across six deals in 2024).

Removing the largest deal from a dataset invites suspicion, and rightly so; it can be made to prove almost anything. The defence has to be structural rather than statistical, and it is the argument the rest of this article makes.

What the Eleven Show

Below SILQ, the stage-by-stage figures are the strongest evidence in the report.

Pre-seed collapsed from USD 2 million across eight deals to USD 60,000 across one. Series A produced nothing, against USD 21 million across six rounds in 2024, and debt financing produced nothing, against USD 7 million the year before. Angel investors made no investments in 2025, down from roughly USD 1 million across five deals in 2024. Early-stage funding’s share of the total fell from thirty-three per cent to eight per cent, even as the average early-stage ticket rose from USD 451,000 to USD 948,000: fewer companies raising larger amounts at the stage where a market’s future gets planted. Late-stage average ticket size rose from USD 3 million to USD 57 million, a figure with essentially one input.

The market did not have a strong year carried by one great deal. It had a poor year that happened to contain one very large transaction.

Context matters here. 2021 remains the decade’s peak in Bangladesh, at USD 435 million across ninety-four deals, itself carried by the USD 250 million bKash–SoftBank round. LightCastle ranks SILQ as the second-largest transaction in the market’s history. A single deal defining Bangladesh’s annual funding number has happened before, which keeps this year’s reading structural rather than a complaint about 2025 specifically.

Rebuilding the Table

One boundary inside 2025’s foreign ninety-nine per cent cannot be fully audited from public sources: whether a Bangladeshi founder’s foreign holding company, raising from a fund run by the diaspora, counts as domestic capital or global. Bangladesh Bank’s Share Swap Circular of July 2025 grants general permission to establish a foreign holding entity for USD 10,000, expressly so founders can consolidate a split capitalisation table into one an international investor can underwrite. The regulator has made the flip cheap and legal, and it shifts what the domestic-versus-foreign line measures, from where a company started to where it is registered now.

LightCastle’s dataset is one house’s count of an admittedly opaque market, and its published report does not break the ex-SILQ total into a deal-by-deal table. The check available to a European investor is the named-transaction trackers and regional outlets that feed them — ExitStack, Future Startup, Wamda and similar. Eleven candidate 2025 transactions were tested against original source reporting, one by one; the table below carries the source each verdict was confirmed or excluded against.

Five rows rest on a single outlet rather than a corroborated pair: Markopolo.ai, iFarmer, Pluto, Wind and Truely are each confirmed against exactly one named source. One credible outlet is enough to rule a transaction out on date or domicile; it is not enough, alone, to carry the article’s one inclusion.

One candidate survives cleanly, and on two independent legs. Cassetex, Bangladesh’s first battery-swapping company for electric three-wheelers, raised USD 1.6 million from Infusion Partners Ltd. and Waverider International Consulting Co. Ltd.; ExitStack reported the round directly, and The Business Standard named the same investors and the same figure in separate coverage weeks later. The two sources place the announcement in January and February 2025 respectively; both sit inside the calendar year. The other nine fail on one of two separate grounds: five closed outside the 2025 calendar year, by margins ranging from a few months to nearly three years, and three are registered in the United Arab Emirates or Singapore regardless of when they closed. Fasset fails on both counts simultaneously.

At USD 51 million, Fasset would have overwhelmed LightCastle’s USD 14 million ex-SILQ figure had it belonged in the count, and either finding would have mattered: a material omission in the primary dataset, or confirmation that a widely circulated “Bangladesh deal” measures a founder’s birthplace rather than a company’s domicile. It resolves toward the second, by a wide margin. Fasset’s Series B closed in May 2026, headquartered in Los Angeles and Dubai. Its only connection to Bangladesh is its founder, Mohammad Raafi Hossain, and the local press has kept a distinction the trackers dropped: the Daily Star‘s own headline describes a “Bangladeshi-born founder’s startup,” a founder’s origin rather than a company’s domicile.

Markopolo.ai resists a clean verdict either way. It raised USD 2 million on 11 May 2025, a seed round supported by Saudi Arabia’s National Technology Development Programme that arrived bundled with the relocation of its headquarters to the Kingdom. Startup Bangladesh Limited, the state fund, had backed the company previously. Markopolo.ai is a Bangladeshi company in the act of becoming something else, in the same year and through the same route the Share Swap Circular was written to make easy. Part 1’s domicile question now has a face: one named transaction, mid-flip.

Against LightCastle’s approximately USD 14 million across eleven deals, independent verification confirms USD 1.6 million and flags one transaction whose domicile is genuinely in motion. LightCastle’s figure survives the test undisturbed: nothing here contradicts it, and a thin market generates small transactions that never reach an English-language tracker at all. What the test shows instead is that the data built to check the aggregate needs the same scrutiny the aggregate received. A European investor reaching for a named-deal list to sense-check a headline is reaching for a list with the same reliability problem, one level down.

Why Financial Services, Why That Size

A USD 110 million financial-services transaction in a market this size calls for an explanation, and it sits in the country’s payments infrastructure rather than in venture appetite. Mobile financial services (MFS) accounts in Bangladesh reached approximately 239.3 million in January 2025, up 9.21 per cent on the year before. Accounts are not unique users, and multiple-account holding is common; Bangladesh Bank states that caveat itself. Monthly transaction volume reached approximately BDT 1.72 trillion, up 32.56 per cent year on year, close to USD 439 million moving through the system daily at the rate prevailing in January 2025. bKash, the market leader, processed more than eleven million transactions a day in 2024 across over sixty-eight million users; Nagad, second-placed, processed roughly seven million.

Capital of SILQ’s size arrived where the rails already existed. That connection is the year’s most useful reframing: the concentration in financial services is legible once the payments infrastructure beneath it is visible.

SILQ is structurally distinct from the venture activity a funding aggregate is built to capture. It is a merger between a Bangladeshi company and a Saudi one, co-led by Sanabil Investments — wholly owned by Saudi Arabia’s Public Investment Fund (PIF), roughly USD 925 billion in assets, in what was reported as Sanabil’s first major South Asia commitment — alongside Valar Ventures, with Qatar Development Bank, Flourish Ventures and Kuwait’s Wafra further down the syndicate. The structure combines equity with a financing facility for SILQ Financial, the group’s financial-services arm. Afeef Zaman, ShopUp’s founder, became SILQ Group’s chief executive; Mohammed Aldossary, Sary’s founder, leads SILQ Financial. Both brands are retained in their home markets, and an initial public offering (IPO) is mooted for around 2027. The strategic frame, as reported, is a Gulf–South Asia trade corridor projected above USD 682 billion over the coming decade.

The clearest evidence of what kind of business SILQ actually is arrived after the merger closed. On 22 July 2026, SILQ secured a USD 75 million Shariah-compliant debt facility from Fasanara Capital, the London-based investment firm, to fund working-capital loans for small businesses in Saudi Arabia through Fina, its embedded-finance product. The facility is projected to provide close to SAR 3 billion, nearly USD 800 million, in liquidity to around two thousand local businesses over the funding year. A company that carried eighty-nine per cent of a venture headline has since raised debt against receivables, the ordinary financing behaviour of a trade-finance operator scaling its balance sheet. Venture-backed startups rarely follow that path in their first year.

The Corridor View

A European investor reading a frontier funding headline should ask for three things before trusting it: the deal-level table behind the aggregate, the distribution across stages rather than the total alone, and the median ticket size rather than the mean, which one large transaction can distort past recognition.

Bangladesh in 2025 fails that first test twice over. LightCastle’s own report does not publish a deal-by-deal table beyond its largest few transactions, and the trackers built to fill the gap carry the wrong year or the wrong domicile for nine of the eleven candidates this article tested. An aggregate figure with no deal-level table behind it is asking the reader to trust a number it has not shown its working for. Verification, in that setting, is the entire exercise.

Taka figures are converted at USD/BDT 123.45, the rate on 29 July 2026, except where a contemporaneous rate is stated.

Sources:

• LightCastle Partners, with Startup Bangladesh Limited, Anchorless Bangladesh and ExitStack, Bangladesh Startup Investments Report 2025: Year in Review (January 2026). lightcastlepartners.com

• LightCastle Partners, Bangladesh Startup Investments Report 2025-H1 (July 2025). lightcastlepartners.com

• The Business Standard, ShopUp, Sary merge to form SILQ with $110m Saudi, US investment. tbsnews.net

• The Daily Star, ShopUp, Sary to merge, raise $110m. thedailystar.net

• Latham & Watkins, Latham Advises SILQ on US$75 Million Financing Facility (July 2026). lw.com

• FinTech Futures, SILQ lands $75m Fasanara debt line for Saudi SME lender Fina (July 2026). fintechfutures.com

• The Financial Express, Bangladesh MFS accounts surge by 20 million in a year, transactions up 32pc. thefinancialexpress.com.bd

• ExitStack, Cassetex Secures US$1.6Mn to Expand Battery-as-a-Service in Bangladesh and Beyond. exitstack.co

• The Business Standard, Bangladeshi entrepreneur joins Earth Action Report programme (March 2025). tbsnews.net

• The Daily Star, Markopolo.ai raises USD 2 million to scale in Saudi Arabia. thedailystar.net

• The Daily Star, Bangladeshi-born founder’s startup raises $51 million in series B. thedailystar.net

• FinTech Futures, Stablecoin neobank Fasset raises $51m Series B (May 2026). fintechfutures.com

• The Business Standard, Pathao secures $12m from foreign investors to boost fintech initiatives (September 2024). tbsnews.net

• Wamda, Pluto closes $4.1 million pre-Series A to drive KSA expansion (January 2025). wamda.com

• TechCrunch, Wind.app makes DeFi accessible to the average consumer (November 2023). techcrunch.com

• The Daily Star, Bangladesh-based startup Tiger New Energy secures USD 3.5mln funding (June 2024). thedailystar.net

• DealStreetAsia, Singapore traveltech startup Truely secures additional $2m (June 2025). dealstreetasia.com

• Future Startup, iFarmer Secures $1.5 Million from Symbiotics as International Capital Continues to Back Bangladesh’s Agritech Infrastructure (April 2026). futurestartup.com

• The Business Standard, Jobtech firm Shomvob raises $1 million in pre-seed funding (May 2024). tbsnews.net

• The Daily Star, Bangladeshi startup MedEasy raises $750K in seed funding (September 2023). thedailystar.net

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Saint Clair Asia (saintclair.asia) builds international investors’ access to frontier innovation ecosystems that institutional channels do not reach.

Learn more: saintclair.sg | saintclair.asia | Contact: contact@saintclair.sg

On business cards across Korea, from large companies and government offices alike, the email is often a personal one, and the handle in front of the @ is often not a name at all. To a foreign eye it can look careless. The reasons are more ordinary than that, and the standard it misses is almost trivially simple: your own name.

Saint Clair · Market Intelligence | August 2026

A living ritual

In Europe the paper card has all but disappeared. A LinkedIn connection or a WhatsApp number does the work now, and even that comes only once a conversation has earned it. In Korea the exchange is still a piece of formal etiquette: offered with both hands, received with both, read once on the spot. The card stands closer to a formal document, and it sets a first impression. So the few characters it carries should be deliberate. What do they tell the other person, and how do you wish to be seen?

Since 2016 Saint Clair has gathered thousands of cards across the Europe–Korea corridor. One pattern recurs. Even at companies whose names everyone knows, the address is frequently a personal one, a Gmail or a Naver rather than the company domain. It appears across ages and industries, well beyond startups and young founders. It reaches into government as well: Korea Pro, an English-language news and analysis service on South Korea, reported in 2024 that officials’ cards commonly show a personal Gmail, and that where they do not, staff often write one on by hand, a practice widespread enough to raise security concerns. And the handle in front of the @ is often not the person’s name.

Why the personal address

The reasons are usually plain, and they come down to convenience. Naver is the portal and mail service most Koreans live inside, tied to the phone and to the rhythm of the day, while official and corporate systems are cumbersome. That friction is exactly what Korea Pro described in government offices, where staff default to accounts that simply work. A personal address carries one further, practical advantage: it survives a change of employer, and Korean careers move often. A company address expires the day its holder leaves. Seen this way, the personal address is habit and infrastructure, and to read it as a lapse of seriousness is to misread the person.

And often, no name at all

The company domain is one thing; the handle is another, and what sits in its place is seldom a name. In its slot you find a self-chosen Daniel or Jenny that appears on no passport or badge; a number, most often a birth year or a birthday, or 1004, which in Korean reads as 천사, “angel”. Then come taste and self: a persona, as in coolguy or alwayssmile; a passion, as in coffeelover, tennismania, or Nadal, after the tennis star; Konglish, as in best-ing (”doing one’s best”); a bare phrase such as thiscountry; a Korean word typed with the keyboard still in English, which surfaces like a difficult password; even a personal mobile number. Chosen by a hand that does not work in English, some read oddly, and a few carry a meaning their owner never intended.

To a foreign counterpart this is especially strange. The expectation is that Koreans rarely show emotion or taste in the first place; yet on the card, the professional document itself, the personal spills out all at once. It reads as too personal, and not quite serious. That reaction deserves to be named plainly rather than explained away.

Perhaps the deeper reason is that Koreans seldom use one another’s names at all. At work people address one another by title rather than by name; a manager is “bujang-nim”, rarely “Jiwon”. Leading with one’s own name is, in that light, an unfamiliar move. The international standard runs the other way, and it is almost embarrassingly simple: you use the name you actually go by. Measured against something that plain, a handle with no name in it stands out on its own. (The examples here are illustrative, not drawn from any real card.)

No one was taught otherwise

Underneath sits a gap no one addresses. In Korea, nobody is taught to hold a professional identity, an email or a handle, to an international standard. Part of it is that English is still not quite second nature: on the 2025 EF English Proficiency Index, Korea ranks 48th, inside the “moderate” band. An unfamiliar language is easy to treat as ornament, and ornament is where the simple standard slips. The instant a Konglish string, a password-like handle, or an English name used nowhere in real life goes onto a card, form and substance part company.

What is at stake, then, is the international standard, which rewards a match between form and substance. The pull to look modern or Western is easy to read into the practice, yet the standard asks for something plainer. A single address moves the needle: Livermore, Scafe and Asher (2016) found that an email built from a real name on an institutional domain read as the most credible, while nicknames and generic accounts pulled credibility sharply down.

Reading what is in front of you

Capital Diplomacy begins by reading the other side on their own terms rather than through an imported idea of what “professional” should look like. For the person receiving the card, that means the discipline to resist marking the holder down, in seriousness or in seniority, for the want of a company domain; to doubt someone’s standing over a personal Gmail is to misread a practice that is ordinary here. A hobby in the handle is, if anything, a low-cost opening: a word about it costs nothing and is usually welcome.

For the person handing the card over, the task runs the other way. Credibility across a border holds when the identity on the card and the identity actually lived coincide, and using one’s own name covers most of that distance. The reading, in the end, runs both ways: one side reads a person in those few characters, the other sets a true one down there. Trust across a border is built where those two kinds of precision meet.

Sources:

• Korea Pro (2024), “Korean government’s widespread use of personal emails poses security risks” — personal Gmail on official business cards; cumbersome official software; exposure to phishing. https://koreapro.org/2024/03/korean-governments-widespread-use-of-personal-emails-poses-security-risks/

• EF English Proficiency Index 2025 — South Korea 48th, score 522, “moderate” band. https://www.ef.com/wwen/epi/regions/asia/south-korea/

• Livermore, Scafe & Asher (2016), “The Impact of E-mail Address on Credibility,” Journal of Academic Administration in Higher Education — a real name on an institutional domain reads as most credible; nicknames and generic accounts reduce it. https://files.eric.ed.gov/fulltext/EJ1139123.pdf

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Since 2016.

Learn more: saintclair.sg | Contact: contact@saintclair.sg

한국에서 받는 명함의 이메일은 대기업이든 관공서든 개인 주소인 경우가 많다. 골뱅이 앞 아이디는 정작 이름이 아닐 때가 잦다. 외국인의 눈에는 허술해 보이지만, 이유는 생각보다 평범하고, 그 관행이 놓치는 국제 표준은 뜻밖에 간단하다. 실제로 쓰는 자기 이름이다.

세인트 클레어 · 마켓 인텔리전스 | 2026년 8월

아직 살아 있는 의례

유럽에서 종이 명함은 거의 사라졌다. 링크드인으로 대신하거나 왓츠앱 연락처를 주고받고, 그것도 충분히 대화를 나눈 뒤의 일이다. 반면 한국에서 명함 교환은 여전히 살아 있는 격식이라, 두 손으로 건네고 두 손으로 받아 그 자리에서 한 번 읽는다. 그만큼 명함은 첫인상을 좌우하는 공식 문서에 가깝고, 거기 담기는 몇 글자는 신중해야 한다. 나는 상대에게 무엇을 전하고, 어떻게 보이고 싶은가.

세인트클레어는 2016년 이래 유럽–한국 회랑에서 수천 장의 명함을 모았다. 그 안에서 되풀이되는 패턴이 하나 있다. 이름만 대면 알 만한 대기업의 직원이라도 회사 도메인 대신 지메일이나 네이버 같은 개인 주소를 적는 경우가 흔하다는 것이다. 스타트업이나 젊은 창업가에 국한되지 않고, 연령과 업종을 가리지 않는다. 관공서도 예외가 아니다. 코리아프로는 2024년, 공무원 명함에 개인 지메일이 적혀 있고 없으면 손으로 적어 주기까지 하는 일이 보안 우려를 부를 만큼 흔하다고 전했다. 게다가 골뱅이 앞 아이디는 정작 사람 이름이 아닐 때가 많다.

왜 개인 주소인가

이유는 대개 단순하다. 편의의 문제다. 네이버는 대다수 한국인이 일상과 휴대폰 속에서 늘 쓰는 포털이자 메일인데, 관공서와 회사의 공식 메일은 번거롭다. 코리아프로가 정부 부처에서 짚은 마찰도 바로 이것으로, 공식 시스템이 불편하니 잘 되는 계정으로 손이 간다. 게다가 개인 주소에는 회사를 옮겨도 그대로 남는다는 실질적 이점이 있다. 한국의 직장 이력은 자주 움직이고, 회사가 준 주소는 떠나는 날 사라지기 때문이다. 이렇게 보면 개인 주소는 습관과 인프라의 문제이며, 이를 진지함의 결여로 읽으면 사람을 잘못 보는 것이다.

그런데 정작 이름이 없다

회사 도메인은 그렇다 치더라도, 아이디는 또 다른 문제다. 그 자리를 차지하는 것은 좀처럼 이름이 아니어서, 여권에도 사원증에도 없는 대니얼이나 제니가 오르고, 숫자라면 대개 태어난 해나 생일이며, 천사로 읽히는 1004 같은 숫자놀이도 흔하다. 좋아하는 운동이나 음식, 꽃, 오래 응원해 온 가수가 그대로 아이디가 되기도 한다. hi나 love처럼 단어 하나가, 혹은 ‘I ain’t alone’을 붙여 쓴 문장이 앉기도 하고, 한글을 영문 자판 그대로 쳐 비밀번호처럼 보이는 문자열이나 자기 휴대폰 번호가 오르기도 한다. 영어가 모국어가 아닌 손에서 고른 만큼, 뜻이 엉뚱하게 읽히거나 본인도 의도하지 않은 뜻을 담기도 한다.

외국인의 눈에는 이것이 낯설다. 명함은 공식 문서인데, 거기에 사적인 취향이 얹히면 이상하고 때로는 프로페셔널하지 않게 비친다. 그래서 이 반응은 에둘러 넘기기보다 그대로 인정하는 편이 낫다.

어쩌면 이름이 빠지는 더 깊은 이유는, 우리가 서로 이름을 잘 부르지 않는다는 데 있는지도 모른다. 직장에서는 부장님, 대리님, 프로님이 이름을 대신하니, 자기 이름을 앞세우는 일 자체가 익숙지 않다. 그런데 국제 무대의 기준은 오히려 그 반대여서, 실제로 쓰는 자기 이름을 그대로 쓰면 그만이다. 그 간단한 기준에서 보면, 이름이 빠진 아이디는 그 자체로 눈에 띈다. (아이디 예시는 모두 설명을 위한 것이며, 특정인의 명함에서 옮긴 것이 아니다.)

아무도 가르쳐 주지 않았다

그 아래에는 아무도 짚지 않는 공백이 있다. 한국에서는 이메일이든 아이디든, 직업적 정체성을 어떻게 국제 기준에 맞추는지 가르치지 않는다. 영어가 아직 완전히 몸에 배지 않은 탓도 있다. EF 영어능력지수 2025년판에서 한국은 세계 48위, ‘보통’ 등급이다. 익숙하지 않은 언어는 장식처럼 다루기 쉽고, 장식으로 다루는 순간 그 간단한 기준을 놓친다. 콩글리시나 비밀번호 같은 문자열, 실생활에서 쓰지도 않는 영어 이름이 명함에 오르는 순간, 형식과 실질은 어긋난다.

그러니 걸려 있는 것은 국제 기준이다. 멋이나 서구화의 문제로 읽기 쉽지만, 무대가 요구하는 것은 형식과 실질의 일치다. 이메일 주소 하나가 인상을 실제로 바꾼다. 리버모어(Livermore)와 동료들이 2016년 미국 대학 구성원을 상대로 한 조사에서, 정식 이름과 소속 도메인을 쓴 주소가 가장 높은 신뢰를 얻었고, 별명이나 상용 메일 계정은 신뢰를 크게 떨어뜨렸다.

앞에 놓인 것을 읽는다는 것

자본 외교(Capital Diplomacy)는 프로페셔널의 수입된 정의를 들이대기보다, 상대를 그의 방식대로 읽는 데서 시작한다. 명함을 받아 드는 쪽에 필요한 것은, 회사 도메인이 없다는 이유로 상대의 무게나 직급을 낮잡지 않는 눈이다. 개인 지메일 하나로 진지함을 의심한다면, 한국에서 지극히 흔한 관행을 오독하는 셈이다. 취미가 담긴 아이디라면 오히려 관계로 들어가는 낮은 문턱이 되어, 한마디 건네는 데 드는 비용은 거의 없고 대체로 반갑게 받아들여진다.

명함을 건네는 쪽의 몫은 반대편에 있다. 국경 밖에서 신뢰는 명함 위의 정체성과 실제로 살아가는 정체성이 맞아떨어질 때 생기고, 자기 이름을 그대로 쓰는 것만으로 그 거리는 대개 좁혀진다. 결국 그 읽기는 양쪽 모두의 일이다. 받는 이는 몇 글자에서 상대를 읽고, 건네는 이는 그 몇 글자에 자기를 정확히 담는다. 국제 무대의 신뢰는 그 두 정확함이 만나는 지점에서 만들어진다.

출처 및 근거

• 코리아프로(Korea Pro), 2024, “Korean government’s widespread use of personal emails poses security risks” — 관공서 명함의 개인 지메일, 번거로운 공식 소프트웨어, 피싱 노출. https://koreapro.org/2024/03/korean-governments-widespread-use-of-personal-emails-poses-security-risks/

• EF 영어능력지수(EF EPI) 2025 — 한국 세계 48위, 522점, ‘보통’ 등급. https://www.ef.com/wwen/epi/regions/asia/south-korea/

• Livermore, Scafe & Asher (2016), “The Impact of E-mail Address on Credibility,” 『Journal of Academic Administration in Higher Education』 — 정식 이름+소속 도메인 주소가 신뢰도 최상, 별명·상용 메일 계정은 신뢰도 저하. https://files.eric.ed.gov/fulltext/EJ1139123.pdf

고지: 이 글은 정보 제공만을 목적으로 하며 투자 자문에 해당하지 않는다. 모든 결정은 독자적인 조사와 자격을 갖춘 자문가와의 상담을 바탕으로 내려야 한다.

세인트 클레어 소개: 세인트 클레어(Saint Clair)는 유럽과 아시아를 잇는 국경 간 투자사다. 2016년 이래 자본이 오가는 인프라까지 직접 구축해 왔다.

더 알아보기: saintclair.sg | 문의: contact@saintclair.sg

Bangladeshi startups raised USD 124 million in 2025, roughly three times the previous year. Investors inside the country provided USD 684,000 of it, across three deals, in a market of 170 million people; the country’s angel investors and its financial institutions provided nothing. In May, a company owned by thirty-nine Bangladeshi banks set out to build the domestic layer that is missing. It has yet to make an investment, or to appoint a chief investment officer.

Ground Truth: Bangladesh | Saint Clair Market Intelligence | 3 August 2026

Based on published data and reporting from LightCastle Partners, Bangladesh Bank, Startup Bangladesh Limited, the World Bank Group and the International Finance Corporation, the Asian Development Bank, The Business Standard, The Daily Star and Prothom Alo. Saint Clair’s analysis follows.

Bangladeshi startups raised USD 124 million across twelve deals in 2025, roughly three times the 2024 total. The headline has been read as a recovery.

The One Per Cent, Disaggregated

Ninety-nine per cent of that capital, USD 123 million across nine deals, came from investors outside Bangladesh. The domestic contribution was USD 684,000 across three deals. Bangladeshi angel investors made no investments at all in 2025, against roughly USD 1 million across five deals the year before, and the country’s financial institutions and development finance institutions made none either.

A soft year would pull domestic and foreign capital down together. Bangladesh saw foreign capital arrive in volume while domestic capital thinned out beneath it. The headline tripled; the local deal count fell from eight to three, and the local total fell from roughly USD 1 million to under seven hundred thousand.

The state is the closest thing the market has to a domestic institutional investor. Startup Bangladesh Limited, the flagship venture fund of the Information and Communication Technology (ICT) Division and the only government-sponsored venture capital fund in the country, launched in March 2020 with BDT 500 crore of allocated capital, described at the time as roughly USD 65 million. At today’s rate the same nominal sum is USD 40.5 million. By June 2026 the fund had deployed Tk 109 crore of it, or USD 8.8 million, across thirty-six approved startups. Over six years. Foreign investors put USD 123 million into Bangladeshi startups in 2025 alone.

Those thirty-six came through a due-diligence process that rejected nineteen others. The fund has written follow-on cheques into 10 Minute School and Pathao, and can point to more than seven thousand direct jobs and over Tk 180 crore, some USD 1.5 million, in tax paid by its portfolio.

India is the comparator that matters. Bangladesh’s gross domestic product (GDP) per capita is USD 2,734, within a hundred dollars of India’s USD 2,818. Startup investment per capita is USD 0.7 against India’s USD 9, thirteen times the Bangladeshi figure, and 0.03 per cent of GDP against 0.3 per cent.

Why Domestic Capital Does Not Arrive

Bangladesh has capital, and it moves at scale. Registered mobile financial services accounts stood at roughly 239 million in January 2025, on monthly transaction volumes near BDT 1.72 trillion, about USD 14 billion. What the country lacks are institutions that carry private money into companies needing equity.

The International Finance Corporation puts the financing gap for micro, small and medium enterprises (MSME) at approximately USD 2.8 billion. Formal small and medium enterprises account for around 40 per cent of the workforce and a quarter of GDP, and sit largely outside affordable formal finance. The causes are structural: informality, thin credit histories, scarce collateral, and a documentation burden that prices small borrowers out. The World Bank’s Country Private Sector Diagnostic of April 2025 records that financial inclusion has stalled after a decade of progress, most sharply for women.

Where an institution has built the route, the money arrives. City Bank’s nano-loans distributed through bKash had disbursed around USD 190 million to some 950,000 users by April 2025, on transaction data converted into credit at scale. For growth equity, no institution has built one, and the absence in venture is the absence the MSME gap measures, one asset class further out.

Whose Money Actually Came

“Ninety-nine per cent foreign” tells a European investor very little about risk. Which foreigners, and on what horizon, tells them a great deal.

Four distinct kinds of capital sit inside that figure. Sanabil Investments, wholly owned by Saudi Arabia’s Public Investment Fund (PIF), made what was reported as its first major South Asia investment. Valar Ventures is a New York growth fund. Qatar Development Bank and Kuwait’s Wafra are sovereign-adjacent Gulf institutions. The Asian Development Bank (ADB), through ADB Ventures, is a multilateral writing genuinely early-stage cheques into cleantech and agriculture, and one of very few institutional investors doing so on the ground. A sovereign wealth fund’s venture arm, a growth fund, two development-linked Gulf balance sheets and a multilateral hold four different views of when they need their money back.

All but one of those names reached Bangladesh through a single syndicate: the USD 110 million financing behind the merger of ShopUp and Sary into SILQ Group, announced in April 2025. LightCastle’s report describes roughly one-third of 2025’s global funding as coming from Gulf-based investors, and gives no absolute figure and no named investors behind that share. The named syndicate is verifiable. The one-third aggregate is unverified, and appears here as the report’s own framing.

The top three deals accounted for around 95 per cent of the capital. The average ticket was USD 10 million, and about USD 1 million once the single merger is set aside. There were no Series A rounds at all in 2025, against USD 21 million across six in 2024, and no debt rounds. Nine deals from abroad and three at home priced an entire market between them, and the largest of those transactions was a cross-border strategic merger that says little about the health of Bangladeshi companies raising their second round.

Thirty-Nine Banks

The development that could change the picture arrived in May. Bangladesh Startup Investment Company PLC (BSIC), owned by thirty-nine Bangladeshi commercial banks, launched on 12 May 2026 in Dhaka with an inaugural fund, Onkur Bangladesh Fund 1, of Tk 425 crore. BSIC states the fund size as USD 35 million; at today’s rate the paid-up figure converts to USD 34.4 million. Authorised capital is Tk 2,000 crore, some USD 162 million, and annual inflows are projected at around Tk 200 crore, or USD 16 million, because the banks’ contributions repeat. Onkur targets late-seed and Series A, the stage at which Bangladesh recorded nothing in 2025.

The funding mechanism is what makes it structural. Since 2021 every scheduled bank in Bangladesh has been required to set aside 1 per cent of annual net profit for startup financing. Bangladesh Bank’s Startup Finance Master Circular of July 2025 reclassified that set-aside as equity, barred its use for lending, and directed the whole pool into a central venture capital company. BSIC’s founding corpus is those thirty-nine banks’ obligations for the years 2020 to 2024, and the central bank authorised the allocation and shaped the prudential framework around it.

Here the public record stops short of what a reader would want. The circular described a venture capital company that Bangladesh Bank would establish and manage. BSIC is a public limited company owned by the banks, governed by its own board and investment committee, with the central bank in a supervisory role. No public source states that BSIC is the vehicle the circular named, and none explains the difference in structure. The relationship is evident in the funding line and unspecified everywhere else.

The figures are no clearer. In April 2025 the Governor of Bangladesh Bank announced a BDT 800 crore equity fund for startups backed by commercial banks, some USD 65 million, alongside a BDT 400 crore debt fund. Thirteen months later the bank-backed equity vehicle that launched was capitalised at Tk 425 crore, roughly half, and nothing in the public record connects the two figures.

One nominal sum makes the point on its own. BDT 500 crore, USD 40.5 million, currently attaches to four separate instruments: Bangladesh Bank’s refinancing facility for startup lending, at 0.5 per cent to the banks and a 4 per cent cap to the borrower; Startup Bangladesh Limited’s allocated capital; BSIC’s working-capital facility for its portfolio companies; and a new startup fund proposed through the ICT Division in the FY2026-27 budget. Three of the four are live at the same time, and no public document sets them out together. A reader meeting any one of them alone would reasonably take it for the others.

And Onkur has yet to invest. BSIC has said it will appoint a managing director and a chief investment officer during the third quarter of 2026, complete its first three investments before the year ends, and hold eight to twelve portfolio companies by 2028. As of the beginning of August, none of that has happened. The gap between an announcement and a deployment is well established here: Startup Bangladesh Limited holds BDT 500 crore of allocated capital and has deployed Tk 109 crore of it since 2020.

The Corridor View

A market with no domestic institutional bid has no price discovery. The foreign investor sets the price alone, with no local counterparty in a position to disagree, and no local institution positioned to lead the next round if the first one goes well.

For the European investor with patience and governance discipline, that cuts in two directions at once. Entry valuations face no competitive domestic pressure, and competition for good companies is close to absent. The same absence means no second opinion on what a company is worth, no natural follow-on capital, and an exit route that the Dhaka Stock Exchange has yet to build for technology companies at venture-stage valuations.

If Onkur deploys, thirty-nine banks produce that second opinion. A domestic institution writing late-seed and Series A cheques creates a local price for Bangladeshi risk, and a counterparty for the round after the one a foreign investor leads. The banks have found the capital and built the vehicle. Whether they can invest it is the question the next five years turn on. The first three cheques will answer it.

Taka figures are converted at USD/BDT 123.45, the rate on 29 July 2026, except where a contemporaneous rate is stated.

Sources:

• LightCastle Partners, with Startup Bangladesh Limited, Anchorless Bangladesh and ExitStack, Bangladesh Startup Investments Report 2025: Year in Review (January 2026). lightcastlepartners.com

• LightCastle Partners, Bangladesh Bank Startup Directives 2025 Explained (24 August 2025). lightcastlepartners.com

• LightCastle Partners, Startup Ecosystem: Bangladesh’s FY2026-27 Budget (6 July 2026). lightcastlepartners.com

• Bangladesh Startup Investment Company PLC, corporate disclosures. bsic.vc

• Prothom Alo, Bangladesh’s banking sector launches institutionally governed venture capital platform (12 May 2026). en.prothomalo.com

• The Daily Star, 39 banks launch Tk 425cr venture capital platform for startups (12 May 2026). thedailystar.net

• The Daily Star, Bangladesh Bank announces Tk 800 crore equity fund for startups (7 April 2025). thedailystar.net

• The Daily Star, Startup Bangladesh fund set for expansion to Tk 1,000 crore (9 June 2026). thedailystar.net

• The Financial Express, Tk 1.09b invested in startups: ICT Minister (9 June 2026). thefinancialexpress.com.bd

• The Business Standard, ShopUp, Sary merge to form SILQ with $110m Saudi, US investment. tbsnews.net

• The Business Standard, 39 banks launch $35m venture capital to boost Bangladesh startups. tbsnews.net

• Startup Bangladesh Limited, Fund of Funds overview. startupbangladesh.vc

• International Finance Corporation and World Bank Group, Country Private Sector Diagnostic: Bangladesh (April 2025). ifc.org

• World Bank, Bangladesh: Inclusive Digital Financial Services. digitalfinance.worldbank.org

• Asian Development Bank, Asia Small and Medium-Sized Enterprise Monitor 2025. adb.org

• Trading Economics, USD/BDT reference rate. tradingeconomics.com

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Saint Clair Asia (saintclair.asia) positions international investors within frontier innovation ecosystems that institutional channels do not reach.

Learn more: saintclair.sg | saintclair.asia | Contact: contact@saintclair.sg

Saint Clair Capital · Ground Truth | July 2026

Executive Summary

Foreign capital reached Market Kurly’s Series C and found governance as the binding constraint. Harmonising consent rights across roughly eighteen institutional investors — the company’s existing domestic shareholder base — took four months, the same figure Article 2 measured at Coupang’s contract layer, recurring here one level down. The governance a foreign investor requires is specific: an independent board and a functioning audit committee, disciplined documentation of resolutions and intellectual property (IP) assignments, standardised reporting on a fixed cadence, and clear ownership of the IP the valuation rests on. Few Korean companies meet the full set before a foreign round arrives. The domestic system never tested for it, and nothing before that point required a founder to build toward a standard nobody around them enforced. Park Hee-duk’s Kurly case establishes the sharper claim: governance readiness gates the transaction ahead of product-market fit, and the rebuild it demands runs six to eighteen months that most founders do not have when the foreign investor is already at the table. The implication generalises. Every cross-border transaction in this series has needed a translation between the Korean domestic standard and the global one. Where that translation exists in advance, capital moves. Where it does not, the company builds it under deadline pressure, at cost.

The Kurly Delay

Foreign capital reached Market Kurly’s Series C already knowing what it wanted: dawn delivery at national scale, a merchandiser-selection engine investors elsewhere would recognise as defensible, and a company growing into infrastructure Korean e-commerce lacked. What it found, once due diligence moved from the product to the paperwork, was a governance structure built for a different kind of company.

Some eighteen institutional investors — its existing domestic shareholders — held positions on Kurly’s cap table by the time foreign capital arrived at the round. Each carried consent rights negotiated under the Korean shareholder-agreement convention: individually held, individually exercised, individually renegotiated whenever a new investor sought to enter. Harmonising those eighteen positions into a structure the incoming capital could actually underwrite took four months.

Article 2 of this series measured a near-identical figure at Coupang’s Series C: four months for Sequoia Capital to secure consent from roughly twenty Korean shareholders before its capital could close. That article read the number as contract architecture: the cost of reconciling individual veto rights with a syndicated global round. The number recurs at Kurly because the underlying problem recurs. What took four months to fix, underneath the contract, was a governance structure with no board discipline, no independent oversight, and no mechanism for resolving disagreement short of renegotiating from zero each time an investor changed.

Look at the same four months from the governance side rather than the contract side, and a different diagnosis appears. The consent-rights problem at Kurly’s Series C sat downstream of a company that had never been asked to operate a board an international investor would recognise. Eighteen investors held individual veto positions because no board structure existed to aggregate their interests into a single governing voice. The contract chaos was a symptom. The governance vacuum underneath it was the condition.

This is Park Hee-duk’s framing of the case, and it is the sharpest form the diagnosis takes across his interview record: Kurly’s internationalisation was gatekept by governance readiness, ahead of product-market fit. In his words: “For companies to expand globally, they need to be prepared. Global expansion is only possible when companies meet global standards.”

Kurly’s product worked. Its dawn-delivery model was, by the account of investors who examined it, a competitive asset, reproducible and defensible on its own merits. None of that reduced the four months required to rebuild the company’s governance into a form foreign capital could underwrite. The product question and the governance question, in this framing, are sequential.

Four months is what it costs to build, retroactively, the governance a company needed from the start. The remainder of this article examines why that governance was never built, and what building it actually requires.

What Global Governance Requires

Global governance, in the vocabulary an international investor actually uses, breaks into four requirements: board composition, documentation discipline, standardised reporting, and IP ownership clarity. None of them is exotic. All four are, for most Korean companies before a foreign round, unmet.

The first is board composition. An international investor expects a board with independent directors who owe no operational loyalty to the founder and no capital relationship to any single shareholder, and an audit committee that meets on a schedule, reviews financial controls, and produces minutes an external auditor can rely on. Korean companies at Series B or C scale typically have neither. The board, where one exists formally, is often an assembly of major shareholders rather than a governing body with independent judgement, closer to a shareholders’ meeting under another name than to the oversight structure a global fund’s own limited partners require it to confirm exists before capital moves.

The second is documentation discipline. Board resolutions, properly minuted and dated. Shareholder consents, executed on a standard template. IP assignments, executed at the moment intellectual property is created rather than reconstructed under deadline pressure once a due-diligence request arrives. Korean legal practitioners who work this seam professionally confirm the pattern: documentation chaos is the norm. The paperwork usually exists somewhere, unindexed and undated, spread across founders’ personal drives and early employees’ inboxes, in a form no due-diligence team can rely on without weeks of reconstruction.

The third is standardised reporting. An international investor expects a key performance indicator (KPI) cadence tracked as a matter of course, set well ahead of any data room. It expects financial controls that produce numbers consistent from one reporting period to the next, prepared to a standard an outside auditor recognises. Korean companies commonly report to domestic investors on an ad hoc basis, numbers assembled when asked rather than tracked continuously against a standing framework. The difference is one of infrastructure more than rigour: nobody built the system that produces the numbers on a schedule, because nobody asked for one on a schedule until the foreign round arrived.

The fourth is IP ownership clarity, and it is the requirement founders most often discover too late. Employee invention agreements, in the Korean corporate default, are frequently informal or unresolved: engineers build the product; the company assumes it owns what they build; nobody signs anything that says so. International investors do not make that assumption. Diligence counsel for a foreign round will ask, company by company and patent by patent, who actually owns the intellectual property the valuation is built on. Where the answer is unclear, the round slows, or the price moves, or both.

Korea’s own governance law has begun to move. The Commercial Act amendment that took effect in July 2025 extended a director’s duty of loyalty from the company alone to the company and its shareholders, the fiduciary standard international investors rely on. Much of the wider reform package around it is drawn for listed companies, and what a duty owed to shareholders asks of a private venture-stage board is a question Korean practice is still answering. A standard has been raised. The board file at a Series B or C company reads much as it did before.

The contract layer moved next. On 30 June 2026 the standard venture investment contracts were revised for the first time in three years, and the revision separated the investment agreement from the shareholders’ agreement. It is in the shareholders’ agreement that unanimous investor consent gives way to collective consent within a round, at a two-thirds threshold: the harmonisation problem this article opened with, addressed at its source. The revision is a model contract rather than a rule, and it governs only agreements written after it. Korea has named the problem and solved it forward, which leaves every cap table assembled before June 2026 exactly where it stood.

Each of the four requirements is invisible under one specific condition: as long as no global investor has underwritten the company, none of them is tested. A Korean company can run for years on a board of major shareholders, informally minuted decisions, ad hoc reporting, and undocumented IP assignments, and nothing in its domestic operating environment will surface the gap. Domestic investors do not ask these questions, because domestic investors operate inside the same convention the company does.

That is what makes the moment foreign capital arrives distinctive. That moment does more than test price or terms. It is the first moment the company’s governance is measured against a standard it was never built to meet, by a counterparty with no reason to compromise on the measurement.

Why Founders Do Not Know What They Do Not Know

The natural response to the governance gap, at the policy level, is procedural: publish the requirements, run the training programme, close the awareness deficit. That response assumes founders do not know the rules. The more accurate diagnosis is narrower and harder to fix: founders do not know the rules exist, because nobody inside their operating environment has ever had reason to invoke them.

A Korean founder building a company through Series A and Series B, funded by domestic venture capitalists operating under domestic convention, is never asked to seat an independent director. Domestic term sheets do not require it. A Korean founder is never asked to run an audit committee, because the shareholder base evaluating the company does not expect one. Korean corporate counsel, retained to close domestic rounds competently and on schedule, is rarely asked to build a governance file that would satisfy a foreign investor’s diligence team, because the founder has never needed one and the counsel has never been asked to deliver one.

This is a familiar structural problem wearing a specific costume: competence inside one system does not transfer automatically to a different system, particularly when the first system gives no signal that a second, incompatible standard exists. A founder who has closed three funding rounds successfully, on time, at reasonable valuations, has every reason to believe the company’s governance is adequate. It has been adequate, for every purpose the company has so far been tested against.

The awareness gap belongs to the system the founder operates inside. Nothing about running a company well by domestic standards prepares a founder for the itemised list of requirements a foreign round will impose. Domestic standards and the foreign standard are separate specifications, built by separate institutional histories, that happen to apply to the same company at different points in its life.

The gap becomes visible only when a global counterparty appears and asks the questions domestic capital never asked. Before that moment, the governance a company has is, from the inside, simply the governance a company has. No signal inside the domestic system tells a founder the standard is about to change. The Korean legal profession, for the most part, serves the domestic system it operates inside. A founder’s own counsel is retained to close Korean rounds under Korean convention, and is rarely instructed to build toward a standard no domestic counterparty will test. Where the demand has never existed, neither has the practice.

The consequence surfaces at exactly the wrong moment: not during a calm planning cycle when it could be addressed at leisure, but in the middle of a live fundraising process, under a term sheet with a closing date, when a foreign investor’s diligence team asks for documents that do not exist in the form requested and a board structure the company has never operated.

Governance as Pre-Condition

Park’s Kurly case study makes an argument stronger than “governance matters.” It makes the argument that governance is the pre-condition, sequenced ahead of the product-market question a foreign investor is nominally there to evaluate.

This is an inversion of how the domestic system trains founders to think. Inside Korea’s venture convention, the operating assumption is that product and traction come first: build something that works, prove customers want it, and the capital follows because the numbers earn it. That sequence holds for domestic capital. It is incomplete for the specific transaction a foreign round represents. A foreign investor evaluates product-market fit as one input among several, and treats governance readiness as a gate the company must clear before the product evaluation reaches term-sheet stage. A company that cannot demonstrate a board, a documentation trail, and clean IP ownership does not get to the point where its product is judged on its merits. The transaction stalls at the gate, regardless of how strong the underlying business is — Kurly’s own four months, examined above, is the case in point: a competitive product bought no discount on the governance rebuild.

The practical implication is severe, and most Korean founders discover it only once they are already inside a live negotiation. Meeting the four requirements set out above is not a weekend’s paperwork; each takes years to build properly and cannot be manufactured to order once a term sheet appears. A company that wants global capital must, in effect, rebuild itself before it can be underwritten.

This rebuild is neither fast nor free. By the pattern this series has observed across the governance requirement set, the process runs six to eighteen months for a company starting from a typical Korean domestic governance baseline, longer where the documentation gap runs deep or the IP ownership questions touch multiple generations of early hires. The cost runs beyond legal fees, into the operating time of a founder and a leadership team who are simultaneously expected to run the company, close the round, and rebuild the governance the round requires, on overlapping calendars.

Here is where the mismatch bites hardest. Most Korean founders do not have that runway when the foreign investor actually arrives. A company that has grown to the point where it attracts serious international interest is, by definition, past the stage where a leisurely governance rebuild fits comfortably into the operating calendar. The foreign investor’s interest itself creates the time pressure: a term sheet carries a closing window, a competing domestic round may be running in parallel, and the company’s own growth trajectory does not pause while independent directors are recruited and an audit committee learns to function. The governance rebuild that should, ideally, have started years earlier is instead compressed into the months a live transaction allows.

This is the structural bind Kurly’s Series C made visible, and it recurs wherever Korean companies reach the threshold of serious foreign interest. Governance readiness is infrastructure: either built during the years when there was no external pressure to build it, or absent when the foreign investor finally arrives, in which case the company pays, in months, in negotiating leverage, and sometimes in valuation, for infrastructure it should already have had.

The Translation Gap

Step back from Kurly’s specific four months, and the pattern generalises across every serious cross-border transaction this series has examined. Article 2 found it in contract architecture: the individual consent rights, the club-deal structure, the absence of a syndication convention that would let a lead investor bind the round. Article 4 finds the same pattern one layer down, in governance: no independent board function, no standing audit discipline, no documentation trail, no settled IP ownership. Different layer, same shape. A domestic standard, internally coherent and functionally adequate for the domestic system it serves, meets a global standard built by a different institutional history, and the two do not translate on contact.

The translation has to happen somewhere. Two ends of it already exist and are already visible: the Korean domestic governance convention the company was actually built under, and the independent global governance standard — assembled across jurisdictions and owed to none of them alone — the foreign investor’s own institutional obligations require it to enforce. What sits between them is the actual mechanism: the way a company’s governance moves from one standard to the other on a timeline a live transaction can survive. This series has so far only described that mechanism by its absence: the four months Kurly spent building it under deadline pressure, the four months Coupang spent building the contract equivalent under the same pressure two years earlier.

Every cross-border investment relationship this series has examined needs that translation to exist somewhere, performed by someone, before capital can move. Where it does not exist in advance, the company builds it in real time, expensively, during the transaction that most needs it to have been built already. Where it does exist in advance, the translation happens off the critical path, and the four months these two articles have each measured, in two different companies, at two different layers of the same structural problem, does not need to happen at all.

A gap this specific, this recurring, and this costly does not stay unaddressed indefinitely. Somewhere, the translation is being built. The rest of this series is, in part, the search for where.

Sources:

• Park Hee-duk, interview and lecture series, eight items 2022–2025. V1–V3, V5, V6 video interviews on Korean channels (이해, 야단법석, VC인터뷰). Consolidated analysis on file.

• Park Hee-duk, V6 governance case study (Market Kurly) and consolidated findings memorandum, 17 March 2026: on file.

• 벤처투자 표준계약서 (standard venture investment contracts), revised edition announced by 중소벤처기업부 (Ministry of SMEs and Startups) and 한국벤처투자 (Korea Venture Investment Corporation) at the 벤처투자 계약문화 발전 선포식 (Venture Investment Contract Culture Development Declaration Ceremony), 30 June 2026 — the first revision in three years. The revision separates the investment agreement from the shareholders’ agreement; in the shareholders’ agreement, unanimous investor consent is replaced by collective consent of holders of two-thirds or more within the same round. A model contract; adoption is voluntary and it applies to agreements concluded after that date.

• Korean Commercial Act amendment (이사 충실의무 확대 / directors’ duty of loyalty extended to “the company and its shareholders”): passed the National Assembly 3 July 2025, in force 22 July 2025.

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Since 2016.

Saint Clair Capital · 그라운드 트루스 | 2026년 7월

핵심 요약

해외 자본이 마켓컬리 시리즈 C에 들어왔을 때, 정작 발목을 잡은 것은 거버넌스였다. 회사의 기존 국내 주주인 기관투자자 열여덟 곳의 동의권을 하나로 모으는 데만 넉 달이 걸렸다. 앞서 이 시리즈의 두 번째 글이 쿠팡의 계약에서 확인한 그 넉 달이, 이번에는 한 겹 아래 거버넌스에서 똑같이 되풀이됐다. 해외 투자자가 원하는 거버넌스는 분명하다. 독립적인 이사회와 실제로 돌아가는 감사위원회, 이사회 결의와 지식재산권(IP) 양도를 빠짐없이 정리한 서류, 정해진 주기로 이뤄지는 표준 보고, 그리고 기업가치의 근거가 되는 IP의 소유권. 이 네 가지다. 해외 투자 유치에 나서기 전에 이 조건을 모두 갖춘 한국 기업은 드물다. 국내에서는 애초에 이런 것을 요구한 적이 없고, 그때까지 창업자에게 ‘주변 누구도 지키지 않는 기준을 미리 갖춰 두라’고 일러 준 사람도 없었기 때문이다. 박희덕의 컬리 사례는 여기서 한발 더 나아간다. 거버넌스 준비가 제품·시장 적합성(PMF)보다 먼저 거래를 여닫는 관문이 되고, 그것을 다시 갖추는 데만 6개월에서 18개월이 걸리는데, 막상 해외 투자자가 눈앞에 앉은 시점에 대다수 창업자에게는 그럴 시간이 없다는 것이다. 그리고 이 이야기는 컬리 한 곳으로 끝나지 않는다. 이 시리즈가 다룬 모든 국경 간 거래에는 한국의 국내 기준과 글로벌 기준을 잇는 ‘번역’이 필요했다. 그 번역이 미리 돼 있으면 자본은 곧바로 움직이고, 그렇지 않으면 기업이 마감에 쫓기며 그 번역을 직접, 값비싸게 떠맡는다.

컬리의 넉 달

해외 자본은 마켓컬리 시리즈 C에 들어오기 전부터 무엇을 원하는지 알고 있었다. 전국 단위의 새벽 배송, 다른 시장의 투자자들이 봐도 경쟁력이 분명한 상품기획(MD) 선별 시스템, 그리고 한국 전자상거래에 없던 인프라로 커 가는 회사였다. 그러나 실사가 제품에서 서류로 넘어가는 순간, 그들 앞에 나타난 것은 전혀 다른 종류의 회사를 위해 지어진 거버넌스 구조였다.

그 무렵 컬리 캡테이블에는 기존 국내 주주인 기관투자자 열여덟 곳이 지분을 나눠 들고 있었다. 저마다 한국식 주주간계약 관행에 따라 협상한 동의권을 쥐고 있었고, 그 권리는 주주마다 따로 갖고 따로 행사하는 것이었다. 새 투자자가 들어오려 할 때마다 하나하나 다시 협상해야 했던 것도 그래서다. 이 열여덟 갈래의 지분을 새 자본이 실제로 감당할 하나의 구조로 묶는 데 넉 달이 걸렸다.

이 시리즈의 두 번째 글도 쿠팡 시리즈 C에서 거의 같은 넉 달을 마주했다. Sequoia Capital이 자본을 넣기까지 한국 주주 스무 명 안팎의 동의를 받는 데 걸린 시간이다. 그 글은 이를 계약 구조의 문제로 봤다. 개별 거부권을 글로벌 신디케이트 라운드에 맞추는 비용이라는 것이다. 컬리에서 같은 넉 달이 되풀이된 것도, 그 아래 깔린 문제가 똑같기 때문이다.

그 문제를 거버넌스의 눈으로 보면 진단이 또렷해진다. 컬리에는 국제 투자자가 인정할 이사회도, 독립적인 감시 장치도 없었다. 열여덟 투자자가 저마다 거부권을 쥔 것도, 그들의 이해를 하나로 묶어 줄 이사회가 없어서다. 투자자가 새로 들어올 때마다 원점에서 다시 협상하는 것 말고는 이견을 풀 길이 없었다. 계약의 혼란은 겉으로 드러난 증상일 뿐, 진짜 원인은 그 아래 거버넌스의 공백이었다.

박희덕의 여러 인터뷰를 통틀어, 그가 이 사례를 가장 날카롭게 파고든 지점이 바로 여기다. 컬리의 국제화를 가로막은 것은 제품이 아니라 거버넌스 준비였다는 것이다. 그의 말은 이렇다. “기업이 세계로 나가려면 준비가 돼 있어야 한다. 글로벌 기준을 갖춰야 비로소 세계로 나갈 수 있다.”

컬리의 제품은 통했다. 새벽 배송 모델은 이를 검토한 투자자들도 인정한 경쟁력이었다. 그 자체로 탄탄했고, 쉽게 따라 하기도 어려웠다. 그러나 그런 강점도, 해외 자본이 받아들일 거버넌스를 다시 세우는 데 든 넉 달을 줄여 주지는 못했다. 박 대표에게 순서는 분명하다. 거버넌스가 제품에 앞선다.

결국 넉 달은, 처음부터 있었어야 할 거버넌스를 뒤늦게 갖추는 값이다. 그 거버넌스가 왜 없었는지, 갖추려면 무엇이 필요한지를 아래에서 짚는다.

이사회부터 IP까지, 해외 자본이 요구하는 네 가지

글로벌 거버넌스는 국제 투자자가 실제로 쓰는 말로 옮기면 네 가지 요건으로 나뉜다. 이사회 구성, 문서화 규율, 표준 보고, 그리고 지식재산권 소유의 명확성이다. 어느 하나 특별할 것이 없다. 그런데도 해외 라운드를 앞둔 한국 기업 대다수는 이 넷 중 어느 것도 온전히 갖추지 못하고 있는 것이 현실이다.

첫째는 이사회 구성이다. 국제 투자자가 기대하는 이사회에는 독립적인 사외이사가 있어야 한다. 창업자에게 매여 있지도, 특정 주주와 자본으로 얽혀 있지도 않은 사람들이다. 감사위원회도 마찬가지다. 정해진 일정에 따라 열려 재무 상태를 점검하고, 외부 감사인이 믿고 쓸 수 있는 의사록을 남겨야 한다. 그러나 시리즈 B나 C 단계의 한국 기업은 대개 이 둘 다 갖추지 못한다. 이사회가 형식적으로 있더라도, 독립적으로 판단하는 기구라기보다 주요 주주들이 모이는 자리인 경우가 많다. 글로벌 펀드는 자본을 넣기 전에 자사 LP에게 ‘우리가 감독하는 이사회가 있다’는 사실부터 확인시켜야 하는데, 한국 기업의 이사회는 그 눈높이에 한참 못 미친다. 사실상 이름만 바꾼 주주총회다.

둘째는 문서화 규율이다. 이사회 결의는 날짜를 붙여 정식 의사록으로 남기고, 주주 동의는 표준 양식으로 받아 둔다. 지식재산권 양도도 실사 요청이 온 뒤 부랴부랴 짜맞출 것이 아니라, 그 권리가 생기는 순간 곧바로 서면으로 마무리해 둬야 한다. 이 일을 전문으로 다루는 한국 변호사들도 같은 이야기를 한다. 문서 관리가 엉망인 것이 오히려 보통이라는 것이다. 서류는 대개 어딘가에 있긴 하다. 그러나 색인도 날짜도 없이 창업자의 개인 드라이브와 초기 직원의 메일함에 흩어져 있어, 몇 주를 들여 처음부터 다시 정리하지 않고서는 어떤 실사팀도 믿고 쓸 수 없다.

셋째는 표준 보고다. 국제 투자자는 데이터룸을 열기 한참 전부터 핵심성과지표(KPI)를 당연한 일상처럼 챙긴다고 전제한다. 회계 기간이 바뀌어도 앞뒤가 맞는 숫자가, 그것도 외부 감사인이 인정하는 수준으로 나오기를 기대한다. 반면 한국 기업은 국내 투자자에게 그때그때 필요할 때만 보고하는 일이 잦다. 정해진 틀에 맞춰 꾸준히 챙기기보다, 요청이 와야 비로소 숫자를 맞춘다. 이 차이는 한국 기업이 덜 꼼꼼해서가 아니라, 정해진 주기로 숫자를 뽑아내는 체계 자체가 없어서 생긴다. 해외 라운드가 오기 전에는 그런 보고를 요구하는 쪽이 없었으니, 그 체계를 만들 이유도 없었다.

넷째는 지식재산권을 누가 갖고 있느냐가 분명해야 한다는 것이다. 창업자가 가장 뒤늦게, 대개는 너무 늦게 깨닫는 대목이다. 한국 기업에서는 관행적으로, 직원이 만든 발명을 회사가 넘겨받는 계약, 곧 직무발명 승계를 대충 넘기거나 아예 매듭짓지 않는 경우가 흔하다. 개발자가 제품을 만들면, 회사는 당연히 그 결과물이 자기 것이라고 여긴다. 그런데 정작 ‘그렇다’고 못 박은 서류에 서명해 둔 사람은 아무도 없다. 해외 투자자는 이것을 당연하게 여기지 않는다. 해외 라운드의 실사를 맡은 법률 자문은 기업가치의 근거가 된 지식재산권을 두고, 회사마다 또 특허 하나하나마다 ‘실제 주인이 누구냐’를 따진다. 답이 흐릿하면 라운드가 늦춰지거나 가격이 깎이거나, 둘 다 벌어진다.

한국의 거버넌스 법제도 움직이기 시작했다. 2025년 7월 시행된 개정 상법은 이사의 충실의무 대상을 ‘회사’에서 ‘회사와 주주’로 넓혔다. 해외 투자자가 신뢰의 바탕으로 삼는 것이 바로 이 신인의무 기준이다. 다만 이를 둘러싼 더 넓은 개혁은 대부분 상장사를 겨냥한 것이고, 주주에 대한 충실의무가 비상장 벤처 단계의 이사회에 무엇을 요구하는지는 한국의 실무도 아직 답을 찾는 중이다. 기준은 높아졌다. 그러나 시리즈 B나 C 기업의 이사회 서류는 예전과 별로 다르지 않다.

계약 쪽에서도 곧 변화가 이어졌다. 2026년 6월 30일 벤처투자 표준계약서가 3년 만에 개정되면서 투자계약서와 주주간계약서가 분리됐다. 투자자 전원의 만장일치를 받아야 하던 사전동의권은 이 가운데 주주간계약서에서, 한 라운드 안 3분의 2 이상의 집합적 동의로 바뀌었다. 이 글이 첫머리에서 던진, 흩어진 동의권을 하나로 모으는 문제를 뿌리에서 손본 셈이다. 다만 이는 강제 규정이 아니라 표준계약서일 뿐이어서, 개정 이후에 새로 쓴 계약에만 적용된다. 한국은 문제를 제대로 짚었고, 앞으로 맺을 계약에는 해법을 마련했다. 다만 2026년 6월 이전에 짜인 캡테이블은 하나같이 예전 그대로다.

이 네 가지 요건은 한 가지 조건에서는 아예 보이지 않는다. 글로벌 투자자가 그 회사에 자본을 대기 전까지는, 어느 요건도 시험대에 오르지 않는다. 한국 기업은 주요 주주로 채운 이사회, 문서로 제대로 남기지 않은 의사 결정, 그때그때의 보고, 서면으로 정리하지 않은 지식재산권 양도만으로도 여러 해를 버틸 수 있다. 국내에서 사업하는 동안에는 이 빈틈이 겉으로 드러날 일도 없다. 국내 투자자 역시 이런 질문을 던지지 않는데, 그들도 회사와 똑같은 관행 안에서 움직이는 탓이다.

그래서 해외 자본과 처음 마주하는 순간은 남다르다. 가격이나 조건을 따지는 자리가 아니다. 회사의 거버넌스가, 애초에 맞출 생각도 없던 기준 앞에 처음으로 놓이는 순간이다. 그것도, 한 치 봐줄 이유가 없는 상대 앞에서다.

세 번 라운드 닫은 창업자도, 규칙의 존재를 몰랐다

이 격차 앞에서 정책은 자연히 절차적인 해법으로 기운다. 요건을 알리고, 교육 프로그램을 돌리고, 부족한 인식을 채워 주는 식이다. 이런 접근은 창업자가 규칙을 모른다고 전제한다. 그러나 진짜 문제는 그보다 좁고, 그래서 더 어렵다. 창업자는 규칙을 몰라서가 아니라, 그런 규칙이 있다는 사실조차 모른다. 그가 몸담은 세계에서는 누구도 그것을 입에 올릴 일이 없었기 때문이다.

국내 벤처캐피털의 돈으로 시리즈 A와 B를 지나온 한국 창업자에게, 독립적인 사외이사를 앉히라고 요구하는 사람은 아무도 없다. 국내 텀시트가 요구하지 않고, 회사를 평가하는 주주들도 감사위원회 같은 것은 기대하지 않는다. 창업자의 법률 자문 역시 국내 라운드를 제때 마무리하려 선임된 사람들이라, 해외 실사팀을 만족시킬 거버넌스 서류를 만들어 달라는 주문을 받을 일이 없다. 찾는 사람이 없으니, 준비하는 사람도 없다.

사실 이것은 익숙한 구조적 문제가 모습만 바꿔 다시 나타난 것이다. 한 체계에서 갈고닦은 역량이 다른 체계에서도 그대로 통하지는 않는다. 더구나 첫 번째 체계가 ‘다른 기준이 있다’는 신호조차 주지 않으면, 그 간극은 더 벌어진다. 세 번의 라운드를 제때, 합리적인 기업가치로 성사시킨 창업자라면 자기 회사의 거버넌스가 충분하다고 믿는 것도 당연하다. 실제로 그 거버넌스는 지금까지 회사가 치른 모든 시험을 무리 없이 통과해 왔다.

결국 이 격차는 창업자 개인이 아니라 그가 속한 시스템의 문제다. 국내 기준과 해외 기준은 서로 다른 역사가 만든 별개의 규격이고, 한 회사의 생애에서 시점만 달리해 차례로 적용될 뿐이다. 그런데 머지않아 그 기준이 바뀐다는 신호를, 국내 시스템은 창업자에게 주지 않는다. 법조계마저 대체로 그 안에서 움직이니, 국내에서는 아무도 시험하지 않을 기준을 미리 갖춰 두라 요구하는 이도 없다. 수요가 없으니, 그 일을 해낼 실무도 자리 잡지 못했다.

그 결과는 하필 가장 나쁜 순간에 터져 나온다. 느긋하게 손볼 수 있는 평온한 시기가 아니라, 마감일이 박힌 텀시트를 앞에 두고 자금 조달이 한창인 와중이다. 바로 그때 해외 실사팀이 요구한다. 요청한 형태로는 존재하지도 않는 서류와, 한 번도 운영해 본 적 없는 이사회를.

제품보다 거버넌스가 먼저, 재정비엔 6~18개월

박희덕의 컬리 사례는 ‘거버넌스가 중요하다’는 말을 넘어선다. 거버넌스가 아예 전제 조건이고, 해외 투자자는 겉으로 제품과 시장을 보러 왔더라도 실제로는 그것부터 따진다는 것이다.

이는 국내가 창업자에게 심어 온 순서를 뒤집는다. 한국 벤처판의 전제는 ‘제품이 먼저’다. 제대로 돌아가는 제품으로 고객 수요를 증명하면 그 숫자만큼 자본이 따라붙는다는 것이다. 국내 자본에는 맞는 순서지만, 해외 라운드에서는 다르다. 해외 투자자에게 제품·시장 적합성은 여러 잣대 중 하나일 뿐, 거버넌스 준비가 그 평가에 앞선 관문이다. 이사회도, 문서 이력도, 정리된 지식재산권 소유도 내놓지 못하는 회사는 제품의 실력을 겨뤄 볼 자리에조차 서지 못한다. 사업이 아무리 탄탄해도 거래는 그 관문에서 멈춘다. 앞서 본 컬리의 넉 달이 그 증거다. 제품이 뛰어나도, 거버넌스를 다시 세우는 일 앞에서는 한 푼도 깎아 주지 않았다.

현실에서 이 말의 무게는 가볍지 않다. 네 가지 요건은 주말 하루에 채울 수 있는 것이 아니라, 하나하나 갖추는 데 몇 년이 걸린다. 이 시리즈가 지켜본 바로는, 전형적인 국내 수준에서 출발하는 회사에 6개월에서 18개월, 문서 공백이 깊거나 지식재산권 문제가 초기 직원 여러 세대에 얽혀 있으면 그 이상이다. 비용도 법률 수수료로 끝나지 않는다. 창업자와 경영진은 회사를 굴리고, 라운드를 닫고, 거버넌스를 다시 세우는 일을 같은 시기에 한꺼번에 떠안는다. 그들의 시간이 통째로 들어간다.

어긋남이 가장 뼈아픈 지점은 여기다. 정작 해외 투자자가 나타난 그때, 대다수 창업자에게는 그럴 여유가 없다. 해외의 굵직한 관심을 받을 만큼 큰 회사라면, 느긋하게 거버넌스를 손볼 단계는 이미 지났다. 관심 자체가 시간을 죈다. 텀시트에는 마감이 걸려 있고, 경쟁 라운드가 나란히 돌아가기도 하며, 사외이사를 뽑고 감사위원회가 자리 잡는 사이에도 회사는 계속 자란다. 그래서 몇 해 전에 시작했어야 할 재정비가, 거래가 허락하는 몇 달로 몰린다.

이것이 컬리 시리즈 C가 드러낸 구조적 딜레마이고, 한국 기업이 해외의 진지한 관심을 받는 문턱마다 되풀이된다. 거버넌스 준비는 결국 인프라다. 압력이 없던 시절에 미리 쌓아 두든가, 아니면 정작 필요한 순간에 그 자리에 없든가 둘 중 하나다. 없다면 그 값은 뒤늦게 돌아온다. 몇 달의 시간으로, 협상력으로, 때로는 기업가치를 깎아 가며.

번역의 공백

컬리의 넉 달에서 한 걸음 물러서면, 같은 패턴이 이 시리즈가 다룬 굵직한 국경 간 거래마다 어김없이 나타난다. 두 번째 글은 그 문제를 계약 구조에서 찾아냈다. 투자자마다 따로 쥔 동의권, 클럽딜 구조, 라운드 전체를 묶어 줄 신디케이트 관행의 부재였다. 이번 글은 같은 패턴을 그 아래 거버넌스에서 확인한다. 독립적 이사회도, 상시 감사 규율도, 문서 이력도, 정리된 지식재산권 소유도 없다. 층위는 달라도 형태는 똑같다. 국내 기준은 제 시스템 안에서는 앞뒤가 맞고 잘 돌아간다. 그러나 다른 역사가 세운 글로벌 기준과 부딪치는 순간, 둘은 서로 말이 통하지 않는다. 번역이 필요해지는 것이다.

번역은 어디선가 반드시 일어나야 한다. 양쪽 끝은 이미 뚜렷하다. 한쪽은 그 회사가 실제로 발 딛고 자라 온 한국의 거버넌스 관행, 다른 한쪽은 여러 나라 제도가 뒤섞이되 어느 하나에도 매이지 않은 글로벌 기준이다. 해외 투자자가 제 의무 때문에 반드시 지켜야 하는 기준이기도 하다. 문제는 그 사이를 잇는 장치다. 회사의 거버넌스를, 거래가 버텨 낼 시간 안에 한 기준에서 다른 기준으로 옮기는 방법 말이다. 이 시리즈는 지금까지 그 장치를 오직 ‘없다’는 사실로만 보여 줬다. 컬리가 마감에 쫓겨 그것을 만드느라 쓴 넉 달, 2년 전 쿠팡이 같은 압박 속에 계약 차원에서 쓴 넉 달이 그 증거다.

이 시리즈가 살펴본 모든 국경 간 투자에서, 자본이 움직이려면 그 전에 누군가 어디선가 이 번역을 끝내 놓아야 한다. 미리 돼 있지 않으면, 회사는 그것을 실시간으로, 비싼 값을 치르며, 정작 이미 끝나 있어야 할 거래 도중에 만든다. 미리 돼 있으면, 그 일은 거래의 급한 일정에서 비켜나 조용히 이뤄진다. 그러면 두 회사에서, 같은 문제의 서로 다른 두 층위에서 각각 재어 본 그 넉 달은 아예 생기지 않는다.

이토록 구체적이고, 이토록 되풀이되며, 이토록 값비싼 격차가 언제까지나 방치될 리 없다. 지금도 어딘가에서 그 번역은 이뤄지고 있다. 이 시리즈의 남은 글들은, 어느 정도는, 그 자리를 찾아가는 여정이다.

Sources:

• 박희덕(Park Hee-duk), 인터뷰·강연 시리즈, 2022~2025년 8건. 한국 채널(이해, 야단법석, VC인터뷰)의 V1~V3·V5·V6 영상 인터뷰. 통합 분석 자료 별도 보관.

• 박희덕, V6 거버넌스 사례 연구(마켓컬리) 및 종합 결과 메모, 2026년 3월 17일: 별도 보관.

• 벤처투자 표준계약서(standard venture investment contracts) 개정판, 중소벤처기업부·한국벤처투자(KVIC)가 「벤처투자 계약문화 발전 선포식」에서 발표, 2026년 6월 30일. 3년 만의 개정으로 투자계약서와 주주간계약서를 분리했고, 주주간계약서에서 투자자 전원의 만장일치 동의를 같은 라운드 안 3분의 2 이상의 집합적 동의로 대체했다. 강제 규칙이 아닌 표준계약서로, 채택은 자율이며 그 이후 체결된 계약에 적용된다.

• 한국 상법 개정(이사 충실의무 확대 / 이사의 충실의무 대상을 ‘회사와 주주’로 확대): 2025년 7월 3일 국회 본회의 통과, 2025년 7월 22일 시행.

고지: 이 글은 정보 제공만을 목적으로 하며 투자 자문에 해당하지 않는다. 모든 결정은 독자적인 조사와 자격을 갖춘 자문가와의 상담을 바탕으로 내려야 한다.

세인트 클레어 소개: 세인트 클레어(Saint Clair)는 유럽과 아시아를 잇는 국경 간 투자사다. 2016년 이래 자본이 오가는 인프라까지 직접 구축해 왔다.

Saint Clair · Market Intelligence | July 2026

The Word Finland Earned

In the 1960s, West German commentators coined an unflattering word for a small country’s foreign policy: Finlandisation. Finland held elections and traded freely with the West. It had also learned, government after government, to calibrate every position against what Moscow would tolerate: a UN vote, an editorial line, a trade term. No single decision surrendered Finnish sovereignty. Three decades of small, defensible accommodations did. By the time a word existed for it, the drift had already become the condition.

Gradual shifts are the dangerous ones, because each accommodation feels harmless taken alone. Crisis announces itself and forces a response. Drift asks for nothing, right up until the accumulated concessions define what remains possible. Fifteen governments now trade inside RCEP, a bloc covering close to a third of world population and output. Intra-Asian trade has been growing faster than trade across the Pacific for several years running. A record share of the world’s central banks, 45 per cent on the most recent count, are adding to gold reserves explicitly to reduce dependence on any single reserve currency. Multi-power Asia is the operating condition already, and Singapore’s principals are inside it whether or not they have named it.

A Structural Property, Not a Posture

Neutrality is a posture, and postures are held at another party’s discretion. Finland’s neutrality lasted exactly as long as Moscow found it useful; when Moscow’s calculus changed in 2022, Helsinki abandoned seventy-five years of policy within fourteen months and joined NATO in April 2023. What had looked, from the outside, like a permanent feature of the European order turned out to have been conditional the entire time, on a condition Helsinki never controlled.

Optionality sits on the principal’s side of the ledger instead. A firm, a family, or a fund holds it when it retains the capacity to operate under several distinct futures at once, built on its own initiative, regardless of what any counterparty currently permits. Singapore’s own neutrality may or may not hold under the pressure this series has traced. Optionality is the discipline for operating well either way, and it comes apart into three buildable parts: where capital sits, which relationships stay credible, and whether the business can actually execute under more than one regime at once.

Where the Equity Sits

Capital optionality is the most literal of the three: where assets, entities, and decision rights physically sit, and under which jurisdiction’s law they can be reached.

TikTok is the clean illustration of losing it, precisely because the loss was not a decision the company made. Under the American divestiture law that took effect in 2024, ByteDance’s US operations were restructured into a joint venture in which Oracle, Silver Lake, and MGX collectively hold the majority stake alongside a ByteDance holding reduced to just under a fifth. Governance followed the same partition: user data now sits on Oracle’s infrastructure, content decisions run through US-based personnel, and outside auditors review the source code. This is what happens to a global platform when two jurisdictions each insist critical infrastructure sit on their side of the line and neither accepts the other’s assurances: a regulator ends up rewriting the ownership structure a boardroom once controlled. Sovereigns are hedging the same exposure at a much larger scale: China’s own share of trade settled in renminbi has risen from roughly a tenth in 2017 to close to a third today, precisely so fewer transactions depend on a currency Washington controls.

The lesson generalises past one platform. Capital optionality means holding the structure before the ultimatum arrives. A holding company, a treasury function, or a family office with assets legally reachable from only one jurisdiction discovers the exposure at the exact moment it can least afford to: when a regulator, not the principal, decides where the equity will sit.

The Client List That Survived a Ban

Network optionality is which relationships stay active and credible on both sides of a divide, independent of formal permission.

Taiwan Semiconductor Manufacturing Company supplied the clearest recent test. Washington allowed the blanket export authorisation covering routine chipmaking-equipment shipments into TSMC’s Nanjing facility to lapse at the end of 2025, alongside similar revocations for Samsung and SK Hynix. Nanjing is a small fraction of TSMC’s global capacity, serving mainly automotive customers on mature nodes, and the restriction threatened to cap it there indefinitely. TSMC’s public response avoided both confrontation and retreat: the company’s China president noted that Chinese clients could reach TSMC’s advanced-node capacity through its global network rather than Nanjing alone, pointing to Xiaomi’s own 3-nanometre chip as proof the route already worked. The regulatory door narrowed. The commercial relationship did not close behind it.

That is network optionality functioning as designed. The counterparty relationship survives a jurisdictional restriction because it never depended on a single node of the network to begin with. A firm whose Asia-Europe relationships run through one office, one bank, or one regulatory approval learns, usually at the worst possible moment, the lesson TSMC did not have to learn twice.

Twenty-Five Per Cent, On Purpose

Operational optionality is execution capacity: whether a business can actually produce, ship, and sell under more than one regulatory and supply-chain regime.

Apple’s shift of iPhone assembly is the current benchmark. India production, negligible five years ago, reached roughly fifty-five million units in 2025, one in every four iPhones made worldwide. Apple frames this explicitly as insurance: against a repeat of pandemic-era disruption, and against the tariff exposure that has attached to China-concentrated manufacturing since. Indian capacity also earns Apple something Chinese capacity cannot: compliance with local-content rules that keep the product price-competitive inside India’s own market. Apple retains its Chinese manufacturing at scale, and has built a second, genuinely functioning production geography alongside it, large enough to matter if the first were ever constrained.

The distinction that matters for principals well below Apple’s scale is between operational optionality and operational aspiration. A contingency plan that exists only on paper, a supplier relationship never tested at volume, a second jurisdiction registered but never used: none of it counts until it has run, under real conditions, at a scale that would matter. Apple’s twenty-five per cent is a real number because someone insisted on testing it well before it was needed.

The Bill for Keeping Both Doors Open

This costs something, and the series has said so throughout.

Optionality carries a coordination tax first: TSMC now manages export-control compliance, customer allocation, and capacity planning across three or more regulatory regimes at once, work a single-jurisdiction competitor does not carry. It carries a relationship tax second: sustaining credibility with counterparties on opposing sides of a divide takes time and trust capital that a firm aligned to one side can spend elsewhere. And it carries a focus tax third: the organisation running two or three parallel structures moves slower than the one that picked a lane and committed. Apple’s India build-out took years of deliberately resourced patience; a China-only strategy would have scaled faster, for a while.

Naming the cost is what makes the posture credible. A firm that claims optionality without naming what it costs has built a slogan. The expense is paid only because the alternative, alignment with its bundled exposure, is judged more expensive still.

What the Principal Does With This

A principal who has worked through exposure, advantage, application, and positioning arrives here with a diagnosis, not yet a posture.

The posture is this: build the structural capacity to operate under more than one plausible future, accept what that capacity costs, and watch for the signals indicating which future is gaining ground, without betting the whole structure on any single one of them. It is a standing disposition, applied continuously to whichever specific decision happens to be in front of the principal that week: a treasury domicile, a counterparty relationship, a manufacturing site. The three dimensions above are where to look. What each principal builds from them is theirs to construct.

Finland, Again

Return to Finland. The accommodations that produced Finlandisation were each defensible: a vote abstained from here, a trade term softened there. None looked like capitulation alone. Together, across three decades, they left a country with fewer options than it had started with, and nobody could name the day it happened.

This firm is built the same way: Singapore-headquartered, with standing offices in Seoul and London, structured to operate the Europe-Asia corridor under more than one version of how the region resolves.

Finland’s eventual reversal, joining NATO within months of deciding to, was possible only because the underlying capacity, industrial, institutional, military, had been quietly maintained beneath the accommodating surface the whole time. The posture changed overnight. The capacity behind it had not.

The principal who has structured for optionality notices the drift while it is still a handful of small decisions. The next one, whatever it is, finds the capacity already built.

Sources:

• ASEAN Secretariat / Wikipedia summary of RCEP membership, ratification and scope (15 members; ~30% of world population and GDP): https://en.wikipedia.org/wiki/Regional_Comprehensive_Economic_Partnership

• UNCTAD, Global Trade Update, December 2025 — global trade >$35 trillion in 2025; East Asia export growth and intra-regional trade growth: https://unctad.org/publication/global-trade-update-december-2025-global-trade-poised-record-breaking-2025-flows

• WTO, Global Trade Outlook and Statistics, October 2025 update: https://www.wto.org/english/news_e/news25_e/stat_07oct25_e.pdf

• World Gold Council, Central Bank Gold Reserves Survey 2026 — 45% of central banks planning to increase gold reserves; 74% expecting lower US dollar reserve share within five years: https://www.gold.org/goldhub/research/central-bank-gold-reserves-survey-2026

• IMF, World Economic Outlook, October 2025 — global and emerging-market growth trajectory: https://www.imf.org/en/publications/weo/issues/2025/10/14/world-economic-outlook-october-2025

• Forbes / RMB internationalisation data — China’s RMB trade settlement share (10% in 2017 to ~30% currently); CIPS network growth: https://www.forbes.com/sites/zennonkapron/2026/02/22/how-renminbi-internationalization-is-changing/

• ITIF, “Five Takeaways from the TikTok Deal,” January 2026 — divestiture structure, Oracle/Silver Lake/MGX joint venture, governance terms: https://itif.org/publications/2026/01/26/five-takeaways-from-the-tiktok-deal/

• TrendForce, “TSMC Dismisses Nanjing Risks as VEU Expires, Flags Global Advanced Capacity for Chinese Clients,” December 2025: https://www.trendforce.com/news/2025/12/26/news-tsmc-dismisses-nanjing-risks-as-veu-expires-flags-global-advanced-capacity-for-chinese-clients/

• South China Morning Post, “TSMC Japan plan for advanced chips seen as hedge against pressure from US, China”: https://www.scmp.com/tech/tech-war/article/3342524/tsmc-japan-plan-advanced-chips-seen-hedge-against-pressure-us-china

• TechBuzz, “Apple hits 25% India iPhone production as China risks mount”: https://www.techbuzz.ai/articles/apple-hits-25-india-iphone-production-as-china-risks-mount

• Bloomberg, “Apple Now Makes About 25% of iPhones in India After China Pivot,” March 2026: https://www.bloomberg.com/news/articles/2026-03-10/apple-now-makes-about-25-of-iphones-in-india-after-china-pivot

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Since 2016.

Learn more: saintclair.sg | Contact: contact@saintclair.sg

The European Union is the second-most-active investor in Indonesia by company count: 225 companies, per Tracxn. A further EUR 1 billion in indicative Team Europe contributions moves through the EU-ASEAN Sustainable Connectivity package, and GIZ already operates a Digital Transformation Center and Digital Hub Network across Jakarta and Bali. The corridor between European institutional capital and Indonesian venture opportunity exists. The commercial bridge across it does not.

Ground Truth: Indonesia | Saint Clair Market Intelligence | 27 July 2026

Based on the European Commission’s EU-ASEAN Sustainable Connectivity Package, the World Bank’s Indonesia Economic Prospects: Funding Indonesia’s Vision 2045, the US Department of State’s 2025 Investment Climate Statement, GIZ’s Digital Hub Network Guideline, and the Google / Temasek / Bain e-Conomy SEA 2025. Saint Clair’s analysis follows.

Two hundred and twenty-five European companies operate in Indonesia today. A billion euros in indicative institutional commitment moves through an active EU connectivity package. Germany’s development agency runs live digital-transformation programming in two of the country’s largest population centres. Every element of a European capital corridor into Indonesia already exists: funded, staffed, and running. What does not yet exist is the commercial firm positioned to move between the European institutions building that corridor and the Indonesian companies it is meant to reach.

Indonesia’s regulatory architecture has moved toward foreign capital by design, and the macro data has held through an external shock this year. The gap sits in domestic financial depth, and in the space between institutional grant-making and commercial venture return. What is different about the corridor with Europe is scale: it is the largest single pool of committed capital examined here, and the most fully built.

The Institutional Weight

The European Commission’s contribution to that architecture is direct: EUR 60 million allocated from the EU budget, against a total programme cost of EUR 90.3 million once co-financing from partner institutions is included. The programme sits inside a larger Team Europe Initiative on ASEAN Sustainable Connectivity, to which EU institutions and Member States have indicated a combined EUR 1 billion. A further EUR 29 million moves through the Energy Transition Partnership, a multi-donor trust fund managed by the United Nations Office for Project Services. The package is built to deliver five connectivity outcomes across trade, energy, digital infrastructure, transport, and people-to-people exchange, the last carrying an explicit mandate on gender equality in digital and green skills. Each outcome has its own regulatory and capacity-building workstream; together they read less like a grant programme than like the early architecture of a market relationship, built one institutional layer at a time.

None of this is a commitment awaiting deployment. GIZ, Germany’s development agency and the EU’s most visible implementing partner on the ground, already runs a Digital Transformation Center and a Digital Hub Network spanning Jakarta and Bali, working directly inside the incubator and accelerator ecosystem that produces Indonesia’s investable companies. The network’s own comparative mapping of existing hub operators concentrates on Java and Bali; the outer provinces see comparatively little of it, a distribution the institutional programme largely reproduces. The European Union’s position, second-most-active foreign investor in Indonesia by company count at 225 companies, sits on top of an institutional layer that is already staffed and operating, not one still being negotiated.

The Regulatory Frame

The corridor arrives into a regulatory environment that has moved, structurally, in the same direction. Indonesia’s Omnibus Law on Job Creation, revised and re-enacted as Law No. 6/2023, repealed the country’s 2016 Negative Investment List and replaced it with a default principle: every business sector is open to foreign investment unless a specific regulation says otherwise. Seven sectors remain closed. Twenty-four Special Economic Zones have been established under the same reform era, offering foreign technology investors exemptions from import and licensing requirements that apply elsewhere in the country. Three of the earliest, the free-trade islands of Batam, Bintan, and Karimun, sit just south of Singapore, positioned to draw exactly the cross-border capital that already moves through Singapore’s legal architecture on its way into the wider archipelago.

The macro backdrop has held, unevenly. Indonesia’s economy grew 5.0 per cent year on year in the first half of 2025. The United States’ 19 per cent reciprocal tariff, effective from August 2025, has since pulled the 2026 growth forecast down from 5.1 to 4.8 per cent. The regulatory direction is toward openness. The growth path is toward a slower, more contested version of the same trajectory, tested now by external demand shocks the country does not control. An investor evaluating Indonesia in 2026 is evaluating a market that has done the policy work openness requires.

The Finance Gap Underneath

The regulatory opening has not been matched by depth in domestic finance. Indonesia’s 2023 tax ratio stood at 10.2 per cent of GDP, among the lowest of its regional peers, with foregone VAT and corporate-income-tax revenue averaging 6.3 per cent of GDP between 2016 and 2021. Public capacity to co-finance the transition this corridor is meant to support is constrained by a tax base considerably smaller than comparable economies collect. Private credit tells a parallel story: credit to the private sector reached 31.4 per cent of GDP at the end of 2023, against an average of 124 per cent of GDP across peer East Asia and Pacific economies. Indonesia’s banking sector is well capitalised and stable. It is also concentrated, cautious about riskier lending, and structurally undersized relative to what an economy of this scale would typically generate in domestic credit. In practice, that caution routes bank lending toward established, asset-rich borrowers, leaving early- and growth-stage companies dependent on risk capital sourced from outside the domestic banking system.

The two constraints compound. A government with a shallow tax base has limited room to co-finance transition and connectivity spending at the pace its own regulatory ambitions imply. A banking sector with shallow private credit has limited appetite to underwrite the venture-stage risk that would convert institutional grant funding into scaled companies. The institutional corridor from Europe arrives into a market that has opened its doors, and has not yet built the internal financial depth to absorb the capital walking through them at commercial pace.

The Bridge With No Operator

The result is institutional traffic on one side of a bridge and nothing built to carry it across. Grant-funded ventures in Indonesia’s climate and environmental sector raise financing that is, by design, non-commercial: eighty per cent of support-organisation financing arrives as grants, more than half of the tickets written by debt or equity funders average below USD 25,000, and traditional investors show a documented reluctance to underwrite deals above USD 500,000. On the founder side, roughly fifteen per cent of Indonesian companies that raise a seed round progress to Series A. Between the grant-funded prototype and the institutional venture round sits the country’s most consistent point of company mortality. The pattern is not confined to climate technology. Indonesia’s USD 20 billion Just Energy Transition Partnership commitment sits in the same category as the EU’s own package: capital that is real, publicly committed, and designed to catalyse commercial deployment it does not itself supply.

GIZ’s presence illustrates the shape of the gap rather than the answer to it. Its Digital Transformation Center and Digital Hub Network build pipeline: they convene, train, and connect the incubators and accelerators that produce investable companies. A development agency that builds pipeline without underwriting the venture-stage risk that pipeline eventually needs is doing exactly what a development agency is for. No institution in the public architecture Europe has built for Indonesia currently supplies the commercial capital the next stage of that pipeline requires.

The gap carries a second dimension the funding numbers alone do not capture. The EU’s own public strategy for digital connectivity in the Indo-Pacific names the stakes explicitly: concentrated digital infrastructure creates strategic dependencies and geopolitical risk, and the EU’s stated response favours diversified, trusted suppliers and human-centric digital standards over the alternative models a Southeast Asian state can otherwise default to. Indonesia’s adoption of a European standard, in digital infrastructure as in capital, will be won on the same terms every corridor in Southeast Asia is contested: whether the approach demonstrably serves Indonesia’s interest better than the alternatives on offer. The institutional case has been made publicly and repeatedly, in funding cycles and connectivity packages. The commercial case is the one an operating bridge would make daily, and it has not yet been made at scale.

The Corridor View

Every element of this corridor exists in public record: two hundred and twenty-five European companies already invested, a billion euros in indicative Team Europe commitment moving through an active connectivity package, a development agency running functioning digital-transformation programming across the country’s largest population centres, and a regulatory environment opened by design. The missing element is a private-sector intermediary able to move between a Brussels-administered grant cycle and a Jakarta term sheet without losing time, trust, or return on either side of the transaction.

Development institutions carry policy risk. Commercial venture funds carry return risk. Between them sits a role neither is built to fill: standing credibility on both sides of the corridor, close enough to a grant committee to read its timeline and close enough to a term sheet to price it. That credibility takes years to build in either direction, which is precisely why the gap has persisted this long despite every institutional signal pointing toward it. Indonesia’s corridor with Europe is built, funded, and regulated. It does not yet have that operator.

Sources:

• European Commission, Action Document for the EU-ASEAN Sustainable Connectivity Package (ACT-61616, March 2023).

• World Bank, Indonesia Economic Prospects: Funding Indonesia’s Vision 2045 (December 2024).

• US Department of State, 2025 Investment Climate Statement: Indonesia (2025).

• GIZ, Digital Hub Network Guideline Research Report, Digital Transformation Center Indonesia (2024).

• Google, Temasek, Bain & Company, e-Conomy SEA 2025, 10th edition (November 2025).

• Tracxn, Indonesia investor-activity data, as referenced in Saint Clair’s Indonesia source-mapping analysis.

• Startup Genome and the Aspen Network of Development Entrepreneurs (ANDE), Southeast Asia Climate and Environmental Entrepreneurial Ecosystem Snapshot (2022).

• DailySocial / DiscoveryShift, Indonesia Startup Report 2026 (January 2026 release).

This article concludes the Ground Truth: Indonesia sub-series.

Disclaimer: This article is for informational purposes only and does not constitute investment advice. Saint Clair has an active engagement with frontier Asian startup ecosystems and may have commercial interests in markets discussed. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair designs and builds cross-border capital infrastructure between Europe and Asia — proposing access where access is scarce, and creating structure where structure is absent. Saint Clair Asia (saintclair.asia) is a frontier investment platform that positions international investors within innovation ecosystems that institutional channels do not reach.

Learn more: saintclair.sg | saintclair.asia | Contact: contact@saintclair.sg

Saint Clair Capital · Ground Truth | July 2026

Executive Summary

Across five years, Korea’s Ministry of SMEs and Startups has deployed more than $10 billion of public budget into the country’s startup sector. The Korea Fund of Funds has separately deployed approximately ₩10 trillion (roughly $6.7 billion at current exchange rates) cumulatively, with fresh mother-fund commitments still flowing into 2026. By funding volume, talent density and sector breadth, Korea now ranks among the world’s most heavily capitalised startup markets. Capital was never the constraint.

What the money could not buy is contract convention, limited-partner evaluation methodology, board governance norms and exit infrastructure calibrated to the global standard. Korean startups’ annual export revenue has grown ninefold since 2017; it still represents only 1.3 percent of total sector sales. New unicorn formation, on the report’s own accounting, has turned down even as the money kept arriving. The veteran practitioner Park Hee-duk frames Korea’s policy instinct as carpet bombing where precision targeting is required. Singapore’s Pavilion Capital deployed capital of a broadly comparable order and became a globally recognised institutional investor; Korea’s Fund of Funds, at similar scale, did not.

Ten Billion Dollars

Over the past five years, Korea’s Ministry of SMEs and Startups has deployed more than $10 billion of public budget into the country’s startup sector. The South Korean Startup Ecosystem Report, prepared by the Global Entrepreneurs Association and The Garrison, reaches for a yardstick to convey what that means: $10 billion is roughly the combined value of ten Korean unicorns. It is industrial-policy money.

Alongside that figure sits the Korea Fund of Funds, the government’s flagship vehicle, which has deployed approximately ₩10 trillion cumulatively, its mother-fund cycles continuing to commit fresh capital into 2026 to seed a new round of venture-fund formation. Government budget and government fund-of-funds capital are two lenses on the same order of magnitude, both landing in the multi-billion-dollar range across the same five-year window.

What kind of money this is matters as much as how much of it there is. Loan instruments, folded into the Ministry’s startup budget from 2022 as an early-stage survival measure through the pandemic period, now make up roughly half of the annual total, against a smaller and shrinking share for commercialisation grants. The research-funding share, by contrast, fell sharply in 2022 before partly recovering in the years since. Park Hee-duk’s account of Korean institutional capital gives this composition a name: banking-structure thinking, the habit of managing public capital for the certainty and repayment discipline a loan book requires, carried over into an asset class that is supposed to tolerate loss. A government startup budget built substantially on loan instruments behaves, by construction, more like a lender than a risk-capital allocator, regardless of the headline figure attached to it.

The report’s own reading of Startup Genome’s scoring places Korea’s startup sector among those with the largest public funding infrastructure in the world; the same reading scores Korea’s talent and knowledge base highly, supported by extensive government and university-linked research funding. The Ministry’s own 2025 policy document sets the ambition explicitly: the Comprehensive Measures for Becoming a Top 4 Venture Power, targeting fifty unicorns and decacorns and ten thousand artificial intelligence (AI) and deep-tech startups by 2030. Korea has, by the government’s own framing, entered the conversation about which countries lead global venture activity by scale.

The current ranking record is more modest than the ambition. Startup Blink’s national comparison places Korea in a cohort with Japan, Denmark and India rather than among the leading four. Rankings move slowly, and the underlying investment is recent by institutional standards. Scale and standing are different variables; standing requires its own investment beyond scale.

This is the scale the rest of the article measures against: what the money built, and what it did not.

If ten billion dollars over five years, a decade of Fund of Funds deployment, and a stated ambition to become a top-four venture power were enough to close the gap the first two articles in this series described, the gap would already be closing. It is not.

What the Money Built

Korea’s startup sector, by the quantitative measures that dominate international comparison, has built scale.

The trigger point was 2017: the abolition of Korea’s joint-guarantor system, which had previously exposed a founder’s extended family to personal liability for a failed venture. Removing that liability legitimised startup founding as a career choice in a way it had not been before. Domestic and foreign capital alike were willing to fund the resulting inflow of entrepreneurial talent. Deployment climbed through the years that followed, alongside a proliferation of government instruments layered across pre-seed, seed and growth stages: the Founder’s Academy, TIPS (the Tech Incubator Program for Startups) and its Pre-TIPS, Post-TIPS and Deep-Tech TIPS variants, and dozens of smaller regional and sector programmes.

Some of these instruments have worked well on their own terms. TIPS, the government’s most-cited success case, has attracted follow-up investment worth 10.4 times its invested budget over ten years; the roughly 3,200 companies that passed through it carry a combined corporate value near ₩88 trillion, and the programme has expanded from five operators in 2013 to more than a hundred. Unicorn formation, sparse and irregular through the 2000s and much of the 2010s, accelerated sharply in 2021 and 2022, concentrated in e-commerce, beauty and gaming. Korean startups’ annual exports have grown alongside the same deployment, rising ninefold since 2017 on the back of cosmetics, electronics and semiconductor-equipment sales abroad, a real export story, on its own terms, for a sector that barely sold outside Korea a decade ago.

By the comparative measures international observers use, this has shifted Korea’s position. Startup Genome’s Global Startup Ecosystem Report scores Seoul strongly on Funding and on Talent & Knowledge, the two dimensions most directly responsive to public capital and public education infrastructure. Korea ranks first globally in AI patent grants per capita as of 2023, with the eighth-fastest rate of increase over the prior decade; the country’s density of technical talent remains unusual by international comparison. Set against twenty-five major economies on a composite technology-competitiveness ranking spanning artificial intelligence, biotechnology, semiconductors, space and quantum, Korea places within the top ten in four of the five fields, a spread few mid-sized economies can match.

A government that spent $10 billion over five years, built the Fund of Funds into a ₩10-trillion institution, abolished a punitive liability regime and watched entrepreneurial formation respond, and produced a cluster of globally competitive companies, put real capital behind a real ambition. Whatever else is true about Korea’s startup sector, the money was there.

But quantitative scale and structural completeness are different achievements, and the sector’s own reviewers have begun to notice the distance between them. The report’s own unicorn data show that most of Korea’s unicorn companies, tallied by founding year, were established before the government’s active startup-support apparatus reached full scale. The money arrived; the response followed it loosely, not tightly. The annual pace of new unicorn formation, having climbed through the years of heaviest public deployment, has since turned downward. The report records the decline even as the funding meant to prevent it kept flowing.

Ten billion dollars built scale. The question is what it was built onto.

What It Did Not Build

Money scales in a way architecture does not. A government can add a zero to a startup budget within a single fiscal year. Adding a zero to the number of Korean limited partners who evaluate general partners on qualitative grounds rather than trailing quantitative returns takes longer. So does adding a zero to the share of Korean shareholder agreements that vest consent in a lead investor rather than in every named subscriber on the cap table, or to the number of Korean venture funds running ten-year-plus primary terms instead of eight. These are the structural findings of the first two articles in this series, and they recur here because they explain why the third billion behaves like the first.

Park Hee-duk’s framing for this gap, developed across his more recent interview record, remains the most useful single image available: capital is the rainfall; the surrounding institutional architecture is the soil. Korean unicorns of the past five years, in his account, raised capital at a scale their global peers would recognise. Many can no longer sustain the pace their early growth implied, because the soil they were built in never equipped them with the operating conventions a global exit, a global fundraise or global governance scrutiny requires. Rainfall on poor soil produces a flush of growth and then a plateau.

The evidence for the plateau sits inside the same report that celebrates Korea’s funding scale. The ninefold export growth noted above reached approximately $2.4 billion against total annual sector sales of $185.4 billion: 1.3 percent. Growth of nine times sounds emphatic until it is set against the base it grew from and the market it was meant to open. Five years on, and by this measure Korea’s startup sector remains almost entirely a domestic-market phenomenon. Startup Genome’s own scoring confirms it, from a different angle: Korea performs well on Funding and on Talent & Knowledge; it scores below average on Global Reach and on market accessibility for international entrepreneurs and investors, a pattern the report explicitly likens to Japan’s. Foreign participation inside Korean fund formation echoes the same pattern from the investor side: in the Korea Venture Capital Association’s own annual tally of who commits capital to new funds, the category grouping foreign and other outside institutional investors sits among the smallest slices year after year, typically in the low single digits. Mother-fund, pension and domestic-corporate money dwarfs it.

The domestic Fund of Funds architecture compounds the mismatch rather than correcting it. Government and policy-institution capital accounts for between 22 and 38 percent of limited-partner participation in new Korean venture funds, by the report’s own tally, a level of dependence the report itself flags as a structural concern. Within that Fund of Funds allocation, the sub-fund share reaching accelerators, the vehicles meant to handle earliest-stage risk, has held in a narrow six-to-nine percent band across 2020 to 2024, even though a conservative reading of the sector’s own observed early-mid-late-stage investment split runs roughly one to two to three. Government and policy-institution capital, taken together, form Korean venture’s largest limited-partner bloc, and its allocation discipline does not consistently match the risk the sector needs it to carry.

A larger cheque would not close this. A fund that runs an eight-year primary term does not become a ten-year fund because its allocation doubled. An evaluation methodology built around trailing quantitative returns does not turn qualitative because more capital flows through it. A cap table with twenty individual consent-rights holders does not consolidate into a syndicate structure because the round is larger. Each of these is a design choice, embedded in Korean fund documentation, limited-partner mandates and regulatory expectation, and each persists regardless of how much money moves through the system built on top of it.

Korean capital lands in the same structure it did five years and ten billion dollars ago. That is the distinction the rest of this article turns on.

Carpet Bombing vs Tomahawk

Park Hee-duk’s sharpest institutional critique, articulated in a July 2022 print interview, concerns not the scale of Korean startup policy but its targeting logic. Korea, in his framing, conflates startup policy with small-and-medium-enterprise (SME) policy: the same ministry, overlapping budget lines and the same instinct for broad dispersion. SME policy is supposed to disperse broadly. A functioning SME support regime spreads risk-tolerant capital across a wide population of small businesses precisely because most will remain small, and the goal is survival and stability across the base, not concentrated breakout. Startup policy, in Park’s account, requires the opposite instinct: precision targeting of the small number of companies capable of a global outcome, backed disproportionately once identified. Korea has spent five years applying carpet-bombing logic to a Tomahawk-shaped problem.

The Ministry’s own multi-year planning record supports the diagnosis from an unexpected angle. Reviewed across 2016 to 2025, the Ministry’s stated policy direction has shifted keyword and emphasis roughly every two years: from global expansion, to job creation, to digital transformation, to a startup-led economy, and back to global expansion again. Advanced markets the report cites for comparison, in Europe, Singapore and Taiwan, plan on seven-to-ten-year horizons. A two-year policy cycle is long enough to fund a broad population of applicants. It is not long enough to build the institutional patience precision targeting requires: identifying which companies merit disproportionate follow-on support, and staying with that judgment through a cycle or two of underperformance before it pays off.

TIPS, the programme this article credited earlier for real results, illustrates the same drift at closer range. The same report that documents its 10.4-times follow-on return also records a quieter side effect: completion of TIPS has become a form of certification in its own right, something struggling founders pursue to signal validity to later investors rather than a growth-stage springboard. Reproducing the programme’s scope and eligibility pool has, on the report’s own account, drawn attention away from deepening support at the growth stage or building the reinvestment habit where successful graduates back the next cohort. A well-designed instrument, run inside a dispersive system, drifts toward the system’s own logic over time.

The re-challenge fund, a real instrument in Korea’s current support architecture, draws Park’s clearest dismissal. He frames it as characteristic Yeouido thinking, named for Seoul’s political and financial district — policy that treats startup failure as a financial loss to be compensated, rather than recognising what a failed company leaves behind: a team, a technical capability, a piece of intellectual property. Silicon Valley’s venture culture recycles failure directly. Founders and early employees from a failed company routinely become the next generation of founders, funded in part by earlier winners paying forward what they learned and earned. Korea’s structural equivalent barely exists. The re-challenge fund, as designed, compensates the loss instead of preserving what the failure built.

Korea’s government could double the startup budget tomorrow and apply the same dispersion logic to twice the money. It would produce twice the breadth and no more depth. The mechanism, not the volume, is the constraint.

Pavilion

Singapore offers the clearest comparator available, and Park Hee-duk returns to it repeatedly: Pavilion Capital, the Temasek-linked global investment platform, deployed capital of a broadly comparable order to Korea’s Fund of Funds and became a globally recognised institutional investor in the process. Korea’s Fund of Funds, deploying at a similar scale over a similar period, has not achieved comparable international standing. Both economies sit inside the same regional neighbourhood, and both built their venture capacity substantially through government-linked institutions rather than private accumulation. The starting conditions were not so different that geography or history alone explains the outcome.

The variable that changed was not the size of the cheque. It was who wrote it, on what terms, and against what evaluation standard.

Park extends the comparison at the institutional level with Temasek and Korea’s National Pension Service. Temasek manages a smaller capital base than the National Pension Service. Its global influence, by any reasonable measure of international recognition and co-investment access, runs substantially greater. The mechanism Park identifies is personnel: Temasek recruited globally trained investment talent and built evaluation processes to match, and institutional competence followed the people rather than the balance sheet. A pension fund or fund-of-funds can, in principle, replicate a personnel strategy at any point regardless of its existing capital base. Few Korean policy-finance institutions have done so at scale.

Park’s sine-cosine problem sharpens the same point. Korea’s financial system, in his framing, runs on a single pro-cyclical curve: when public markets contract, venture capital contracts with them in lockstep, because the same institutions and the same risk appetite govern both. A healthy system needs a second curve, running counter to the first, so that capital available for venture investment does not evaporate at precisely the moment public-market sentiment sours. Korea’s Fund of Funds, evaluated on Korea-only terms by Korea-only personnel, has built no structural source for that second curve.

Singapore built an institution that global limited partners, global co-investors and global portfolio companies recognise as operating on their terms, evaluated by their standards, staffed by people trained inside their systems, at a capital scale Korea’s own Fund of Funds already matches.

Three articles now point at the same structure. The $22.9 billion maturity wall, the four-month contract negotiation, and the ten billion dollars that built scale without building the architecture underneath it are three views of it. More capital, deployed through the same architecture, produces more of the same outcome. The gap closes, if it closes, through the standards the capital operates under, not through the next budget cycle.

Sources:

• Global Entrepreneurs Association & The Garrison, South Korean Startup Ecosystem Report (30 December 2025, on file)

• Park Hee-duk, print interview in GoKorea magazine (July 2022) and six subsequent video interviews (2024–2025)

• Korea Venture Investment Corporation (KVIC), Korea Fund of Funds: https://www.kvic.or.kr/about-business

• Korea Government Policy Briefing, “2026 Mother Fund 1st regular allocation”: https://www.korea.kr/news/policyNewsView.do?newsId=148958492

• Ministry of SMEs and Startups (중소벤처기업부), Venture Four Great Powers Comprehensive Plan (벤처 4대 강국 도약 종합대책), 18 December 2025: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156736553

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Since 2016.

Saint Clair Capital · 그라운드 트루스 | 2026년 7월

핵심 요약

지난 5년간 한국 중소벤처기업부가 스타트업 부문에 투입한 공적 예산은 100억 달러가 넘는다. 여기에 한국벤처투자(KVIC) 모태펀드가 누적 약 10조 원(현재 환율로 약 67억 달러)을 별도로 집행했고, 모태펀드 출자는 2026년에도 이어진다. 자금 규모로 보나 인재 밀도로 보나 산업의 폭으로 보나, 한국은 이제 세계에서 가장 많은 자본이 투입된 스타트업 시장 축에 든다. 정작 부족했던 것은 다른 데 있었다.

돈으로 살 수 없었던 것은 계약 관행과 LP 평가 방식, 이사회 거버넌스 규범, 국제 기준에 맞는 회수 인프라였다. 한국 스타트업의 연간 수출은 2017년보다 9배로 늘었지만, 업종 전체 매출에서 차지하는 비중은 여전히 1.3%에 그친다. 새 유니콘이 생겨나는 속도는 보고서 자체 집계로도 이 자금이 계속 흘러드는 동안 오히려 꺾였다. 베테랑 VC 박희덕 대표는 정밀 타격이 필요한 자리에 한국의 정책이 융단폭격으로 대응해 왔다고 진단한다. 싱가포르의 파빌리온캐피털(Pavilion Capital)은 한국 모태펀드와 엇비슷한 규모의 자본을 집행하고도 세계가 인정하는 기관 투자자로 올라섰다. 같은 규모의 한국 모태펀드는 끝내 그러지 못했다.

100억 달러 쏟았지만, 이제 예산 절반이 융자다

지난 5년간 중소벤처기업부가 한국 스타트업 부문에 투입한 공적 예산은 100억 달러를 웃돈다. Global Entrepreneurs Association과 The Garrison이 함께 펴낸 『한국 스타트업 생태계 보고서(South Korean Startup Ecosystem Report)』는 이 액수를 가늠할 잣대 하나를 제시한다. 100억 달러가 한국 유니콘 열 곳의 기업가치를 합친 규모와 맞먹는다는 것이다. 벤처 지원이라기보다 산업 정책에 가까운 크기다.

여기에 정부의 대표 벤처 자금 창구인 한국벤처투자 모태펀드가 겹친다. 모태펀드는 누적 약 10조 원(현재 환율로 약 67억 달러)을 집행했고, 정례 출자 사이클은 2026년에도 새 벤처펀드를 키우기 위해 돌아간다. 정부 예산이든 모태펀드 출자든 창구가 다를 뿐, 같은 5년 동안 똑같이 수십억 달러대로 집행된 자금이다.

돈은 양만큼이나 성격이 중요하다. 코로나 시기 초기 생존책으로 2022년부터 중기부 스타트업 예산에 편입된 융자성 자금은 이제 연간 총액의 절반가량을 차지한다. 사업화 보조금 비중은 그만큼 줄었고, 연구개발 자금 비중은 2022년 크게 낮아졌다가 이후 일부 회복됐다. 박희덕 대표는 이런 자금 구성을 「은행식 사고」라 부른다. 손실을 견뎌야 하는 자산군에 원금 회수와 상환 규율을 앞세우는, 은행 여신의 습성을 그대로 들여왔다는 진단이다. 융자성 자금이 큰 비중을 차지하는 정부 스타트업 예산은, 겉으로 드러난 숫자와 무관하게 구조상 손실을 감내하는 투자자보다 대출 기관에 가깝게 움직인다.

AI 특허 1위에 팁스 10.4배, 규모는 키웠다

전환점은 2017년이었다. 창업자 일가에 연대 책임을 지우던 연대보증제가 폐지되면서, 창업이 비로소 하나의 진로로 자리 잡았다. 국내외 자본은 그렇게 늘어난 창업 인재를 기꺼이 뒷받침했다. 자금 투입은 해가 갈수록 늘었고, 프리시드·시드·성장 단계마다 정부 프로그램이 겹겹이 쌓였다. 창업사관학교와 팁스(TIPS), 그 아래 프리팁스·포스트팁스·딥테크팁스, 그리고 크고 작은 지역·업종별 지원책이 층층이 더해졌다.

이 가운데 상당수는 저마다 성과를 냈다. 정부가 가장 자주 인용하는 성공 사례인 팁스는 10년간 투입 예산의 10.4배에 이르는 후속 투자를 끌어냈다. 팁스를 거친 약 3,200개 기업의 기업가치를 합치면 88조 원에 이르고, 운영 기관도 2013년 5곳에서 100곳 넘게 늘었다. 유니콘이 나오는 속도는 2000년대와 2010년대 대부분 뜸하다가 2021~2022년 이커머스·뷰티·게임을 중심으로 크게 뛰었다. 같은 기간 한국 스타트업의 연간 수출은 2017년보다 9배로 늘었고, 화장품·전자·반도체 장비의 해외 판매가 그 성장을 이끌었다. 10년 전만 해도 해외 매출이 거의 없던 업종치고는 그 자체로 실질적인 수출 성과다.

국제 비교의 잣대로 보면, 이 흐름은 한국의 위상까지 바꿔 놓았다. Startup Genome의 『글로벌 스타트업 생태계 보고서』는 서울에 펀딩과 인재·지식 두 항목에서 높은 점수를 주는데, 둘 다 공공 자본과 공교육 인프라에 곧바로 좌우되는 지표다. 실제로 한국은 2023년 기준 인구당 AI 특허 등록 건수에서 세계 1위, 지난 10년 증가 속도로도 세계 8위이며, 기술 인재의 밀도 또한 국제적으로 보기 드문 수준이다. AI·바이오·반도체·우주·양자 5개 분야를 아우른 25개 주요국 기술 경쟁력 종합 순위에서도 한국은 5개 가운데 4개 분야에서 상위 10위 안에 든다. 웬만한 중견 경제권은 넘보기 어려운 폭이다.

‘벤처 4대 강국’ 내걸었지만, 새 유니콘은 뜸해졌다

정부는 그 야심을 감추지 않는다. 중기부는 2025년 12월 「벤처 4대 강국 도약 종합대책」을 내놓으며, 2030년까지 유니콘·데카콘 50곳과 AI·딥테크 스타트업 1만 곳을 목표로 제시했다. 규모로 세계 벤처를 이끄는 나라의 대열에 서겠다는 선언이다. 다만 지금의 순위는 그 야심보다 겸손하다. StartupBlink의 국가 비교만 봐도 한국은 일본·덴마크·인도와 같은 구간에 놓여, 선두 4개국과는 아직 거리가 있다. 순위가 더디게 움직이는 데다, 그 바탕이 된 투자마저 제도가 무르익는 데 걸리는 시간으로 보면 이제 겨우 첫발을 뗀 셈이다. 규모와 위상은 별개의 문제이며, 위상에는 규모를 넘어서는 다른 투자가 따로 필요하다.

5년간 100억 달러를 쓰고, 모태펀드를 10조 원 규모의 기관으로 키우고, 징벌적 연대보증제를 폐지해 창업 인재의 응답을 이끌어 내고, 세계적 경쟁력을 갖춘 기업군까지 배출한 정부라면, 실질적인 야심 뒤에 실질적인 자본을 댄 셈이다. 한국 스타트업을 두고 어떤 평가가 엇갈리든, 자금이 부족했던 적은 없다.

다만 양적 규모와 구조적 완성도는 서로 다른 성취다. 업계의 평가자들도 그 간극을 알아차리기 시작했다. 보고서 자체의 유니콘 데이터를 설립 연도로 나눠 보면, 한국 유니콘의 대다수는 정부의 스타트업 지원 체계가 본궤도에 오르기 전에 이미 창업한 기업이다. 자금이 먼저 풀렸고, 업계의 반응은 거기에 느슨하게만 맞물린 채 뒤따랐다. 새 유니콘이 생겨나는 속도는 공공 자금이 가장 집중적으로 투입된 시기에 오르다가 이후 다시 꺾였다. 그 흐름을 되돌리라고 넣은 자금이 계속 흘러드는 동안에도 하락은 멈추지 않았고, 이 대목을 다름 아닌 보고서 스스로가 적어 두었다.

100억 달러는 규모를 만들어 냈다. 관건은 그 규모를 무엇이 떠받치고 있느냐다.

수출 9배 늘어도 매출 비중 1.3%, 구조는 못 세웠다

돈은 하루아침에도 몸집을 키운다. 제도는 그 속도를 따라가지 못한다. 실제로 정부는 한 회계연도 안에 스타트업 예산의 자릿수를 하나 늘릴 수 있지만, 정성적 기준으로 GP를 평가하는 한국 LP의 수는 그만큼 빨리 늘리지 못한다. 리드 투자자 한 곳에 동의권을 모으는 주주간계약의 비중을 끌어올리는 일도, 8년이 기본인 펀드 만기를 10년 이상으로 늘리는 일도 마찬가지다. 이 시리즈의 앞선 두 편이 밝힌 구조적 문제가 여기서 다시 고개를 드는 까닭은, 그것이 왜 세 번째 100억 달러도 첫 번째와 똑같이 움직이는지를 설명하기 때문이다.

박 대표가 최근 인터뷰에서 다듬어 온 비유가 이 간극을 한마디로 압축한다. 자본이 비라면, 그 비를 받아 내는 토양은 자본을 둘러싼 제도적 인프라다. 그의 진단으로는, 지난 5년의 한국 유니콘도 해외 동종 기업 못지않은 규모로 자본을 끌어모았다. 문제는 그다음이다. 상당수가 초기의 성장세를 이어 가지 못하는데, 글로벌 무대의 회수와 조달, 거버넌스 검증을 견딜 운영 관행을 그들이 뿌리내린 토양이 끝내 길러 주지 못한 탓이다. 메마른 땅에 내린 비와 같다. 한때 풀을 무성하게 틔우지만, 그 성장은 곧 정체로 접어든다.

이 정체의 증거는 한국의 자금 규모를 자랑하는 바로 그 보고서 안에 있다. 앞서 본 9배 수출 성장은 약 24억 달러 규모로, 업계 전체의 연간 매출 1,854억 달러에 견주면 1.3%에 불과하다. 9배라는 숫자는 그 자체로는 인상적이지만, 어디서 출발했고 어떤 시장을 열려 했는지를 함께 놓고 보면 이야기가 달라진다. 5년이 지나고 100억 달러를 쏟은 지금도, 이 지표만 놓고 보면 한국 스타트업은 여전히 사실상 내수 시장에 갇혀 있다. Startup Genome의 점수도 다른 각도에서 같은 결론을 뒷받침한다. 한국은 펀딩과 인재·지식에서는 강하지만, 해외 진출과 외국 창업자·투자자의 접근성에서는 평균을 밑돈다. 보고서는 이런 모습을 일본과 같은 유형으로 본다. 투자자 쪽에서도 같은 양상이 되풀이된다. 한국벤처캐피탈협회(KVCA) 집계를 보면, 신규 펀드 조성에 참여하는 해외·기타 기관 투자자의 비중은 해마다 한 자릿수 초반에 머물고, 모태펀드·연기금·국내 기업 자금이 이를 압도한다.

국내 모태펀드 구조는 이 불균형을 바로잡기는커녕 되레 굳힌다. 정부·정책금융기관 자금은 신규 벤처펀드 LP 출자의 22~38%를 차지하는데, 보고서 스스로 구조적 의존이라 부를 만큼 높은 수준이다. 그런데 이 출자 안에서도, 가장 이른 단계의 위험을 감당해야 할 액셀러레이터 서브펀드의 몫은 2020년부터 2024년까지 6~9%라는 좁은 띠에 갇혀 있었다. 업계가 실제로 보이는 초기·중기·후기 투자 비율이 대략 1대 2대 3인 점을 감안하면, 보수적으로 봐도 어긋난 배분이다. 정부·정책금융기관 자금은 이렇게 한국 벤처에서 가장 큰 단일 LP 집단이면서도, 정작 그 배분 규율은 업계가 요구하는 위험 수준에 좀처럼 맞아떨어지지 않는다.

돈을 더 붓는다고 이 어긋남이 사라지지는 않는다. 8년 만기 펀드는 출자액이 두 배로 늘어도 10년 만기 펀드가 되지 않고, 후속 라운드 실적을 좇는 평가 방식은 자금이 더 몰린다고 정성적 기준으로 바뀌지 않는다. 개별 동의권자가 스무 명인 캡테이블 역시 라운드가 커진다고 신디케이트 구조로 저절로 합쳐지지 않는다. 세 가지는 모두 한국의 펀드 문서와 LP 위임 조건, 규제 기대치에 이미 새겨진 설계상의 선택이다. 돈이 아무리 더 들어와도 이 선택들은 그대로 남는다.

한국 자본이 놓이는 구조는 5년 전, 100억 달러 전과 조금도 다르지 않다. 나머지 논점은 모두 그 사실에서 갈린다.

융단폭격과 토마호크, 2년마다 바뀐 정책 구호

박 대표가 2022년 7월 지면 인터뷰에서 던진 가장 날카로운 제도 비판은 한국 스타트업 정책의 조준 방식을 정면으로 겨눈다. 규모가 아니라 방식이 문제라는 것이다. 그의 진단에서 한국은 스타트업 정책과 중소기업 정책을 뒤섞는다. 부처도 같고 예산 항목도 겹치며, 넓게 흩뿌리려는 본능도 같다. 중소기업 정책은 원래 넓게 흩뿌려야 한다. 대다수 중소기업은 끝까지 중소기업으로 남기 마련이고, 그 정책의 목표는 특정 기업의 대박보다 저변 전체의 생존과 안정에 있기 때문이다. 반면 스타트업 정책은 정반대여야 한다는 게 박 대표의 시각이다. 세계적 성과를 낼 소수를 정밀하게 가려내고, 일단 확인되면 그 소수에 자원을 몰아주는 방식이다. 한국은 지난 5년간 정밀 타격이 필요한 문제에 융단폭격으로 맞서 왔다.

뜻밖의 자료도 이 진단을 뒷받침한다. 중기부가 스스로 쌓아 온 다년간의 정책 기록이다. 2016년부터 2025년까지를 훑어보면, 중기부 정책의 핵심 키워드는 대략 2년마다 바뀌었다. 글로벌 진출에서 일자리 창출로, 다시 디지털 전환으로, 스타트업 중심 경제로, 그리고 또다시 글로벌 진출로 돌아왔다. 보고서가 비교 대상으로 드는 유럽과 싱가포르, 대만은 7년에서 10년 단위로 계획을 짠다. 2년 주기는 넓은 지원 대상을 소화하기에는 충분하다. 그러나 정밀 타격에 필요한 제도적 인내심을 기르기에는 너무 짧다. 어느 기업에 자원을 몰아줄지 가려내고, 한두 번의 부진에도 그 판단을 끝까지 밀고 가는 끈기 말이다.

이 글이 앞서 성과로 꼽은 팁스도 가까이서 보면 같은 흐름에 휩쓸려 있다. 팁스의 10.4배 후속 투자 실적을 기록한 바로 그 보고서는, 조용한 부작용도 함께 적는다. 팁스 수료 자체가 하나의 인증서가 됐다는 것이다. 다시 말해 성장 단계로 가는 발판이라기보다, 후속 투자자에게 신뢰를 증명하려 좇는 자격증이 됐다는 뜻이다. 보고서의 서술을 그대로 빌리면, 팁스의 외형과 대상 폭을 넓히는 데 힘을 쏟다 보니 성장 단계 지원을 더 깊이 다지거나, 먼저 성공한 기업이 다음 세대를 되받쳐 주는 재투자 선순환을 만드는 일에는 손이 덜 갔다. 그러다 보니 잘 설계된 제도라도 넓게 흩뿌리는 지원 체계 안에서는, 시간이 갈수록 그 논리에 물들기 마련이다.

재도전 펀드는 한국의 현재 지원 체계에 실재하는 제도인데, 박 대표는 여기에 가장 단호한 평가를 내린다. 그는 이를 「여의도식 사고」, 곧 금융가의 발상이라 부른다. 스타트업의 실패를 그저 메워야 할 재무적 손실로만 본다는 것이다. 하지만 실패한 기업도 무언가를 남긴다. 함께 일한 팀, 쌓인 기술 역량, 특허 같은 지식재산권이다. 실리콘밸리의 벤처 문화는 바로 이것들을 곧바로 재활용한다. 실패한 창업자와 초기 구성원이 다음 세대의 창업자가 되고, 그 밑천은 앞서 성공한 이들이 배우고 번 것을 되돌려주는 데서 나온다. 한국에는 이런 순환이 거의 없다. 재도전 펀드는 돈으로 실패의 손실만 메워 줄 뿐, 그 실패가 남긴 팀과 기술은 살려 두지 못한다.

한국 정부가 내일 당장 스타트업 예산을 두 배로 늘린다 해도, 그 돈을 지금처럼 넓게 흩뿌린다면 폭만 두 배가 될 뿐 깊이는 그대로다. 방식이 곧 제약이다.

싱가포르 파빌리온은 해냈고, 한국 모태펀드는 못 했다

가장 선명한 비교 대상은 싱가포르, 그중에서도 박 대표가 되풀이해 드는 파빌리온캐피털이다. 테마섹(Temasek) 계열의 글로벌 투자 플랫폼인 이 회사는 한국 모태펀드와 큰 틀에서 견줄 규모를 집행하면서도, 세계가 인정하는 기관 투자자로 자리 잡았다. 반면 한국 모태펀드는 같은 규모를 비슷한 기간에 운용하고도 그런 위상은 얻지 못했다. 두 나라는 같은 아시아권에 있고, 벤처 역량의 상당 부분을 정부 연계 기관을 통해 키웠으며, 민간의 자본 축적은 부차적이었다. 출발 조건이 이만큼 비슷했으니, 지리나 역사만으로 두 기관이 걸어간 서로 다른 길을 설명하기는 어렵다.

그렇다면 격차는 어디서 갈렸을까. 규모가 같았던 만큼, 남는 변수는 그 자본을 누가 어떤 조건과 평가 기준으로 집행했느냐다. 갈림길은 바로 거기 있었다.

박 대표는 이 비교를 기관 차원으로 한 단계 더 끌고 간다. 테마섹은 한국 국민연금공단(국민연금)보다 자본 기반이 작다. 그런데도 국제적 인지도나 공동 투자 접근성으로 보면 영향력은 오히려 훨씬 크다. 박 대표가 꼽는 열쇠는 사람이다. 테마섹은 국제적으로 훈련된 투자 인력을 뽑고, 그에 맞는 평가 체계를 세웠다. 그러니 기관의 역량도 그 사람들을 따라 커졌다.

이런 인력 전략은 연기금이든 모태펀드든, 지금 운용하는 자본 규모와 상관없이 마음만 먹으면 언제든 따라 할 수 있다. 다만 한국 정책금융기관 가운데 이를 제대로 된 규모로 실행에 옮긴 곳은 드물다.

박 대표는 같은 문제를 경기순응성이라는 각도에서도 본다. 그가 보기에 한국 금융 시스템은 경기 흐름을 그대로 타고 움직인다. 공개 시장이 위축되면 벤처 자본도 같은 리듬으로 위축된다. 같은 기관, 같은 위험 성향이 양쪽을 다 관장하기 때문이다. 건강한 시스템에는 이 경기순응성을 상쇄할 자금이 필요하다. 공개 시장 심리가 얼어붙는 바로 그 순간에도 벤처 투자 여력이 함께 증발하지 않도록 붙잡아 줄 자금이다. 한국 모태펀드는 한국만의 기준으로, 한국 인력만으로 평가받아 왔고, 그 상쇄 기능을 만들 구조적 원천을 아직 갖추지 못했다.

싱가포르는 글로벌 LP와 공동 투자자, 포트폴리오 기업이 저마다의 기준과 방식으로 인정하는 기관을 세웠다. 그것도 한국 모태펀드가 이미 맞먹는 자본 규모에서 이뤄낸 일이다.

이 시리즈 세 편은 같은 골격을 서로 다른 각도에서 찍은 엑스레이 세 장이다. 229억 달러의 만기 절벽, 넉 달이 걸린 계약 협상, 그리고 구조를 세우지 못한 채 규모만 쌓은 100억 달러다. 그 구조를 그대로 둔 채 자본만 키운다고 결과가 달라지지는 않는다. 격차는 자본이 따르는 기준이 바뀔 때 좁혀진다.

Sources:

• 박희덕 대표 지면 인터뷰: 공감신문(GoKorea), 2022년 7월. 이후 2024~2025년 영상 인터뷰 6회.

• Global Entrepreneurs Association · The Garrison, 『South Korean Startup Ecosystem Report』(2025년 12월 30일), 별도 보관.

• 한국벤처투자(KVIC), 한국 모태펀드: https://www.kvic.or.kr/about-business

• 대한민국 정책브리핑, 「2026년 모태펀드 1차 정례배정」: https://www.korea.kr/news/policyNewsView.do?newsId=148958492

• 중소벤처기업부, 「벤처 4대 강국 도약 종합대책」, 2025년 12월 18일: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156736553

고지: 본 글은 정보 제공만을 목적으로 한다. 투자 자문에 해당하지 않는다. 모든 의사결정은 독자적 조사와 자격을 갖춘 자문가의 상담을 기반으로 이뤄져야 한다.

세인트 클레어 소개: 세인트 클레어(Saint Clair)는 유럽과 아시아를 잇는 국경 간 투자 회사다. 자본이 국경을 넘나드는 인프라까지 함께 구축하는 기관 투자자다. 2016년 이래로.

Saint Clair Global · 마켓 인텔리전스 | 2026년 7월

핵심 요약

2026년 상반기 한국 화장품 수출액은 70억 달러로 집계됐다. 전년 동기보다 27.3% 늘어난 역대 상반기 최대 실적이다. 한국은 이제 EU가 수입하는 화장품의 최대 공급국으로, 그 점유율은 21.7%에 달한다. 그런데 7월부터 8월 12일 사이, 유럽에서는 세 갈래 규제 변화가 한꺼번에 겹친다. 성분 사전 개정, 포장재 규정, 그리고 CMR 물질 퇴출 가속이다. 이에 따라 유럽 시장의 관건도 달라졌다. 수요는 이미 충분히 입증됐고, 이제는 규제 대응 능력이 브랜드의 성패를 가른다. 특히 유럽 현지의 규제 신고 업무를 외부 업체에 맡겨온 브랜드가 가장 큰 부담을 안게 됐다.

상반기 70억 달러, 성장의 축은 스킨케어

식품의약품안전처가 7월 2일 내놓은 상반기 집계를 보면, 시장의 양상이 뚜렷하게 달라졌다. 수출액은 70억 달러로 전년 동기보다 27.3% 늘었고, 분기별 증가세도 꺾이기는커녕 더 가팔라졌다. 실제로 1분기 31억 달러에서 2분기 39억 달러로 올라섰고, 6월 한 달에만 13억 4000만 달러를 수출해 여덟 달 연속 증가를 이어갔다.

그런데 총액보다 눈여겨볼 대목은 그 구성이다. 기초화장품이 상반기 수출의 54억 8000만 달러를 차지하며 25.0% 늘어난 반면, 색조화장품은 4.2% 줄어든 7억 2000만 달러, 목욕·세정용품은 20.6% 줄어든 3억 4000만 달러에 그쳤다.

이로써 2년 가까이 이어진 논쟁도 정리됐다. 시장은 오랫동안 K뷰티 수출 호황을 값싸고 교체 주기가 빠른 신기성 제품 덕으로, 즉 재구매 기반이 취약하다고 평가해왔다. 그러나 상반기 데이터는 정반대를 가리킨다. 성장을 이끈 것은 매일 바르는 스킨케어로, 재구매가 가장 꾸준하고 가격대가 가장 탄탄한 품목이다. 반면 일회성 제품은 오히려 줄고 있다.

지역별로 보면, 가속이 가장 가파른 곳은 유럽이다. 상반기 대영국 수출이 150.6%, 네덜란드가 220.4%, 폴란드가 72.8% 늘었다. 1년 전만 해도 유럽 국가는 한 곳뿐이었지만, 이제는 한국의 10대 수출 대상국에 세 곳이나 이름을 올렸다. 특히 6월 한 달 대EU 수출은 78.9%나 늘었다.

1위의 그림자, 쏟아지는 서류와 단속

코트라 브뤼셀무역관에 따르면, 2025년 EU 화장품 수입(HS 3304)에서 한국은 21.7%를 점유했다. 수입액을 4억 달러가량 늘리며 19.7%에 그친 미국을 앞질렀다. 아시아산 수입이 전반적으로 늘어나는 반면, 미국·영국·스위스에서 들어오는 물량은 줄고 있다.

시장 1위가 되면 규제 당국과의 관계도 달라진다. 유럽에 진출한 한국 브랜드들은 틈새 수입품에 머물던 시절에는 없던 규모의 모니터링 공문과 증빙 자료 요구를 받고 있다. 점유율이 오른 만큼, 유럽 경쟁사와 소비자의 고발도 함께 늘었다. 회원국 세관은 EU 역내 책임자(Responsible Person, 이하 RP)의 위임장 공증에 더해 안전성 평가사 이력서 공증까지 요구하기 시작했고, 한 매장은 사용기한을 유럽 방식이 아닌 한국 방식으로 표기했다는 이유로 벌금을 물었다.

성장과 함께 회색지대 물량도 늘었다. 인증도 없고 지정된 책임자도 없이 아마존 등 플랫폼에서 유통되는 제품은 이미 브뤼셀의 실질적인 단속 대상이다. 그리고 여기서 발생하는 평판 손실은 업계 전체가 함께 떠안는다. 무인증 화물이 유럽 세관에 한 건이라도 적발되면 그 여파가 품목 전체로 번지고, 그 품목의 최대 공급국이 바로 지금의 한국이기 때문이다.

성분·포장·유해물질, 세 규제가 한여름에 겹친다

우선 7월부터 개정 INCI 성분 사전(규정 2025/1175)이 의무 적용되고, 향료 알레르겐 표기 대상도 82종으로 늘어난다. 이미 유럽 시장에서 제품을 판매 중인 브랜드라면 포장을 전면 수정하거나 원료를 대체해야 하는데, 인쇄에 걸리는 시간까지 감안하면 일정이 빠듯하다.

다음으로 8월 12일부터는 EU 포장재·포장폐기물 규정(PPWR)이 본격 적용된다. PFAS와 중금속 농도가 제한되고 추적성 요건이 더해지며, 포장재 협력사로부터 적합성 선언서(DoC)를 받아두어야 한다. 여기에 회원국별 생산자책임재활용(EPR) 등록까지 따라붙는다. 진출하는 나라가 늘수록 행정 부담도 그만큼 커진다.

끝으로 유럽의회와 이사회는 CMR 물질(발암성·변이원성·생식독성물질) 퇴출 유예기간을 6~12개월 단축하는 데 잠정 합의했다. 이에 따라 종전 기준으로 짜둔 재처방 일정도 그만큼 압축된다.

세 가지 모두 서류를 갖추는 데 드는 비용이다. 이 비용은 매출 지표에 잡히지 않고 원가와 법인 구조에만 얹히기 때문에, 앞으로 두 분기 동안은 수출 통계에도 흔적을 남기지 않는다.

승부는 법인 구조에서 갈린다

K뷰티의 유럽 진출을 뒷받침한 유통 구조는 속도에 맞춰 설계됐다. 신생 한국 브랜드는 국경 간 판매를 대행하는 크로스보더 셀러와 중개업체를 거쳐 유럽 소비자에게 제품을 판매하고, 매장 운영과 통관은 물론 CPNP 등록까지 이들이 맡는 경우가 많다. 브랜드는 제품을 공급하고, 규제 서류는 이들이 대신 제출한다. 그러나 그 서류가 곧 법적 책임으로 바뀌는 순간, 이 편리한 방식은 그대로 위험이 된다.

이번 여름 시행되는 규제에서 RP는 규제 당국이 직접 상대하는 주체이자, 세관이 촉박한 시한을 걸어 공증 서류를 요구하는 대상이다. 따라서 물류 중개업체의 책임자를 빌려 쓴 브랜드는 규제상 지위까지 함께 빌린 셈이다. 재처방 일정도, 포장 사양도, EPR 등록도 스스로 결정하지 못한다. 그리고 그 대가는 계획 단계에서는 보이지 않다가 단속 현장에서 비로소 드러난다.

문제는 이 지점이 비어 있다는 점이다. 중개업체와 유럽 현지법인 사이, 곧 한국식 기업 지배구조와 유럽의 소매 규제를 동시에 소화하는 중견 브랜드 관리 사업자가 거의 없다. 그동안 한국 본사는 서울에서 지표만 보며 유럽 사업을 관리하다가, 수요가 몰리면 제3자 물류센터로 물량을 넘기며 대응해왔다. 그러나 8월부터 부과되는 규제 의무는 이렇게 넘길 수 없다.

이제 실사는 서류를 묻는다

거시 지표는 봄에 예상했던 것보다 오히려 견고하다. 상반기 70억 달러, EU 수입 점유율 1위, 고객 유지율이 가장 높은 품목에 집중된 성장, 여기에 지금 속도가 유지되면 연간 137억~140억 달러에 이른다는 전망까지 더해진다. 수요 자체는 이미 입증됐다.

달라진 것은 위험의 성격이다. 1년 전만 해도 한국 뷰티 브랜드의 유럽 사업을 볼 때는 마케팅 효율과 판매 채널 구성을 물었다. 그러나 지금 확인해야 할 항목은 그보다 좁고 구체적이다. CPNP 등록은 누가 보유하고 있는지, 책임자가 브랜드 자신의 법인인지 아니면 외부에서 빌려 온 것인지를 먼저 따져야 한다. 포장재 협력사가 8월 12일 전에 적합성 선언서를 내줄 수 있는지, 회원국별 EPR 등록이 실제로 몇 건이나 확보돼 있는지도 같은 무게로 확인해야 한다.

2026년 하반기, K뷰티 업계는 유럽에서 둘로 갈린다. 유럽을 수출처로만 대한 브랜드는 이 비용을 충격으로 맞고, 법인을 세운 브랜드는 사업 비용으로 흡수한다. 그 차이는 수출 통계에 잡히지 않는다. 지금, 실사 현장에서 드러난다.

Sources:

• 식품의약품안전처, 2026년 상반기 화장품 수출 실적. 서울경제(Seoul Economic Daily), 「K-Beauty Exports Hit Record $7 Billion in First Half, US Share Tops 20%」(2026년 7월 2일): https://en.sedaily.com/markets/2026/07/02/k-beauty-exports-hit-record-7-billion-in-first-half-us

• 산업통상자원부, 2026년 6월 수출입 동향. 씨앤씨뉴스(CNC News), 「6월 화장품 수출 13.4억달러(+43%), 상반기 70억달러... 미국·EU·중국 등 큰 폭 성장」(2026년 7월 1일): https://www.cncnews.co.kr/news/article.html?no=11059

• 코트라 브뤼셀무역관, EU 화장품 수입 점유율 및 2026년 하반기 규제 진단. 씨앤씨뉴스(CNC News), 「유럽 화장품시장, 현지 견제(소비자·기업)+규제(각국 식약처·세관) ‘이중 암초’」(2026년 7월 11일): https://www.cncnews.co.kr/news/article.html?no=11077

• 코리아타임스(The Korea Times), 「Korea becomes world’s No. 2 cosmetics exporter as trade surplus tops $10 bil.」(2026년 5월 22일): https://www.koreatimes.co.kr/business/20260522/korea-becomes-worlds-no-2-cosmetics-exporter-as-trade-surplus-tops-10-bil

Disclaimer: 본 글은 정보 제공만을 목적으로 한다. 투자 또는 사업 자문에 해당하지 않는다. 모든 의사결정은 독자적 조사와 자격을 갖춘 자문가의 상담을 기반으로 이뤄져야 한다.

About Saint Clair: 세인트 클레어는 유럽과 아시아를 잇는 국경 간 투자 회사다. 자본이 국경을 넘나드는 인프라까지 함께 구축하는 기관 투자자다. Saint Clair Global은 기업을 국제 기관 표준에 맞춰 구축한다. 사업이 구하는 자본 앞에서 성공적이고 매력적인 투자 기회로 읽히게 만드는, 그 실천의 기업 측면이다. 2016년 이래로.

Saint Clair Pte. Ltd. · 싱가포르 · saintclair.global

Saint Clair · Market Intelligence | July 2026

The Volume Singapore Celebrates

Singapore’s single family offices passed 2,000 by the end of 2024, up from 1,400 a year earlier. Roughly 600 new offices formed in 2024 alone, twice the number that formed in 2023, according to figures the Monetary Authority of Singapore’s (MAS) Deputy Chairman Chee Hong Tat disclosed in a January 2025 speech. Assets under management across the city-state’s fund management industry reached a record S$6.07 trillion in 2024, up 12% on the year before, with net inflows rising roughly 50% year-on-year. Singapore’s billionaire population grew to 55 in 2025 from 47 the year before, combined wealth estimated at close to S$335 billion. These are the numbers policymakers cite and the numbers the skyline increasingly reflects: more towers, more private-banking floors, more advisors quoting basis points to a widening client base.

Volume Is Not Composition

Volume measures inflow, not character. A city can host record numbers of family offices and record trillions in assets while the underlying composition of that capital shifts in ways the headline figures do not capture. Two kinds of capital can produce identical totals and identical financial-centre statistics. They do not behave the same way when conditions change, and they do not leave the same institutional footprint behind. The question Singapore’s principals now have reason to hold is not how much capital is arriving, but what kind.

Two Kinds of Capital

Parking capital is regulatory-arbitrage motivated. It seeks optionality over yield and discretion over engagement, and it carries a light operational footprint: perhaps two investment professionals administering assets that could relocate within a fiscal quarter. Operating capital is mandate-driven. It funds real businesses, staffs real offices, and answers to structures built around decades rather than quarters, which is what makes it slow to unwind.

These are loose tendencies that shift as circumstances change, and Singapore draws both from every geography represented in its private-banking client base. A family office can arrive holding parking capital and become an operating-capital vehicle as the family builds a genuine Singapore-based enterprise around it. The distinction turns on behaviour after arrival, and on what it would take for the capital to leave.

What the Composition Data Shows

Three data points test that distinction, each in a different way.

The first sits in the gap between two formation routes carrying the same name. More than 2,000 single family offices now hold tax incentives under Singapore’s Section 13O and 13U schemes, a route that requires as few as two Singapore-based investment professionals and no residency commitment at all. The Global Investor Programme, Singapore’s route to permanent residency through committed capital, sets a materially higher bar: a minimum S$10 million direct investment in a Singapore business, a S$25 million commitment to an approved fund, or a S$200 million single family office with at least S$50 million transferred into the country. Thresholds were raised again in 2023, and reported uptake has not shifted materially since. The tax-incentive route alone produced roughly 600 new offices in 2024. The imbalance sits on the parking side of the ledger: the route that asks least of the capital using it is also the route producing, by a wide margin, the most volume.

The second signal is regional rather than local, and it works by elimination. Set Singapore’s growth, noted above, against its neighbours’: Hong Kong’s assets under management grew 13% in 2024, on an 81% jump in fund inflows, per the Securities and Futures Commission’s industry survey, and Hong Kong’s single family offices already outnumber Singapore’s in absolute terms, more than 3,380 against Singapore’s 2,000-plus at the end of 2025, per Invest Hong Kong’s own count. The Dubai International Financial Centre (DIFC) added wealth and asset management firms at a 16% clip, and its 120 largest resident families now manage more than US$1.2 trillion between them. Singapore is one of at least three centres absorbing comparable volumes in the same window. On the parking-versus-operating question, this signal stays silent; what it rules out is the assumption that Singapore’s growth rate, or its formation count, is itself evidence of anything distinctive. The composition question has to be answered elsewhere, because the volume comparison cannot answer it.

The third is the one public data cannot yet answer with precision. MAS does not disclose the beneficial-owner size distribution across Singapore’s family-office population, nor the share of assets under management that entered through tax-incentive vehicles rather than institutional mandates. What is visible is the regulatory response to that opacity: from 15 June 2026, MAS replaced its case-by-case exemption regime for single family offices with a mandatory registration framework, requiring formal notification, annual returns, and a Singapore-resident contact person for every office, old and new. Regulators do not rebuild the architecture of a scheme unless the volume moving through it has already outpaced the assumptions the scheme was designed around — and a registration mandate reaching every office, old and new, is itself a signal about which kind of capital the regulator now believes it is hosting at scale.

The Regulatory Response

Singapore’s regulators have not been passive. The 2023 discovery of a S$3 billion money-laundering operation produced the country’s largest asset-recovery action to date, after all ten individuals charged pleaded guilty: roughly S$2.8 billion in cash, property, and other assets surrendered or forfeited by the convicted and by fugitives who agreed to give up their claims. In July 2025, MAS fined nine financial institutions a combined S$27.45 million for anti-money-laundering lapses connected to the case. In November 2025, MAS revoked tax incentives from two family offices linked to Chen Zhi, a Cambodian businessman indicted in the United States for directing forced-labour scam operations, after authorities across five jurisdictions seized more than US$15 billion in assets tied to his network. Singapore’s own figures put the scale in context: family offices connected to money-laundering convictions represent under 1% of the sector, and MAS rejected roughly 3% of the 1,300 family-office applications it reviewed over the prior three years.

What the record does not settle is whether enforcement is a standing capacity or a sequence of responses to whichever case became public first. Each action followed a case rather than pre-empting one, including the framework that took effect in June 2026.

What a Parking-Capital Centre Risks

A financial centre that skews toward parking capital carries three specific exposures. Operational anchoring is the first: capital that arrived because a spreadsheet favoured Singapore over Zurich or Abu Dhabi can leave for the same reason, taking the private-banking mandates, the professional-services fees, and the office leases built around it. Regulatory-cost burden is the second: the compliance architecture needed to screen thousands of thinly staffed structures is expensive to run, and the political case for that expense weakens if the flows it screens are seen as extractive rather than productive. Reputational tail risk is the third and least forgiving: a single case, however small a share of the sector it represents in the statistics, can recalibrate how counterparties, correspondent banks, and regulators elsewhere treat the entire centre for years afterward.

All three are structural features of hosting capital at these volumes, in Singapore or in any centre absorbing them, and each applies with more or less force depending on the composition of what keeps arriving.

The Principal’s Asymmetry

The asymmetry lands differently depending on which side of it a reader sits. A principal who runs a real business in Singapore, employs local staff, and built a structure meant to outlast a single fiscal cycle is operating capital by definition, and the tightening under way mostly does not touch them: the new registration requirements formalise what a genuine operating presence already does as a matter of course. A principal whose Singapore structure exists chiefly to hold assets at a favourable tax rate sits closer to the other end of the spectrum, and every recalibration, from the June 2026 framework to whatever follows the next high-profile case, adds friction they did not previously carry. Singapore has not chosen what kind of financial centre it becomes. The capital arriving now, formation by formation, is one of the things doing the choosing.

Sources:

• Asia Asset Management, “In Singapore, new single-family offices double in 2024, lifting total to 2,000” (January 2025, citing MAS Deputy Chairman Chee Hong Tat): https://www.asiaasset.com/family-offices/in-singapore-new-single-family-offices-double-in-2024-lifting-total-to-2000/

• The Edge Singapore, “Singapore’s AUM grows 12% to $6.07 tril in 2024; net inflows rebound with 50% y-o-y growth”: https://www.theedgesingapore.com/news/asset-management/singapores-aum-grows-12-607-tril-2024-net-inflows-rebound-50-y-o-y-growth

• Singapore Economic Development Board, Global Investor Programme Factsheet (updated 5 May 2025): https://www.edb.gov.sg/content/dam/edb-en/how-we-help/global-investor-programme/GIP%20Factsheet.pdf

• The Online Citizen, “Singapore grants permanent residency to 450 high-net-worth investors under Global Investor Programme” (27 February 2026): https://theonlinecitizen.com/2026/02/27/singapore-grants-permanent-residency-to-450-high-net-worth-investors-under-global-investor-programme/

• Securities and Futures Commission (Hong Kong), 2024 Survey on Asset and Wealth Management Business: https://apps.sfc.hk/edistributionWeb/gateway/EN/news-and-announcements/news/doc?refNo=25PR112

• Invest Hong Kong, “Hong Kong’s single-family offices total surpasses 3,380, injecting over $10 billion annually into local economy” (10 February 2026, Deloitte study): https://www.investhk.gov.hk/en/news/hong-kongs-single-family-offices-total-surpasses-3-380-injecting-over-10-billion-annually-into-local-economy/

• DIFC, “Record 20th anniversary year results solidify DIFC’s position as region’s number one global financial centre”: https://www.difc.com/whats-on/news/record-20th-anniversary-year-results-solidify-difcs-position

• Wikipedia, “2023 Singapore money laundering case” (compiled record of court proceedings and sentencing): https://en.wikipedia.org/wiki/2023_Singapore_money_laundering_case

• CNBC, “Singapore monetary authority penalizes 9 banks, institutions for 2023 money laundering case” (4 July 2025): https://www.cnbc.com/2025/07/04/singapore-monetary-authority-penalizes-9-banks-institutions-for-2023-money-laundering-case.html

• Asia Asset Management, “Singapore cancels tax breaks for Cambodian tycoon’s family offices” (5 November 2025): https://www.asiaasset.com/post/30230-sfosingapore1105-gte-1106

• Baker McKenzie, “Singapore: Revised Framework for SFOs Comes into Force” (June 2026): https://www.bakermckenzie.com/en/insight/publications/2026/06/singapore-revised-framework-for-sfos-comes-into-force

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Since 2016.

Learn more: saintclair.sg | Contact: contact@saintclair.sg

Saint Clair Global · Market Intelligence | July 2026

Executive Summary

Korean cosmetics exports reached $7.0 billion in the first half of 2026, up 27.3% and the largest half-year on record. Korea is now the single largest source of cosmetics imported into the European Union, holding 21.7% of that import market. Between July and 12 August, three EU regulatory regimes tighten at once: an updated ingredient glossary, a packaging and waste regulation, and an accelerated phase-out of CMR substances. The binding constraint on Korean brands in Europe has moved from demand to documentation, and the brands most exposed are those whose European regulatory filings sit with somebody else.

Where the Growth Actually Sits

The half-year figures the Ministry of Food and Drug Safety published on 2 July describe a market that has changed shape. Exports reached $7.0 billion, 27.3% above the same period last year, and the momentum built rather than faded: $3.1 billion in the first quarter, $3.9 billion in the second. June alone brought $1.34 billion, an eighth consecutive monthly rise.

The composition tells more than the total. Basic skincare accounted for $5.48 billion of the half-year figure, growing 25.0%. Colour cosmetics fell 4.2% to $720 million. Body cleansing products fell 20.6% to $340 million.

That settles a question which has hung over the sector for two years. The received reading of K-Beauty’s export boom places it on cheap, high-turnover novelty formats with little claim on the customer. The half-year data places the growth in the daily skincare routine, the category with the longest customer life and the firmest pricing. The disposable formats are shrinking.

Europe is where the acceleration is sharpest. Exports to the United Kingdom rose 150.6% over the half, the Netherlands 220.4%, Poland 72.8%. Three European markets now sit inside Korea’s top ten destinations, against one a year ago. EU shipments grew 78.9% in June.

First Place, and What Comes With It

KOTRA’s Brussels office reports that Korea took 21.7% of EU cosmetics imports under HS 3304 in 2025, displacing the United States at 19.7% on a rise of roughly $400 million. Imports from Asia are climbing across the board; those from the United States, the United Kingdom and Switzerland are in decline.

Market leadership changes a company’s relationship with its regulator. Korean brands operating in Europe now report a volume of monitoring correspondence and evidentiary requests that simply did not exist when they were a niche import. Complaints from European competitors and consumers have risen alongside the share. Customs authorities in individual member states have begun demanding notarised powers of attorney for Responsible Persons and notarised CVs for safety assessors. One retailer describes being fined for printing an expiry date in Korean convention rather than European.

Growth has pulled grey-market volume along behind it. Product moving through Amazon and comparable platforms without certification, and without a designated Responsible Person, is now a live enforcement concern in Brussels. The reputational exposure there is collective. An uncertified consignment held at a European border damages the category, and the category is now Korean.

Three Regimes, One Summer

From July, the updated INCI ingredient glossary (Regulation 2025/1175) applies as an obligation, and fragrance allergen labelling extends to 82 substances. For a brand with a portfolio already in market, that means either full packaging redesign or ingredient substitution. Print timelines are not short.

From 12 August, the Packaging and Packaging Waste Regulation takes effect. It sets PFAS and heavy-metal limits, imposes traceability requirements, and obliges brands to obtain Declarations of Conformity from their component suppliers. It also requires Extended Producer Responsibility registration country by country, an administrative load that scales with every market a brand sells into.

Behind both, Parliament and Council have provisionally agreed to shorten the phase-out grace period for CMR substances by six to twelve months, compressing reformulation timetables that were set against the old assumption.

Each of these is a documentation cost. They land on a legal structure and a cost base, not on a sales line, which is precisely why the export figures will not show them for another two quarters.

The Entity Question

The distribution model that carried K-Beauty into Europe was built for speed. An emerging Korean brand can reach European consumers through cross-border aggregators and brokers who run the storefront, clear customs and, in many cases, hold the CPNP registration itself. The brand ships. Somebody else files.

That arrangement works beautifully until the filing becomes the liability.

Under the regimes arriving this summer, the Responsible Person is the entity a regulator addresses and the entity a member-state customs office will ask to produce notarised documentation at short notice. A brand that has rented its Responsible Person from a logistics intermediary has rented its regulatory position along with it. It does not control the reformulation timetable, the packaging specification, or the EPR registrations. It will discover the cost of that at the point of enforcement rather than the point of planning.

This is where the corridor is thin. Between the aggregator layer and a full European subsidiary sits a category of operator that barely exists: mid-scale brand management with genuine fluency in both Korean corporate governance and European retail compliance. Korean leadership teams manage European growth from Seoul, through dashboards, routing containers to third-party fulfilment centres and reacting to demand spikes as they appear. The obligations arriving in August do not route.

What Diligence Should Now Ask

The macro case is intact and, if anything, firmer than it looked in the spring. A $7 billion half-year. First place in EU import share. Growth concentrated in the highest-retention product category, and a full-year figure that reaches $13.7 to $14.0 billion if the current run-rate holds. Demand is settled.

What has changed is the shape of the risk. Twelve months ago, assessing a Korean beauty brand’s European position meant asking about marketing efficiency and channel mix. The questions that matter now are narrower and duller: who holds the CPNP registration, whether the Responsible Person is the brand’s own entity or a rented one, whether the packaging suppliers can produce Declarations of Conformity before 12 August, and how many member-state EPR registrations actually exist.

We expect the second half of 2026 to separate the sector visibly. Brands that treated Europe as a shipping destination will absorb the compliance cost as a shock; brands that built the entity will absorb it as a cost of doing business. That distinction is legible in the diligence today. It was never legible in the export data.

Sources:

• Ministry of Food and Drug Safety, H1 2026 cosmetics export figures, via Seoul Economic Daily, “K-Beauty Exports Hit Record $7 Billion in First Half, US Share Tops 20%” (2 July 2026): https://en.sedaily.com/markets/2026/07/02/k-beauty-exports-hit-record-7-billion-in-first-half-us

• Ministry of Trade, Industry and Resources, June 2026 trade figures, via CNC News (씨앤씨뉴스), “6월 화장품 수출 13.4억달러(+43%), 상반기 70억달러... 미국·EU·중국 등 큰 폭 성장” (1 July 2026): https://www.cncnews.co.kr/news/article.html?no=11059

• KOTRA Brussels, EU cosmetics import share and H2 2026 regulatory assessment, via CNC News (씨앤씨뉴스), “유럽 화장품시장, 현지 견제(소비자·기업)+규제(각국 식약처·세관) ‘이중 암초’” (11 July 2026): https://www.cncnews.co.kr/news/article.html?no=11077

• The Korea Times, “Korea becomes world’s No. 2 cosmetics exporter as trade surplus tops $10 bil.” (22 May 2026): https://www.koreatimes.co.kr/business/20260522/korea-becomes-worlds-no-2-cosmetics-exporter-as-trade-surplus-tops-10-bil

Disclaimer: This article is for informational purposes only and does not constitute investment or business advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair is a cross-border investment firm between Europe and Asia: an institutional investor that also builds the infrastructure through which capital crosses borders. Saint Clair Global builds companies to international institutional standards: the company side of the practice, where a business is made to read as a successful, attractive investment opportunity to the capital it seeks. Since 2016.

Saint Clair Pte. Ltd. · Singapore · saintclair.global

Indonesia has built 120 incubators and needs nine million digital workers by 2030; the incubators are underfunded by design, and the founders best positioned to succeed increasingly incorporate in Singapore. A second-generation founder cohort is emerging anyway, small and real. This is the constraint the rest of this series has been describing from four different angles.

Based on the Asian Development Bank’s Indonesia’s Technology Startups: Voices from the Ecosystem, East Ventures’ Digital Competitiveness Index 2025 and Sustainability Report 2025, DailySocial’s DiscoveryShift Indonesia Startup Report 2026, and AC Ventures & Bain & Company’s Indonesia Venture Capital Report 2023. Saint Clair’s analysis follows.

The Scale, and the Depth Underneath

Indonesia has built the scaffolding of a founder pipeline at a scale few Southeast Asian markets can match. Around 120 incubators and accelerators operate across the country, alongside roughly 200 financing institutions that serve startups wholly or in part. Jakarta’s own ambition runs higher still: Indonesia’s Ministry of Communication and Digital Affairs (Komdigi) runs Garuda Spark, a programme targeting the creation of 1,000 tech companies at a USD 100 million scale, against a national need East Ventures’ Digital Competitiveness Index puts at nine million digital workers by 2030. A second-generation founder cohort, drawn from the country’s first wave of tech companies, is just beginning to take shape too.

Set that scale against the conversion rate and the gap between ambition and outcome opens. Only fifteen per cent of Indonesian startups that raise a seed round progress to Series A, per DailySocial’s DiscoveryShift Indonesia Startup Report 2026. Pre-seed and seed rounds make up sixty-seven per cent of all 2025 deals; Series A through C, at just twenty-eight per cent of rounds, absorb eighty per cent of the capital deployed. The Asian Development Bank has a name for the distance between those two figures: the “valley of death,” the high-risk stretch in which a startup must reach scale amid competition, uncertainty, and creative disruption, with no guarantee of reaching the other side. Indonesia is generous at the entry point and unforgiving at the gate. A market this size should not need to filter this hard this early. The filtering says less about the founders than about what built them.

Why the Factory Is Unfunded

The incubator count is real; its quality is the more useful number. Sixty per cent of Indonesia’s roughly 120 programmes sit in Java, leaving founders elsewhere with meaningfully less access. The Asian Development Bank’s ecosystem mapping, drawn from interviews across the incubator and accelerator community, is direct about what that access delivers. Local government incubators, in the report’s own words, “are often managed by government staff who have limited understanding of the business world and how startups are run.” Private incubators face the opposite problem: in the report’s assessment, “the private sector finds it difficult to operate incubators on a profitable basis.”

Behind both sits a funding design that inverts its own logic. Government money, the ADB report finds, “can only be used to directly provide assistance for startups, leaving incubators and accelerators without sufficient funds to cover the costs of their facilities and services.” Indonesia funds the plant’s output and starves the plant. Regulation has tried to correct the geography and failed to: a 2015 rule from the Ministry of Cooperatives and Small and Medium Enterprises requires at least five incubators in every province and one in every district or city, and the ADB’s own researchers found the target unmet, in part because there is no penalty for falling short. A founder who reaches the far end of this pipeline has been underserved twice over: once by an incubator that lacked the sector-specific mentors to prepare them properly, and once by a state that will fund their company but not the institution meant to develop them.

Why Singapore Wins the Best Founders

Where the pipeline does produce a founder with real promise, Indonesia often loses the company at incorporation. Talent scarcity is the proximate cause. A “talent deficit” was frequently cited by the ADB’s informants as an important constraint, and for good reason: unicorns and large established firms compete with early-stage startups for the same small pool of engineers, programmers, and designers, and the unicorns generally win. Indonesia’s own ministry undercuts its own target here: Komdigi projects that fewer than seven million of the nine million digital workers needed by 2030 can realistically be trained in that window. Enrolment in generative AI training courses is climbing fast, up 3.4 times year on year according to Coursera data cited in the Google/Temasek/Bain e-Conomy SEA 2025 report, one of the steeper increases in Southeast Asia. Growth from a small base does not close a two-million-worker shortfall, let alone the nine million the country actually needs.

Founders who can raise capital elsewhere often do, and the money agrees with them. Singapore-headquartered companies absorbed sixty-one per cent of Southeast Asia’s equity funding by value in 2025, against fourteen per cent for Indonesia-headquartered ones: 283 deals worth USD 4.2 billion against Indonesia’s 66 deals worth USD 340 million. DailySocial’s DiscoveryShift report describes Singapore holding structures as “the preferred route for accessing regional capital and managing liquidity,” and rates Indonesia as “a higher-beta allocation relative to Singapore’s role as the region’s legal and financial hub.” The calculation is straightforward: clearer exit routes, faster access to capital, and a legal architecture international limited partners already trust. The departure reflects the incentives Jakarta’s own capital markets and incubator system have built.

The Second Generation, Quietly

The layer that might close Indonesia’s founder-development gap is visible, if still thin. AC Ventures and Bain’s assessment, written into a year of scandal and retrenchment across the sector, singled out the rise of second-generation founders as one of the few structural positives: people who had already worked inside Indonesia’s tech companies, absorbed how they were built, and were turning that experience into ventures of their own. East Ventures counts more than 500 founders across a portfolio of over 300 companies, and twenty-six per cent of that active portfolio has at least one female founder, a figure the firm attributes to a deliberate diversity policy rather than an accident of deal flow.

Five hundred founders inside one firm’s portfolio, set against everything this series has traced about the scale of what Indonesia’s pipeline still needs to produce, is a small, real signal, worth tracking as it grows. DailySocial’s DiscoveryShift report reaches a similar conclusion from the funding side: governance and product-market-fit gaps are, in its words, “more visible among newer founder cohorts,” and the resulting shortfall in early-stage research and founder formation “presents an opportunity to cultivate a more high-quality pipeline of next-generation startups.” The report is diagnosing a funding market. It has, without naming it as such, described a founder-development one.

Capital is already chasing the newer frontier regardless of what today’s pipeline can support. Investment in Indonesia’s AI-driven startups reached USD 542.9 million by 2024, up 141.5 per cent since 2020, a growth rate the founder-development infrastructure underneath it has not matched. Jakarta’s own Garuda Spark ambition is itself an admission of the same shortfall: a government target for the kind of company this pipeline is not yet producing at pace.

The Corridor View

Every article in this series has traced a downstream symptom back to something upstream. The capital that reallocated through Singapore, and the governance failure that went undetected until it collapsed, are each, in part, a founder problem wearing a different label. Better capital cannot compound through founders who were never developed to receive it. Governance cannot be enforced by founders who were never mentored on what enforcement demands. The capital question, in other words, sits downstream of the founder question.

The market has not priced this yet. The visible signals, the funding round, the valuation, the headline company, are what a due-diligence process can measure quickly. Whether a founder came through a programme with genuine sector expertise, or an incubator with the management capacity to prepare them for institutional capital, sits underneath what most term sheets ask about. It shows up in reference calls and founder networks well before it shows up on any term sheet. The investor patient enough to read that slower signal before the rest of the market catches up holds a moat the current pricing does not reflect — and firms that cultivate governance-ready founders, backing them with structural patience, are the ones building it.

Indonesia’s own market researchers have already named the shape of the answer without stepping into the business of providing it. The pipeline needs deliberate cultivation: mentorship with sector expertise, management capacity inside the incubators themselves, and capital patient enough to compound over years rather than a single grant cycle. Whether the capital that eventually closes that gap comes from government programmes, from venture funds, or from advisory relationships willing to stay with founders through the years an incubator’s budget will not cover, the opening exists for whoever occupies it first. Where that opening meets the wider capital corridor between Indonesia and its investors is where this series turns next.

Sources:

• Asian Development Bank, Indonesia’s Technology Startups: Voices from the Ecosystem (Country Report No. 8, June 2023).

• AC Ventures and Bain & Company, Indonesia Venture Capital Report 2023 (November 2023).

• East Ventures, Digital Competitiveness Index 2025.

• East Ventures, Sustainability Report 2025.

• DailySocial / DiscoveryShift, Indonesia Startup Report 2026 (January 2026 release).

• Google, Temasek, Bain & Company, e-Conomy SEA 2025, 10th edition (November 2025).

Disclaimer: This article is for informational purposes only and does not constitute investment advice. Saint Clair has an active engagement with frontier Asian startup ecosystems and may have commercial interests in markets discussed. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair designs and builds cross-border capital infrastructure between Europe and Asia — proposing access where access is scarce, and creating structure where structure is absent. Saint Clair Asia (saintclair.asia) is a frontier investment platform that positions international investors within innovation ecosystems that institutional channels do not reach.

Learn more: saintclair.sg | saintclair.asia | Contact: contact@saintclair.sg

Saint Clair Capital · Ground Truth | July 2026

Executive Summary

When Sequoia Capital entered Coupang’s Series C round, the term sheet took weeks. The consent process took four months. The friction came from a structural source: the Korean cap table held approximately twenty individual shareholders whose veto rights were structured under domestic contract convention. Reconciling that convention with the global syndicated structure Sequoia’s own limited partners required cost four months of dedicated negotiation, separately from diligence, price, and round structuring.

This is the architecture. It operates at five interlocking levels: fund-term convention, contract convention, board-governance convention, regulatory configuration, and capital culture. Each layer is internally coherent inside the Korean system. Each layer carries a translation cost when foreign institutional capital arrives.

Most operators inside the system see the layers as the way the work gets done. The view from inside a coherent operating system rarely registers the system as a system. The Korean policy community has begun, in the past eighteen months, to acknowledge that the layers exist. The diagnosis of the interlock itself, that the five layers form a single coordinated system that systematically taxes cross-border capital flow, remains rare even inside Korea. The veteran practitioner Park Hee-duk has articulated it across eight interviews and lectures between 2022 and 2025, and by September 2025 had presented it directly to National Assembly members. Korean policy has reached the layers. The mechanism waits.

This article describes what foreign capital meets. It proposes no solution. The reader infers the category.

Four Months

Sequoia Capital’s Coupang Series C closed in 2014 after four months of consent negotiation. The term sheet had taken weeks. The remaining four months were architecture.

The valuation at entry was approximately three hundred billion Korean won, by the recollection of an investor who held a position in the company at the time. The capital that subsequently mobilised behind Sequoia (DST Global, Hillhouse Capital, and additional institutional capital from Hong Kong, Australia, the United States, and Europe) brought Coupang’s total raise to a figure in the order of one trillion won and built what became Asia’s largest e-commerce platform. The structural sequel is a public-market success that requires no further introduction.

The structural antecedent is the four months.

Park Hee-duk, the Korean venture practitioner who held an investment position in the company at the time, has described the mechanism on the public record. The Coupang Series C cap table at the moment Sequoia entered held approximately twenty Korean shareholders, including domestic VCs and individual investors, most of whom had subscribed into the company on Korean shareholder-agreement terms. Those terms granted each shareholder consent rights over subsequent funding rounds. The rights were structured as individual veto positions held by each named subscriber on the cap table; coordination at the lead-investor level sat outside the convention.

Sequoia’s institutional limited partners would deploy capital into a Series C only under documentation reconciled to international convention. The contract architecture had to convert. Conversion required affirmative consent from each existing shareholder. Four months of dedicated negotiation reconciled the prior Korean architecture to a structure that accepted the new institutional capital. The four months operated alongside price and diligence as a separate workstream.

A Silicon Valley equivalent is illustrative. The contractual architecture that supports a US Series C is constructed in stages from the first round forward. A US seed-stage term sheet, often a one-page SAFE, establishes board governance, syndicate structure, and rights-waiver conventions. By Series C, every prior round has been negotiated against a documentary backbone that contemplates further institutional rounds. The consent process is built into the architecture. New capital enters in days or weeks because the structure was designed to absorb it.

The Korean architecture was designed for a different purpose, which it serves competently inside the domestic system. The translation cost is what foreign capital meets at the threshold.

The Coupang transaction closed because a threshold-based consent mechanism was reached: a workable approximation of the one-hundred-percent unanimous standard the original Korean shareholder agreements appeared to require, negotiated under pressure of a deal that mattered to every party at the table. Without that workaround, the round would have failed. Park’s framing on the public record is direct: Korean VCs failed Coupang at the contractual level; global VCs carried it across, at the cost of four months of architectural translation.

That is the asymmetry. A US Series C closes in weeks because the architecture is designed for the next round. A Seoul Series C closes in four months because the architecture is designed for the previous one.

The Five Levels

Park’s diagnosis, articulated across his interview record, identifies five distinct levels at which Korean venture contract architecture diverges from international institutional convention. The levels reinforce each other and form, in his framing, a coordination problem: every layer holds the others in place, and no single actor commands the position to shift one layer in isolation.

The fund level sets the calendar. Korean limited-partnership agreements run on eight-year primary terms; the international institutional standard runs ten years and beyond. The compression propagates downward into the rest of the architecture: a fund that must return capital in year eight operates against a different return clock than an international fund of the same vintage. The Korean Fund of Funds programme, by the cumulative comparison Park advances, has deployed sums of comparable order to Singapore’s Pavilion Capital. The difference in international standing reflects operating convention. Korea’s operating convention is internally coherent. The convention sits at a structural distance from the international one.

The contract level sets the relationship between investors. Korean shareholder agreements grant individual consent rights to each named investor over subsequent funding rounds. Club-deal structures replace syndicated rounds. Term sheets are negotiated on a per-investor basis. The result is a cap table that, by Series B, holds ten to fifteen consent-rights holders, each operating an effective veto. The Coupang Series C cap table held approximately twenty. International convention vests consent in a lead investor whose decision binds the syndicate; the Korean convention vests consent individually. Both arrangements function on their own terms. They translate poorly.

The governance level sets the relationship between investor and company. Board seats are taken inconsistently; company-building participation runs through informal channels alongside or outside the formal seat. The patient operating engagement that defines international venture practice operates at lower intensity inside the Korean convention. Korean VCs commonly exit at Series C, before the company-building phase has matured, because the fund-level eight-year clock and the LP evaluation framework reward earlier returns. The exit passes without reputational consequence inside the domestic system. The same exit executed by a Silicon Valley fund of comparable vintage would carry one.

The regulatory level sets the institutional context. KOSDAQ listing standards configure listing thresholds and founder-holding requirements around domestic capital structure assumptions, which shape company financing patterns from Series B onward. The Financial Services Commission’s standards apparatus is configured for the Korean financial-services market; global investment-finance convention sits outside its operating brief. The configuration is internally coherent. The international interface remains to be constructed.

The cultural level is the load-bearing layer. Korean capital operates inside a relational dynamic that retains the hierarchy between the party providing capital and the party receiving it. Silicon Valley capital underwent its transition in the 1970s and 1980s, when capital oversupply forced a recalibration that elevated founder authority. Korean capital, by the operating evidence Park assembles, holds the prior posture. Korean entrepreneurs are commonly born as sub-vendors to large corporations: first-tier and second-tier suppliers solving problems their corporate customers have already defined. The supply-chain logic of Korea’s industrial-conglomerate economy shapes the assumptions that follow capital through its life cycle, from the kind of company that gets funded to the contract that governs the funding.

The five layers reinforce one another. The fund-level clock generates pressure that produces the contract-level individual consent structure that supports the governance-level early exit that operates inside the regulatory-level listing constraint, on top of the cultural-level relational posture that justifies the whole.

This is what makes reform difficult. A Korean VC could, in theory, adopt international syndicated contract conventions; the eight-year fund-term clock would punish the resulting patience. A Korean LP could fund ten-year vehicles; the contract-level individual consent architecture would persist downstream. The Korean Financial Services Commission could constitute a global-standards unit; the cultural-level relational posture would persist beyond regulatory mandate. The architecture moves at the speed of the slowest layer, and the cultural layer is the slowest.

The interlock is the point.

Three Tiers of Awareness

The architecture has been present in the Korean venture market for the entire duration of the industry’s existence. What is recent is the beginning of explicit recognition that the architecture functions as architecture. The procedural frictions, taken individually, are visible to anyone operating inside the system; the systemic property is the new recognition.

The recognition runs at three tiers.

Tier one holds the structural diagnosis. The architecture exists as a single coordinated system; its five layers interlock; the system imposes a translation cost on cross-border capital flow that is invisible to actors operating inside it. Park Hee-duk has articulated this view across eight interviews and lectures on the public record, beginning in July 2022 and, by September 2025, presenting it directly to National Assembly members at a Democratic Party study session. A small cohort of Korean practitioners with comparable cross-border operational experience shares the view. The cohort is small enough that its members tend to know one another by name. Each member typically came to the diagnosis through the same route: personally attempting, and personally absorbing the cost of, executing a transaction across the architectural seam.

Tier two holds general policy awareness. The Korean institutional layer has begun, in the past eighteen months, to acknowledge that the layers exist and that they impose costs on the venture market. The acknowledgment runs at the level of symptoms; the mechanism that connects them sits outside the current frame. The Korean media and policy discourse frame the problem as a fairness issue between VCs and founders: an issue of relative power inside the domestic system. The framing names a real dynamic. It also stops short of the interlock: the contract architecture itself, taken as a coordinated system, is what makes the market incompatible with international institutional capital. Tier two has reached the layers; the structural mechanism beneath them sits a layer further down.

Tier three holds no awareness of the architecture as architecture. Most Korean entrepreneurs, most domestic VCs, and most of the business community operate inside the existing system as given. The contract conventions are received practice. The fund-term standard is the calendar. The governance posture is the way the work gets done. The architecture functions, for tier three, as invisible infrastructure.

Tier three is by far the largest population. Tier two is the institutional layer that has begun, in policy speeches and government forums, to surface the questions. Tier one is the practitioner cohort that holds the unified diagnosis.

The international stakeholder map adds a complication. If tier-one understanding is rare inside Korea, equivalent understanding among non-Korean actors is rarer still. The analytical chain required is exacting. One must understand the granular mechanics of Korean venture contract conventions, which are not documented in English at any sufficient depth. One must compare them rigorously against international institutional standards, which requires operational experience on both sides of the translation. One must identify the specific incompatibilities. One must map those incompatibilities onto investment outcomes for capital deployed across the seam. The chain is built from practitioner experience: transactions attempted, friction absorbed, the structure read from the inside. Market reports and consulting decks carry the layers in summary form; the interlock comes from the work.

The cohort of non-Korean actors with that experience is small. The cohort that has synthesised the experience into a structural framework is smaller. The intersection with the cohort that is currently building capital infrastructure designed for the architectural translation is smaller still.

The December 2025 Forum

Two institutional artefacts surfaced inside a single fortnight in late December 2025. On the eighteenth, the Ministry of SMEs and Startups (중소벤처기업부) published the Venture Four Great Powers Comprehensive Plan (벤처 4대 강국 도약 종합대책), naming alignment with global standards as a policy objective. Five days later, the Ministry and the Korea Venture Investment Corporation inaugurated the Venture Investment Contract Culture Development Forum (벤처투자 계약문화 발전 포럼), a standing body addressing precisely the questions Park has articulated on the public record since 2022: contract conventions, fund-term architecture, alignment with global standards. Both are tier-two artefacts. They identify the layers. The structural mechanism that holds the layers in coordinated tension sits outside the Comprehensive Plan’s framing.

This is the institutional moment.

The Korean policy apparatus has now placed contract culture, fund convention, and global-standards alignment inside its formal agenda. The forum is a listening exercise rather than a reform programme. The Comprehensive Plan is a framework; binding instruments will follow on whatever schedule the framework establishes for them.

The institutional shift the forum signals will, in time, work its way into Korean contract conventions, fund-term standards, and LP evaluation frameworks. Park’s own assessment of the timeline for that cultural change is ten to twenty years. The trust network on which a healthy venture market depends, where startups trust VCs and VCs trust their commitments and LPs trust their general partners, takes time. Policy mandate accelerates the institutional architecture; the cultural layer arrives on its own clock.

For the Korean ecosystem, the forum is the beginning of the reform path. For capital reaching the end of fund life in 2026 and 2027, it is information about a horizon beyond the calendar.

The forum confirms the gap. The diagnosis of the interlock waits for a later cycle of Korean institutional reflection. The structure that the diagnosis describes operates, unchanged, in the meantime.

What Global Capital Meets

A European limited partner subscribing into a Korean venture fund in 2026 meets the eight-year primary fund term and the LP evaluation convention configured around it. A European general partner attempting direct entry into a Korean Series C round meets the cap table populated by individual consent-rights holders under Korean shareholder-agreement convention. A Korean institutional limited partner reaching the end of a 2018-vintage fund meets the domestic exit infrastructure built for the previous architecture.

All three are meeting the same thing. The five layers, interlocked, operating exactly as they were designed to operate.

The capital that has been placed into the Korean venture market on Korean architectural terms will, on the calendar the architecture has set, complete its life cycle. The question raised by the first article in this series (where the $22.9 billion of Korean venture capital reaching the end of its fund life through 2027 actually goes) sits on top of the question raised by this one. The shortfall in domestic absorption capacity is the visible figure. The architecture inside which the absorption operates is the structural fact.

The contract is the gate. Capital reaching the wall meets the architecture before it meets any alternative.

The Korean policy response, as it has formed, will rebuild the architecture on a generational horizon. The non-Korean response, as it has formed, has often been to look elsewhere, to other Asian destinations whose contractual conventions are more legible to international institutional capital. Both responses are, on their own terms, rational. Closing the present gap on the calendar the present gap requires sits outside the scope of both.

The category that the architecture has made necessary is the subject of the rest of the series.

Sources:

• Park Hee-duk, interview and lecture series, eight items 2022–2025. Print interview: GoKorea (공감신문), 25 July 2022. Video interviews on Korean YouTube channels (이해, 야단법석, VC인터뷰). Video interview on 투자자 인사이트 (bkamp), YouTube. Lecture: Democratic Party 경제는 민주당 study session, 30 September 2025, YouTube.

• Park Hee-duk interview, “Singapore Foreign Capital” segment — primary reference for the Coupang Series C consent precedent.

• Ministry of SMEs and Startups (중소벤처기업부), Venture Four Great Powers Comprehensive Plan (벤처 4대 강국 도약 종합대책), 18 December 2025: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156736553

• Ministry of SMEs and Startups and Korea Venture Investment Corporation, Venture Investment Contract Culture Development Forum (벤처투자 계약문화 발전 포럼), inaugurated 23 December 2025: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156750991

• Public reporting on Coupang’s growth-round capital structure (2014).

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All decisions should be made based on independent research and consultation with qualified advisors.

About Saint Clair: Saint Clair designs and builds cross-border capital infrastructure between Europe and Asia, proposing access where access is scarce and creating structure where structure is absent. Since 2016.

Saint Clair Capital · 그라운드 트루스 | 2026년 7월

핵심 요약

쿠팡 시리즈 C에서 가장 오래 걸린 건 가격도, 실사도 아니었다. 구조였다. Sequoia Capital이 들어올 때 텀시트는 몇 주 만에 끝났지만, 동의 절차에만 넉 달이 걸렸다. 당시 한국 캡테이블에는 스무 명 가까운 주주가 저마다 지분을 들고 있었고, 그 동의권은 하나같이 국내 계약 관행에 따라 주주별로 따로 행사되도록 짜여 있었다. 그러니 넉 달이 걸릴 수밖에 없었다. 스무 명의 거부권을 Sequoia의 LP가 요구하는 글로벌 신디케이트 구조에 하나씩 맞춰 넣어야 했기 때문이다.

한국 벤처의 ‘구조’란 이런 것이다. 펀드 만기, 계약 관행, 이사회 거버넌스, 규제 환경, 그리고 자본 문화. 이 다섯 층위가 서로 맞물려 하나의 시스템을 이루고, 한국 안에서는 그 시스템이 빈틈없이 돌아간다. 문제는 해외 기관 자본이 들어오는 순간 불거진다. 층위 하나하나가 국내용으로 짜여 있어, 그때마다 ‘전환 비용’이 발생하기 때문이다. 그런데 정작 그 안에서 일하는 사람들에게 다섯 층위는 그저 일이 돌아가는 방식일 뿐이다. 잘 짜인 체계 안에 있으면, 그 체계는 보이지 않는 법이다.

한국 정책 당국도 지난 18개월 사이 이 다섯 층위의 존재를 인정하기 시작했다. 그러나 여기서 한 걸음 더 들어간 진단, 곧 이 다섯 층위가 하나의 조정된 시스템으로 맞물려 국경 간 자본에 구조적 비용을 치르게 한다는 진단은 한국 안에서도 좀처럼 나오지 않는다. 베테랑 VC 박희덕 대표가 2022년부터 2025년까지 여덟 차례 인터뷰와 강연을 통해 이 진단을 다듬어 왔고, 2025년 9월에는 이를 국회의원들 앞에 직접 내놓았다.

정책은 다섯 층위를 짚었다. 다만 그 층위들이 서로 어떻게 맞물리는지, 그 작동 원리까지는 아직 규명하지 못했다. 이 글 역시 해외 자본이 한국에서 마주하는 현실을 보여줄 뿐, 해법을 내놓지는 않는다. 그 답은 독자의 몫이다.

네 달

넉 달. Sequoia Capital의 쿠팡 시리즈 C가 2014년 마무리되기까지, 동의 협상에만 걸린 시간이다. 텀시트는 몇 주 만에 끝났고, 나머지가 모두 설계였다.

진입 당시 기업가치는 약 3,000억 원으로 추산된다. 쿠팡 지분을 보유했던 한 투자자의 회고에 근거한 수치다. Sequoia를 시작으로 DST Global과 Hillhouse Capital이 뒤를 이었고, 홍콩·호주·미국·유럽의 기관 자본이 차례로 합류했다. 이들의 자본이 누적 조달 규모를 1조 원대로 끌어올렸고, 그 위에서 쿠팡은 아시아 최대 전자상거래 플랫폼으로 올라섰다. 이후 공모 시장의 성과는 굳이 설명할 필요가 없다. 그 모든 성공의 구조적 출발점이 이 넉 달이었다.

당시 쿠팡 지분을 들고 있던 박희덕 대표는 이 과정을 공식 석상에서 직접 설명해 왔다. Sequoia 진입 시점에 쿠팡 캡테이블에는 한국 주주가 20명 가까이 있었다. 국내 VC와 개인 투자자가 뒤섞인 구성이었다. 대다수가 한국식 주주간계약 조건으로 들어왔고, 이 계약은 주주마다 후속 라운드 동의권을 부여했다. 결국 각자가 거부권을 따로 행사하는 구조였고, 리드 투자자가 한데 모아 조율하는 방식은 애초에 한국 관행 밖이었다.

Sequoia의 기관 LP는 국제 관행에 맞는 문서가 있어야만 자본을 집행한다. 계약 구조 자체를 갈아엎어야 했고, 그러려면 기존 주주 전원의 동의가 필요했다. 넉 달의 협상 끝에 한국식 계약은 새 기관 자본을 담을 구조로 바뀌었다. 가격이나 실사와는 전혀 별개의 일이었다.

실리콘밸리는 다르다. 미국 시리즈 C를 떠받치는 계약 구조는 첫 라운드부터 차곡차곡 쌓인다. 시드 단계 텀시트는 한 장짜리 SAFE인 경우가 많지만, 그 안에 이미 이사회 거버넌스와 신디케이트 구조, 권리 포기 관행이 정해져 있다. 그래서 시리즈 C에 이를 무렵이면, 이전 단계가 모두 향후 기관 라운드를 내다본 골격을 따라 협상을 마친 상태다. 동의 절차가 계약 안에 처음부터 들어 있는 셈이다. 신규 자본이 며칠, 길어야 몇 주면 들어오는 이유가 여기 있다. 구조 자체가 그 흡수를 전제로 짜였기 때문이다.

한국식 계약은 본래 다른 목적을 위해 설계됐고, 국내에서는 그 목적을 충실히 해낸다. 그런 계약이 전환 비용을 낳는 것은, 해외 자본이 그 문턱을 넘는 순간부터다.

쿠팡 거래가 성사된 건 ‘기준선 동의’라는 절충안 덕분이다. 한국 주주간계약이 겉으로 요구하던 100% 만장일치를 대신한 실무적 합의였다. 모두에게 중요한 거래라는 압박이 협상을 끌어냈다. 이 우회로가 없었다면 라운드는 무산됐을 것이다. 박 대표의 진단은 명료하다. 계약 차원에서 한국 VC는 쿠팡을 끝까지 받쳐주지 못했고, 그 빈자리를 글로벌 VC가 메웠다. 그리고 그 글로벌 자본을 받아들이기 위해 한국이 치른 비용이, 바로 넉 달에 걸친 계약 구조의 재설계였다.

그렇다면 이 비대칭은 어디서 오는가. 미국 시리즈 C가 몇 주 만에 끝나는 건 구조가 ‘다음 라운드’를 내다보고 설계된 덕분이다. 서울 시리즈 C가 넉 달을 끄는 건 구조가 ‘직전 라운드’만 바라보고 설계된 탓이다.

다섯 층위

박 대표가 인터뷰 곳곳에서 정리한 진단은 다섯 층위로 압축된다. 한국 벤처 계약이 국제 기관 관행과 갈라지는 다섯 지점이다. 이 다섯은 서로를 떠받친다. 박 대표의 표현으로는 하나의 ‘조정 문제(coordination problem)’다. 각 층위가 나머지를 붙들어 두기 때문에, 누구도 혼자서는 한 층위를 흔들 수 없다.

첫째, 펀드 차원이 시간표를 정한다. 한국 LP 계약은 8년 만기를 기준으로 삼는다. 국제 기관은 10년 이상이다. 짧은 만기는 구조 전체로 흘러내린다. 8년 차에 자본을 돌려줘야 하는 펀드는 같은 빈티지의 국제 펀드와 회수 시간표가 다를 수밖에 없다. 박 대표가 누적으로 비교한 바로는, 한국 모태펀드는 싱가포르 Pavilion Capital과 엇비슷한 규모를 집행해 왔다. 다만 국제 평판의 차이는 운영 관행에서 갈린다. 한국식 관행은 그 자체로는 앞뒤가 맞지만, 국제 기준과는 구조적으로 거리가 있다.

둘째, 계약 차원이 투자자 사이의 관계를 정한다. 한국식 주주간계약은 후속 라운드 동의권을 투자자마다 따로 부여한다. 그 결과 신디케이트 대신 클럽딜이 자리 잡고, 텀시트는 투자자별로 협상된다. 시리즈 B쯤이면 동의권 보유자가 10~15명에 이르고, 저마다 사실상 거부권을 쥔다. 쿠팡 시리즈 C에서는 20명 안팎이었다. 국제 관행에서는 리드 투자자의 결정이 신디케이트 전체를 묶는다. 반면 한국은 동의권을 잘게 쪼갠다. 둘 다 제 체계 안에서는 매끄럽다. 마찰은 두 체계를 맞춰야 하는 순간 터진다.

셋째, 거버넌스 차원이 투자자와 기업의 관계를 정한다. 이사회 의석은 일관되게 확보되지 않고, 기업 형성기의 관여도 공식 경로보다 비공식 채널로 흐른다. 국제 벤처의 핵심인 ‘지속적 운영 관여’가 한국에서는 약하게만 이뤄진다. 그래서 한국 VC는 시리즈 C 단계에서 발을 빼는 경우가 흔하다. 기업이 채 무르익기도 전이다. 8년 만기와 LP 평가 체계가 조기 회수를 보상하는 데다, 그 회수에 국내에서는 평판 비용이 따라붙지 않기 때문이다. 같은 결정을 실리콘밸리 동급 펀드가 내렸다면 상당한 평판 부담을 졌을 것이다.

넷째, 규제 차원이 제도적 맥락을 정한다. 코스닥 상장 요건과 창업자 지분 기준은 국내 자본구조를 전제로 짜여 있다. 그 영향은 시리즈 B 이후 자금조달 패턴에 곧장 미친다. 금융위원회의 기준도 국내 시장을 겨냥한 만큼, 글로벌 투자금융 관행은 애초에 사정권 밖이다. 그 구성은 안에서 보면 빈틈이 없다. 다만 국제 자본과 맞닿는 접점은 아직 설계되지 않았다.

다섯째, 문화 차원이 가장 무거운 층위다. 한국 자본은 돈을 대는 쪽과 받는 쪽의 위계를 그대로 둔 채 움직인다. 실리콘밸리는 1970~80년대에 한 차례 변혁을 거쳤다. 자본 과잉이 부른 재조정이 창업자의 입지를 끌어올렸다. 박 대표가 모은 현장 증거로는, 한국 자본은 여전히 그 변혁 이전 단계에 머물러 있다. 한국 창업자는 출발부터 대기업의 1·2차 협력사로 형성되는 경우가 많다. 대기업이 이미 정의해 둔 문제를 푸는 자리다. 산업자본주의의 공급망 논리가 자본의 생애 주기 전반으로 그대로 이어진다. 어떤 기업이 돈을 받고 그 돈이 어떤 계약 안에서 움직이는지, 모든 단계에 같은 가정이 깔려 있다.

다섯 층위는 이렇게 서로를 떠받친다. 펀드의 빠듯한 시간표가 계약 차원의 개별 동의 구조를 낳고, 그 구조가 거버넌스 차원의 조기 회수를 떠받친다. 그 회수는 규제 차원의 상장 제약 안에서 이뤄지고, 이 전체를 문화 차원의 관계적 태도가 맨 아래에서 지탱한다.

개혁이 어려운 이유가 여기 있다. 한국 VC는 이론상 국제 신디케이트 계약을 받아들일 수 있다. 그러나 그 인내심은 8년 만기라는 시한 앞에서 대가를 치른다. 한국 LP는 10년 만기 비클을 만들 수 있지만, 그 아래 개별 동의 구조는 그대로다. 금융위가 글로벌 표준 전담 부서를 둬도, 문화 차원의 태도만큼은 규제 권한 밖에 남는다. 결국 구조 전체는 가장 느린 층위의 속도에 맞춰 움직이고, 그 가장 느린 층위가 바로 문화다. 다섯 층위를 하나로 묶어 두는 이 맞물림이야말로 구조의 핵심이다.

인식의 세 계층

이 구조는 한국 벤처 산업의 역사만큼 오래됐다. 새로운 것은 그것을 하나의 맞물린 시스템으로 바라보는 시선이며, 이 시선은 최근에야 등장했다. 절차 곳곳에서 생기는 마찰이야 안에서 일하는 사람이라면 누구나 안다. 다만 그 마찰들을 한데 묶어 하나의 시스템으로 파악하는 눈, 그것이 새롭다. 그리고 이렇게 구조를 알아보는 인식은 다시 세 계층으로 나뉜다.

첫 번째 계층은 구조적 진단을 갖춘 집단이다. 구조가 하나의 조정된 시스템이고, 다섯 층위가 맞물리며, 그 시스템이 국경 간 자본에 전환 비용을 치르게 한다는 진단이다. 정작 안에서 일하는 사람의 눈에는 이 비용이 잘 보이지 않는다. 박 대표는 2022년 7월부터 여덟 차례에 걸친 인터뷰와 강연으로 이 시각을 가다듬어 왔고, 2025년 9월에는 국회의원들 앞에서 이를 직접 발표하기에 이르렀다. 비슷한 경험을 가진 소수의 한국 실무자가 이 진단을 공유한다. 서로 이름까지 알 만큼 작은 집단이다. 이들은 대개 같은 경로로 그 진단에 다다른다. 체계의 경계를 넘나드는 거래를 직접 해보고, 그 비용을 스스로 치러 본 경험이다.

두 번째 계층은 일반적인 정책 차원의 인식이다. 한국 제도권은 지난 18개월 사이 다섯 층위의 존재와 그 비용을 인정하기 시작했다. 다만 개별 현상을 하나씩 지적하는 단계에 머물러 있을 뿐, 그 현상들이 어떻게 서로 이어지는지까지는 나아가지 못했다. 한국 언론과 정책 담론은 이를 흔히 VC와 창업자 사이의 공정성 문제로, 곧 국내에서 누가 더 힘을 쥐느냐의 문제로 풀어낸다. 물론 그 안에도 실제 역학은 존재한다. 그러나 다섯 층위가 하나로 맞물려 돌아간다는 사실, 계약 구조 전체가 하나의 조정된 시스템으로서 한국 시장과 국제 기관 자본을 어긋나게 만든다는 사실까지는 끝내 미치지 못한다. 요컨대 두 번째 계층은 다섯 층위를 발견했지만, 그 아래에서 층위들을 움직이는 작동 원리는 아직 붙들지 못했다.

세 번째 계층은 구조를 구조로 보지 않는다. 다수의 한국 창업자와 국내 VC, 비즈니스 커뮤니티가 기존 시스템을 주어진 조건으로 받아들이고 그 안에서 움직인다. 이들에게 계약 관행은 원래 그런 것이고, 펀드 만기는 시장이 정해 둔 시간표이며, 거버넌스 태도는 몸에 밴 습관이다. 그렇게 세 번째 계층에서 구조는 끝내 보이지 않는 인프라로 남는다.

세 번째 계층이 단연 가장 크다. 두 번째 계층은 정책 발표와 정부 포럼을 통해 이 질문을 공론의 장으로 끌어낸 제도권이며, 첫 번째 계층은 통합된 진단을 갖춘 소수의 실무자 집단이다.

여기에 국제 이해관계자라는 변수가 하나 더 겹친다. 첫 번째 계층 수준의 이해는 한국 안에서도 드물고, 해외로 나가면 더더욱 드물다. 그 수준에 이르기까지 거쳐야 할 분석이 워낙 까다롭기 때문이다. 영어로는 제대로 정리된 적 없는 한국 벤처 계약의 세부를 정확히 꿰고, 그것을 국제 기관 기준과 엄밀히 견줄 수 있어야 한다. 그러려면 양쪽 실무를 다 겪어봐야 한다. 구체적인 비호환 지점을 짚어내고, 그 비호환이 경계를 넘나든 자본의 성과로 어떻게 이어졌는지까지 설명할 수 있어야 한다.

이런 이해는 오직 현장에서만 나온다. 직접 거래를 성사시키고, 그 과정의 마찰을 몸소 겪고, 구조를 안에서 지켜본 경험이 있어야 한다. 시장 보고서나 자문사 자료는 다섯 층위를 요약해 줄 뿐, 그 경험을 대신하지는 못한다. 그런 경험을 갖춘 비(非)한국계 투자자는 드물고, 그것을 구조적 틀로 정리해 낸 이는 더욱 드물다. 그리고 그 이해를 바탕으로 국경 간 자본 인프라를 실제로 세우는 데까지 나아간 이는, 그중에서도 극소수다.

2025년 12월의 포럼

2025년 12월 후반, 한국 정책 당국은 보름 사이에 두 질문을 잇따라 공론장에 올렸다. 18일, 중소벤처기업부는 「벤처 4대 강국 도약 종합대책」을 발표하며 글로벌 표준 정합성을 정책 목표로 명시했다. 닷새 뒤, 중기부와 한국벤처투자는 「벤처투자 계약문화 발전 포럼」을 발족시켰다. 박 대표가 2022년부터 공식 석상에서 정리해 온 질문들, 곧 계약 관행과 펀드 만기, 글로벌 표준 정합성을 그대로 다루는 상설 협의체다. 두 문서 모두 두 번째 계층의 산물답게 다섯 층위를 짚어낸다. 다만 그 층위들이 서로를 붙들고 놓지 않는 작동 원리까지는, 여전히 대책이 다루는 범위 밖에 있다.

그럼에도 지금은 분명한 제도적 변곡점이다. 한국 정책 당국이 계약 문화와 펀드 관행, 글로벌 표준 정합성을 처음으로 공식 의제에 올렸기 때문이다. 물론 포럼은 개혁이라기보다 ‘청취’에 가깝고, 종합대책 역시 어디까지나 틀에 머문다. 구속력 있는 수단은 그 틀이 정한 일정을 따라 뒤에 온다.

포럼이 보낸 신호는 시간을 두고 한국의 계약 관행과 펀드 만기 기준, LP 평가 체계 속으로 서서히 스며들 것이다. 다만 그 시간은 짧지 않다. 박 대표는 문화가 바뀌는 데 10년에서 20년을 내다본다. 건강한 벤처 시장을 떠받치는 신뢰의 네트워크, 곧 창업자가 VC를, VC가 제 판단을, LP가 GP를 믿는 관계가 자리 잡기까지 그만한 시간이 들기 때문이다. 그래서 정책이 제도의 골격을 앞당길 수는 있어도, 문화는 끝내 제 속도로 움직인다.

한국 벤처 시장에 이 포럼은 개혁의 출발점이다. 그러나 2026년과 2027년에 만기를 맞는 자본에는, 제 시한을 한참 넘어선 뒷날의 이야기일 뿐이다. 포럼이 확인해 주는 것은 ‘격차’가 있다는 사실까지다. 그 격차가 왜 좀처럼 좁혀지지 않는지, 다섯 층위가 어떻게 맞물려 있는지에 대한 진단은 한국 제도권의 다음 성찰 주기를 기다려야 한다. 그리고 그사이에도 구조는 변함없이 돌아간다.

해외 자본이 마주하는 것

2026년 한국 벤처펀드에 출자하는 유럽 LP는 8년 만기, 그리고 그것을 둘러싼 LP 평가 체계와 마주한다. 한국 시리즈 C에 직접 들어오려는 유럽 GP는 개별 동의권 보유자로 채워진 캡테이블과 마주한다. 2018년 빈티지 만기를 맞은 한국 기관 LP는 이전 구조를 위해 만들어진 국내 출구 인프라와 마주한다. 셋 모두 같은 것과 마주한다. 다섯 층위가 맞물린 채 설계대로 돌아가는 현실이다.

한국식 계약 구조에 들어간 자본은 그 구조가 정한 시간표대로 생애를 마친다. 본 시리즈 첫 글이 던진 질문이 여기서 다시 고개를 든다. 2027년까지 만기를 맞는 229억 달러는 어디로 갈 것인가. 국내 흡수 여력의 부족은 눈에 보이는 숫자다. 그러나 그 흡수가 일어나는 자리의 구조는 그 숫자 너머의 사실이다. 결국 계약이 관문이다. 만기 절벽에 다다른 자본은 어떤 대안을 만나기 전에 구조부터 만난다.

한국 정책의 대응은 지금 모습대로 한 세대에 걸쳐 구조를 다시 쌓는다. 한국 밖의 대응은 지금 모습대로 다른 곳을 살피는 쪽으로 움직여 왔다. 국제 기관 자본에 더 잘 맞는 계약 관행을 갖춘 다른 아시아 시장이다. 둘 다 제 논리 안에서는 합리적이다. 다만 그 격차를, 격차가 요구하는 속도로 메우는 일은 어느 쪽 사정권에도 없다. 구조가 만들어낸 이 새로운 범주를, 이번 시리즈의 다음 글들에서 이어 다룬다.

Sources:

• 박희덕(Park Hee-duk) 대표 인터뷰·강연 시리즈, 2022~2025년 총 8회. 지면 인터뷰: 『공감신문』(GoKorea), 2022년 7월 25일. 한국 YouTube 채널(이해, 야단법석, VC인터뷰) 영상 인터뷰. 투자자 인사이트(bkamp) YouTube 영상 인터뷰. 더불어민주당 「경제는 민주당」 스터디 세션 강연, 2025년 9월 30일, YouTube.

• 박희덕 인터뷰 ‘싱가포르 해외 자본’ 섹션 — 쿠팡 시리즈 C 동의 사례 일차 참조 자료.

• 중소벤처기업부, 「벤처 4대 강국 도약 종합대책」, 2025년 12월 18일: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156736553

• 중소벤처기업부·한국벤처투자, 「벤처투자 계약문화 발전 포럼」 발족식, 2025년 12월 23일: https://www.korea.kr/briefing/pressReleaseView.do?newsId=156750991

• 쿠팡 시리즈 C 자본 구조에 관한 공개 보도 (2014년 전후).

Disclaimer: 이 글은 정보 제공만을 목적으로 한다. 투자 자문에 해당하지 않는다.

About Saint Clair: 세인트 클레어(Saint Clair)는 유럽과 아시아를 잇는 국경 간 자본 인프라를 설계하고 구축한다. 접근이 희소한 자리에 접근을, 구조가 부재한 자리에 구조를 제안해 왔다. 2016년 이래로.