1. The Night of the National Security Law
June 30, 2020, 11:00 p.m. — minutes before midnight in Beijing time. Hong Kong's National Security Law (NSL) took effect. Secession, subversion, terrorism, and collusion with foreign forces — up to life imprisonment. The law, the text made explicit, also reached non-residents outside Hong Kong.
That night, the lights stayed on in the law-firm offices of Central. Emergency calls went back and forth. The content of most of the calls was the same: Begin the Singapore incorporation.
The numbers tell what happened in the four years that followed.
Hong Kong's benchmark Hang Seng Index fell from its all-time high of 33,223 points in January 2018 to roughly half that level by January 2024 — a decline of nearly 50 percent. In 2023 alone it dropped another 14 percent. This was not an ordinary business cycle. The share of foreign shareholders fell 4 percentage points from its 2019 peak — a 14 percent relative contraction.
The number of foreign firms with regional headquarters in Hong Kong fell 13.3 percent between 2019 and 2023. French companies dropped from 96 to 80; American companies from 278 to 240. FedEx moved its Asia-Pacific headquarters to Singapore. Dyson had already relocated its global headquarters from the United Kingdom to Singapore in 2019.
It was not just capital that left. People left too. Through the BN(O) visa introduced by the British government for Hong Kong citizens, 163,400 people migrated to the United Kingdom between 2021 and March 2025. Seventy percent of them held college degrees or higher; 39 percent worked in professional occupations. Hong Kong's labor force shrank by 94,100 in 2022 alone — the largest annual decline since statistics began to be compiled in 1985.
And a considerable share of that capital and talent headed for — a small city-state four hours away by plane.
2. The Great Migration of the Family Offices
Official statistics from the Monetary Authority of Singapore (MAS) show the scale of the migration.
At the end of 2020, about 400 tax-advantaged Single Family Offices (SFOs) were registered in Singapore. By the end of 2022: 1,100. End of 2023: 1,400. End of 2024: roughly 2,000. A fivefold increase in four years. More than 59 percent of Asian family offices are now based in Singapore.
In a September 2024 speech, MAS Deputy Chairman Chee Hong Tat disclosed that Singapore's total assets under management (AUM) had reached S$5.4 trillion, with a five-year compound annual growth rate of about 10 percent.
The core driver of this surge was the migration of wealthy Chinese mainlanders and Hong Kong–based family offices. The industry called it the Singapore family-office mad rush. What flew from Hong Kong to Singapore was not only money. Lawyers, accountants, and asset managers moved with it. An entire financial ecosystem was transplanted.
Hong Kong hit back. In 2023 it created a new tax incentive for family offices and threw everything at retaining the wealthy. According to Deloitte's 2024 report, the number of SFOs in Hong Kong reached roughly 2,700 — ahead of Singapore in raw count. Total AUM, at about $4.53 trillion, narrowly edged Singapore's $4.46 trillion.
The structure behind the numbers, however, is different. A large share of Hong Kong's family offices hold mainland Chinese capital. This means Hong Kong's economic dependence on China is deepening. Singapore is absorbing diversified global capital — from greater China, Southeast Asia, India, and the Middle East. The true competitiveness of a hub is measured less by total volume than by the diversity of the capital it holds.
The IPO market contracted sharply as well. Hong Kong IPO proceeds declined for three consecutive years from the 2021 peak. The $7.3 billion raised in 2023 was the lowest level since 1999. In 2024, Hong Kong recovered to fourth place globally, but the figure is still far below the peak.
Capital leaving Hong Kong did not flow only to Singapore. The Dubai International Financial Centre (DIFC) emerged as a third destination. In 2024 alone, 200 new family offices were established at DIFC — a 33 percent increase from the previous year. The top 120 families at DIFC manage combined assets in excess of $1.2 trillion. Specific moves are on the record. Patrick Tsang, chairman of the Hong Kong-based Tsang Group, set up a base in Dubai in 2022 and then announced expansions into Abu Dhabi and Riyadh. Hong Kong-based Landmark Family Office also opened a Dubai office. These families have not left Hong Kong entirely, but the geographic dispersion of their asset structures reflects an erosion of confidence in Hong Kong.
Hong Kong → Singapore → Dubai. Capital flows to where trust exists, and leaves where trust has disappeared.
3. A Safe Place for Both Sides
In Chapter 3 we saw Lee Kuan Yew's founding of Singapore. Declaring his country a poisonous shrimp, he built a nation too toxic to swallow. In what form is that strategy operating in the twenty-first century?
Singapore functions as the most refined capital-transit hub in the world. As of 2023, bilateral trade between China and Singapore totaled $108.39 billion. Singapore has been China's largest source of new foreign investment for eleven consecutive years since 2013. At the same time, most major American firms — Google, Meta, Amazon, Microsoft — keep their Asian regional headquarters in Singapore. More than 4,200 multinational corporations maintain regional headquarters here, and over 37,000 international firms in total operate from the island.
For American firms, Singapore is the gateway into Asian markets without Chinese mainland risk. For Chinese firms, it is the window through which they can reach global capital markets while avoiding direct U.S. sanctions. This contradictory coexistence is Singapore's core value.
The most striking recent phenomenon is Singaporization. The AI startup Manus, originally headquartered in Beijing, moved its corporate base to Singapore in 2024. Chinese-founded companies migrating to Singapore and repositioning themselves as Singaporean firms in order to reach Western capital have become a spreading strategy. A laundering effect that severs the direct link to mainland China.
In Chapter 5 we saw Israel choose the American side decisively over the Haifa port question. Singapore is different. It does not choose either side. Instead, it provides both sides with functions that neither can do without.
The China–Singapore Suzhou Industrial Park, established in 1994, combined the strategic visions of Lee Kuan Yew and Deng Xiaoping. What China needed was not a factory shell but an administrative system — urban planning, policy discipline, investment know-how. The software. Singapore provided it, and in return received the status of gateway into the mainland. Lee Kuan Yew acknowledged a shared Han cultural heritage with China while drawing a sharp line between Singapore and any position as a Chinese satellite — an independent, multi-ethnic nation, not a client. We are Chinese but not China. Thirty years later, that identity strategy became the institutional foundation of Singaporization.
Meanwhile, the smallest of the ten ASEAN countries exercises, paradoxically, the greatest diplomatic influence. As an honest broker, Singapore serves as the adjuster where the interests of great powers cross. ASEAN itself is Singapore's strategic buffer. If in Chapter 3 Lee Kuan Yew declared his country a poisonous shrimp, in the 1980s his son, Brigadier-General Lee Hsien Loong, extended the metaphor to that of a porcupine. The poisonous shrimp was passive deterrence; the porcupine is active defense. From a country that cannot be swallowed, to a country that cannot be touched.
Changi Naval Base is one of a handful of facilities in the world that can accommodate an American aircraft carrier. More than 800 U.S. troops and 15 military command centers are stationed in Singapore — the largest permanent American presence in Asia south of Japan. But Singapore does not call this an alliance. It calls it a strategic partnership. Not a base, but access. It meshes perfectly with the U.S. places, not bases strategy.
Through this carefully calibrated positioning, Singapore takes shelter under the American security umbrella without letting itself be labeled an anti-China ally. At the same time, it maintains deep economic ties with China through the Suzhou Industrial Park. Belonging fully to neither side while remaining indispensable to both — the design Lee Kuan Yew drew up continues to operate as the twenty-first-century working principle of Singapore.
4. Regulatory Goldilocks
Singapore's most refined product is not a semiconductor, nor a berth for an aircraft carrier. It is the regulatory environment itself.
National AI Strategy 2.0 (NAIS 2.0), announced in December 2023. The subtitle signals the character of the strategy: AI for the Public Good — For Singapore and the World. Three core shifts. AI is reframed from a good-to-have opportunity into a must-know imperative. A single-project mindset is replaced by a comprehensive systems approach — three systems, ten enablers, fifteen actions. And Singapore begins to pursue a role as global AI standard-setter, not merely a domestic adopter.
The central tool of the strategy is the AI Verify Foundation, led by the Infocomm Media Development Authority (IMDA). Its aim is to develop an open-source AI testing framework so that firms and regulators can verify the trustworthiness of AI systems. Its premier members include Google, IBM, and Microsoft. The framework is built to be consistent with the EU AI Act and the OECD's AI Principles.
Singapore's strategic insight shows itself here. This is not rule-making; it is tool-making. Rules apply only within borders. Tools can be exported. Where the EU AI Act produced a law that regulates, Singapore built a tool that verifies regulation. AI Verify's framework is consistent with EU rules, OECD principles, and Singapore's own guidelines all at once. A firm in any jurisdiction can use the tool to test whether its AI meets global standards. The regulatory environment itself becomes an export product.
In parallel, MAS runs a regulatory sandbox. Fintech firms are granted temporary exemptions from parts of the existing rulebook and can test their products in the live market. Failure incurs no regulatory penalty. Success opens a path to a formal license. Experiment first, regulate later. While Korea was busy banning Tada under pressure from the taxi industry and oscillating between allowing and blocking telemedicine, Singapore was busy institutionalizing experimentation.
In 2025, Singapore committed a further S$1 billion over five years to the National AI Research and Development (NAIRD) plan. A country of 5.8 million people spending, on a GDP-adjusted basis, roughly eighteen times the U.S. level on AI R&D.
Why is Singapore's institutional speed different from other countries'? It is the result of overlapping structural conditions.
A population of 5.8 million is the first factor. The cost of policy experimentation is low. A new AI policy can be implemented nationwide and revised quickly. Layered on top of this is the long incumbency of the People's Action Party (PAP). In power continuously for 67 years since 1959, the PAP guarantees policy continuity across political cycles and enables 5-to-10-year strategic execution. The inter-agency structure is different as well. The Smart Nation Group designs the strategy; IMDA, AI Singapore, and A*STAR execute it — a vertically integrated, single-track chain of command that delivers policy signals directly to the ground. Finally, a unicameral parliament. Every bill passes through a single deliberative process, and the legislative speed is simply on a different order from other countries'.
Combine these conditions and what operates is a regulatory goldilocks — a regulatory environment neither too strict nor too loose. Flexible enough to let firms innovate, predictable enough for investors to trust. This is the product Singapore sells.
5. The Geopolitics of the Sovereign Wealth Fund
Singapore does not sell only regulation. It buys indispensability with money as well.
GIC. Assets under management of roughly $900 billion to $940 billion. One of the world's largest sovereign wealth funds. In February 2026, GIC co-led Anthropic's $30 billion Series G round, valuing the American AI firm at $380 billion. With that single investment, GIC secured a core position in the world's third-largest private company, after OpenAI and xAI.
Temasek Holdings is a sovereign wealth fund of a different character from GIC. Where GIC concentrates on long-term preservation of overseas assets, Temasek invests aggressively in strategic industries. In the first half of 2024, Temasek joined the AI Infrastructure Partnership consortium led by BlackRock, Microsoft, and MGX, committing to more than $30 billion in AI-infrastructure projects.
The GIC+Temasek dual structure is the core of Singapore's sovereign-wealth strategy. GIC is responsible for long-term asset preservation and for securing equity positions in the frontier AI firms of the world; Temasek is responsible for locking in leadership of the Southeast Asian digital economy. The roles are separated but strategically linked. One hand takes a seat in the best AI firms in the world; the other hand lays the infrastructure on which that technology will run across Southeast Asia.
Korea's National Pension Service (NPS) manages roughly ₩1,100 trillion (about $800 billion), a scale comparable to GIC's. But its degree of investment freedom is fundamentally different. Because the NPS carries the liability of paying out retirement benefits, high-risk investment is restricted. Any loss of principal becomes a political issue. Investment decisions have to pass through a multi-layered structure that includes the Investment Committee and parliamentary audits. The Korea Investment Corporation (KIC) manages $206.5 billion — roughly a quarter of GIC's scale — and has nothing resembling a GIC-plus-Temasek dual structure.
In December 2025, the idea of reshaping KIC into a Temasek model was formally raised in Korea. It has not yet been made concrete. In the AI sovereignty race, a country that does not secure a strategic equity stake falls permanently behind on technology transfer, talent acquisition, and standard-setting.
6. The Ceiling of the Poisonous Shrimp
Before idealizing Singapore, one has to see the ceiling of this country.
The most fundamental vulnerability. Singapore has no core technology of its own. It is a hub but not a core. OpenAI is not a Singaporean company. Neither is Anthropic. Neither is Google DeepMind. In semiconductors, biotech, or advanced materials, Singapore has no firm that holds an original source technology. The irreplaceable technology assets we saw in Chapter 4's TSMC and Chapter 5's Israeli cybersecurity startups simply do not exist here.
Singapore is a connection point in the global supply chain, not an origin point. It is excellent at attracting foreign firms, but those firms do not treat Singapore as a true home. As they scale, they relocate headquarters to the United States or de-prioritize Singapore.
There is also the wall of its population. Of 5.8 million people, foreign workers make up about 40 percent of the total labor force. This is at once a strength and a fragility. Talent-attraction capacity is national competitiveness, but it carries the risk of talent outflow under geopolitical change. Unlike Korea (52 million) or Japan (125 million), Singapore has no path of domestic validation followed by global expansion.
There is post-Lee-Kuan-Yew political risk as well. Lawrence Wong, inaugurated as the fourth prime minister in May 2024, was elected in part as a compromise choice among the PAP's fourth generation. Wong himself admitted that the era of single-party dominance is over and cannot go back to. The opposition is increasing its seat count, and pressure toward political pluralization is building. The deeper the U.S.–China rivalry cuts, the harder Singapore's neutrality becomes to maintain. Whether Wong's pro-Singapore line can function as an operational strategy has not yet been tested.
And there is the vulnerability created by the fact that roughly 74 to 76 percent of the population is ethnic Chinese. A 2022 Pew Research survey showed 67 percent of Singaporeans holding a favorable view of China — the highest figure among 19 countries surveyed. Attempts to exploit this opinion landscape are being documented in specific detail. The Washington Post and others have reported that the Chinese-language Singaporean daily Lianhe Zaobao (联合早报) tends to echo Beijing's messaging. There are also reports that Chinese authorities identify Singapore's high-education, high-net-worth ethnic-Chinese academics as candidates sympathetic to the Party's interests. Chinese-immigrant lifestyle sites have been documented propagating Beijing's geopolitical positions. Media, academia, and online — a simultaneous approach on three channels.
Lee Kuan Yew's identity strategy — We are Chinese but not China — becomes harder to sustain the deeper U.S.–China rivalry cuts. Singapore's enactment of a Foreign Interference Countermeasures Act (FICA) is a sign that the government recognizes the risk. But there is a limit to what a law can block. Severing a cultural bond shared by three quarters of the population by statute is not a problem in the domain of institutions. It is a problem in the domain of identity.
7. Capital Only Rents Trust
Hong Kong's failure compresses the core lesson of this chapter.
In the second half of the twentieth century, Hong Kong was the most successful case of the in-between strategy. British colonial rule of law, geographic and cultural proximity to the Chinese mainland, and the function of Asia's financial hub. Between East and West, between Communism and capitalism — Hong Kong created a value nothing else could match.
Three assets made Hong Kong's in-between possible. Institutional trust (common law and an independent judiciary). Geographic and cultural access (the Chinese language, proximity to the mainland). Free movement of capital (no foreign-exchange controls, low tax rates). The 2020 National Security Law destroyed the first asset — institutional trust.
But that destruction did not happen overnight. It was the product of a twenty-year erosion. 2003: Hong Kong's own attempt to enact Article 23, a local national-security law. Half a million people marched in the streets, and the bill was withdrawn. 2014: the Umbrella Revolution. Protesters demanding universal suffrage occupied the downtown for 79 days. 2019: the anti-extradition-bill protests. This was when foreign firms in Hong Kong first began drawing up exit scenarios. June 2020: the National Security Law. March 2024: Hong Kong's own Article 23 — the very bill that had been withdrawn in 2003, reintroduced twenty years later and passed with penalty provisions broader than those of the 2020 NSL. For twenty years, geopolitical pressure steadily eroded the institutional space. It did not collapse at a single moment. The cracks accumulated and accumulated until they crossed a threshold.
Apply the core formula of this book to Hong Kong, and a reverse chain appears. Hong Kong achieved the concentration of capital — Asia's largest financial hub. But its answer to social instability, the 2019 protests, was not institutional redesign. It was institutional suppression: the 2020 NSL. That choice destroyed the institutional trust that is the precondition for concentrating capital in the first place, and capital departed. The formula of the series —
Technological innovation → Concentration of capital → Social instability → Institutional redesign
— run in reverse, became: institutional suppression → trust collapse → capital flight → weakened hub status.
Hong Kong still has a 16.5 percent corporate tax rate, a 0 percent capital-gains tax, and a top-three ranking on the Global Financial Centres Index. Its physical infrastructure, financial technology, and professional talent are still at world-class levels. And capital departed anyway. Because the neutrality of the institutions had been compromised. No matter how strong the outward measures of competitiveness are, the moment political power crosses the boundary of the rule of law, the value of the between collapses.
Capital does not purchase institutional trust. It only rents it. And a rental agreement can be cancelled overnight.
Singapore, at the same fork in the road, took a different path. It maintained authoritarian control while refining the design of its economic institutions so carefully that capital's trust held. This is where Lee Kuan Yew's inheritance lies. The true toxin of the poisonous shrimp was not military force — it was institutional credibility.
Korea in the Mirror
The face of Korea that Singapore reflects is ironic.
A core without a hub. What Singapore lacks, Korea has. A domestic market of 52 million, world-class semiconductors, batteries, and shipbuilding, the potential to develop AI on its own, and the soft power of K-culture. If Singapore is the country that is a hub without a core, Korea is the country that is a core that cannot run a hub. Seoul's ambition to become an Asian financial hub has been frustrated repeatedly. Foreign-exchange controls, a bank-centered financial system, a high corporate tax rate, rigid regulation, and limited English-language access. During the Roh Moo-hyun administration, on the day the president announced the financial-hub strategy, the head of a major commercial bank publicly declared that a ceiling should be set on foreign investment — a single anecdote that has become a symbol of the structural contradiction.
The speed of regulation. In Korea, the AI Basic Act (인공지능 기본법) took three and a half years to pass the National Assembly. Tada was banned under pressure from the taxi industry. Telemedicine was allowed during the pandemic and blocked again the moment COVID receded. The gap between Singapore's institutional speed and Korea's regulatory rigidity is not merely an administrative difference. It is a difference in a nation's capacity to execute a strategy. In a country where the name of the AI strategy and the allocation of its budget change every time the administration changes, five-to-ten-year strategic execution is structurally close to impossible.
The other side of Seoul's GFCI rank of 10. On the 2024 Global Financial Centres Index, Seoul ranked 10th in the world and 8th in fintech. It finished ahead of Tokyo (15th) and Paris (18th). On the rankings alone, it is an encouraging result. But the gap in foreign direct investment tells a different story. Singapore at $192 billion (39 percent of GDP) and Korea at $15.1 billion (1 to 2 percent of GDP): a gap of thirteen to one. While Singapore hosts 132 banks (126 of them foreign), Korea has 19 commercial banks. Multinational regional headquarters: Singapore 4,200, Seoul roughly 100. The distance between ranking and reality is wide.
The warning of the martial-law moment. The 2024 martial-law incident in Korea showed that Hong Kong's lesson is not abstract possibility. When a political crisis damages trust in institutions, the shock propagates into capital and talent flight far faster than any competitiveness index moves. Hong Kong has walked that path. For Korea to put an in-between strategy into practice, institutional neutrality — an independent judiciary, predictable regulation, and financial supervision free of political power — is a precondition. Under any geopolitical pressure, judicial independence and equality before the law must not be compromised. Non-negotiable rule of law. This is the most essential lesson that Hong Kong and Singapore leave for Korea.
But Korea does not have to become Singapore. Singapore is a country that must survive as a hub without a core. Korea has a core. Semiconductors, batteries, defense, K-culture. Korea's true task is not to imitate Singapore but to design a Korea-specific in-between strategy that adds financial function, AI governance, and neutral-platform capacity on top of its manufacturing power.
The next mirror reflects a country whose indispensability was built by a single firm — the Netherlands and ASML.