Three Forests Answer

All three countries earn dollars through trade. What happens next differs. Over the latest 12 months of balance-of-payments data, Korea earned a current-account surplus of $266.2bn and sent $117.7bn abroad through resident overseas portfolio investment — an offset ratio of 44%. Nearly half its earned dollars become foreign assets. Japan earned a $227.7bn surplus while overseas portfolio investment showed +$19.2bn of net inflows. Money is returning rather than leaving. China's corresponding statistic is not meaningful: the channel for households to buy foreign securities is closed.

This difference sums up the strategies. Korea sends money out, Japan turned it back, and China prohibits the outflow at source. Before examining each, look at their biggest footprint: exchange rates.

During 2021–24, when the won and yen fell sharply, exchange rates per dollar rose +52.1% for the yen, +35.3% for the won and +11.7% for the yuan(our daily measurements). In currency-value terms, the yen fell −34%, the won −26% and the yuan −10%. The yuan fell far less.

That is unusual. Interest-rate gaps should have made the yuan the weakest. From 2022, Chinese rates were below US rates, with deflation on top. Carry logic makes it a candidate for the weakest currency. Yet it fell least. The §01 flows explain what carry cannot: only currencies with structural channels that leak household savings into US stocks — overseas retail investing, NISA and institutional foreign investment — weakened structurally. The presence or absence of an outflow channel split their paths. This cross-checks the absorption-mechanism hypothesis through exchange rates.

In #07 we wrote that Japan had caught the shareholder-payout leader “to the second decimal place” (translated). In #10 we found that half the convergence came from the throne lowering itself. The other half — Japan's own rise — is Japan's answer.

The defensive tools are the institutions seen in #08: share buybacks of a record JPY 17 trillion(2026); the Tokyo Stock Exchange campaign to improve companies below 1x P/B — over half of TOPIX disclosed their cost of capital (2026-05); and faster unwinding of strategic shareholdings. The results are the Nikkei near record highs and the §01 observation: resident overseas portfolio investment turned to +$19.2bn of net inflows. Money that was leaving is returning — the firmest evidence that restoring the domestic market's appeal works.

Japan's strategy does not block outflows. Pension funds and insurers retain their accumulated US assets (#08's capital-exporter structure), and some NISA household savings still enter US funds (allocation data unavailable; qualitative judgment). Keep earning from existing US assets while restoring the domestic market's appeal so new money has less reason to leave — Japan pursues both together.

Korea's answer is institutional catch-up. Value-up plan disclosures by 177 companies; the 1st Commercial Act amendment (directors' duty of loyalty to all shareholders), the 2nd (mandatory cumulative voting), and the 3rd (mandatory treasury-share cancellation + separate taxation of dividend income); SK hynix's KRW 12.2 trillion cancellation and Samsung Electronics' KRW 6.1 trillion of treasury shares. The direction clearly follows Japan.

The grade is mixed. The KOSPI broke through 5000 and 6000, but July's −31% exposed the cycle behind the rise (#09 — the broken branch). The total shareholder payout yield remains 1.61%, lowest of 4 markets, with the structural discount to an Asian basket unchanged at −14 percentage points. And the §01 figure: 44% of the surplus leaves for overseas securities. The rules follow Japan while the money moves the other way.

This is where we stand. The capital account is open — complete freedom to leave — while the payout yield is the lowest, offering little reason to stay. For US equities, Korea is an ideal funding source. Overseas-stock custody holdings of KRW 256 trillion show it. If the hypothesis in #10 §05 is right, that KRW 256 trillion sits where it absorbs losses when US stocks turn down. Actual implementation of the 3rd Commercial Act amendment — whether the 1.61% payout yield moves — is the real test of Korea's strategy.

A new scene appeared on August 10. The KOSDAQ surged +6.97%, led by biotech, semiconductor equipment and batteries while Samsung Electronics and SK hynix were flat. The flows were unusual: KOSDAQ foreigners turned net buyers by +KRW 103.1 billion, and of institutions' +KRW 566.1 billion, investment trusts supplied +KRW 169 billion and pension funds +KRW 82.3 billion. Institutional flows described as purely mechanical throughout August saw directional capital appear for the first time. On the same day, foreigners sold −KRW 1.49 trillion in KOSPI stocks. Rather than leaving Korea, they were closer to “leaving the two stocks accounting for 94% of the concentration and moving into the chains benefiting from that investment cycle” (translated). The three rising groups — equipment, ESS and biotech — sit precisely where expanding US AI investment spills over.

This rotation adds depth to the argument rather than overturning it. Korea supplies funding for US AI investment and also benefits from that investment. Outward money (retail overseas investing) and circulating money (rotation) are two faces of the same cycle. One footnote: amplifiers produced much of the rally's size — daily KOSDAQ150 leveraged ETF turnover of KRW 1.15 trillion (3rd among all ETFs), and margin loans rising for three consecutive days to KRW 29.2 trillion. The catalyst set the direction; the machinery set the size. A rebound arriving with debt has been on our watch list since #03.

China's answer has two moves. Block: the household overseas-securities channel is closed (§01/§02), and the country is also stepping out of the absorption layer. US Treasury holdings are far below their peak, while the PBOC bought gold for 20 consecutive months, accumulating 2,346 tonnes. Replicate: steer household savings displaced from property into A-shares (CXMT IPO $8.6bn, first day +466%); have the central bank fund share buybacks through relending (a CNY 300bn facility); and have the securities regulator effectively require dividends.

The results appear in our measurements. Total A-share dividends rose from CNY 1.13 trillion in 2018 to a 2025 total of CNY 2.63 trillion — 2.3x in 7 years, increasing for 8 consecutive years. The dividend yield for all A-shares is 2.42% — nearly twice the US dividend yield of 1.28%.

The structure mirrors the US. US payouts center on share buybacks (buybacks 1.78% + dividends 1.28%): payouts that are easy to stop, and were stopped this year. China centers on dividends (dividends 2.42% + buybacks 0.11%), enforced by policy: payouts that are hard to stop.

In one line: The US pays with stories; China pays with cash. The state cannot sell a Musk-style growth story, so it pays cash to keep savings inside the domestic mechanism. One footnote: at least 27.9% of that cash returns to the treasury through state-owned stakes(our measurements of the top 20 companies, a lower bound). These are shareholder payouts and fiscal recycling at once.

Put the three answers together and a pair of scissors appears. Water enters the US even as payouts fall (total payout halved to 3.06%, a turn to net issuance — yet record inflows of $901.8bn). Asia must raise payouts to keep the water: Japan's JPY 17 trillion, Korea's 3rd Commercial Act amendment and China's dividend mandate. The scissors of privilege and cost.

The same scissors open on the institutional axis. US companies leave Delaware and move beyond enforcement of directors' duty of loyalty (Tesla led the move to Texas), while Korea changes its Commercial Act to extend that duty to all shareholders. Governance standards themselves, as well as payout yields, move in opposite directions.

Give this cultural divide a name: the birth of the story shareholder. Tesla demonstrated a $1 trillion market cap without dividends. A contract paying shareholders in roadmaps rather than cash flow became a norm. Alphabet is its latest descendant: $0 in share buybacks and $49.6bn of new equity issuance in the same quarter. It did more than stop returning money; it started collecting it (#10). Markets that can sell a story can stop distributing fruit. Markets that cannot must offer more. The two blades ultimately differ in what they pay with.

Exchange rates have already graded outflow prevention: prohibition (China) won, reversal (Japan) worked, and Korea paid in currency value. But all three strategies fail to prevent co-movement. July demonstrated it: US −4.86%, Korea −21.18% (β 1.469), Japan −6.86%. All three responses address outflows; they do not prevent markets from shaking alongside US stocks.

The real test is still ahead. The two pillars of #10 — water (liquidity) and proof of fruit (monetization) — face overlapping stress in Q4 2026–2027 window. That is the test. If the window opens, the difference among the three answers will be whose balance sheet is left holding the losses, rather than different growth rates.

The day US-bound money begins asking about returns — #10's “day the water asks about fruit” — has yet to come. Before it does, we must know exactly where our money sits in this structure.