The Forest That Redirected the Water

If you remember Japan as a country living off exports, that memory stopped 20 years ago.

Counterexamples exist on both sides. At the yen's record strength in 2011 (JPY 79.81), Japan posted its first trade deficit. When the yen strengthened in 2016 (JPY 108.79), a +JPY 5.5 trillion surplus returned. The correlation over the 15 years from 2011 is −0.14. The weak-yen/trade-surplus link has effectively broken.

What a weaker yen clearly inflates instead is income from assets planted abroad. It arrives in foreign currencies, so a weaker yen increases its yen value. From 2011 onward, the correlation is +0.94.

The lost trade surplus and rising primary income are not separate events. They are two sides of the same corporate behavior: Japanese manufacturers moved factories abroad. The Bank of Korea diagnosis says exactly this: export-volume growth weakened because of more overseas production.

When a factory leaves, two things happen. Less is made and sold at home, reducing the trade surplus. Holdings in foreign subsidiaries instead produce dividends and retained earnings. The balance-of-payments ledger calls that direct investment income.

Over 30 years, direct investment income grew 17.1x, and its share of investment income changed to 25% → 63%. In 1996, Japan was a country that bought overseas bonds and earned interest(portfolio investment 70.8%). In 2025, Japan owns companies abroad and receives its share.

The first answer was ROE. In August 2014, the Ito Review from a METI project set a number.

There was a basis for 8%. In the same report, global institutional investors' average expected cost of capital for Japanese companies exceeded 7%. ROE above 8% would exceed the cost of capital for roughly 90% of investors. Thus, in 2014 the framework was already return on capital relative to its cost.

The Stewardship Code arrived that year, followed by the Corporate Governance Code the next. The yardstick and the pressure appeared together.

The scorecard 8 years later was sobering.

The second answer came in 2023. The TSE sent all listed companies its request, Action to Implement Management that is Conscious of Cost of Capital and Stock Price. It diagnosed:

The same 8% from 10 years earlier reappeared, but this time alongside P/B, rather than alone. The document also explained why P/B:

But there is a trap. The exchange's own words reveal it: P/B below 1x points to two different causes.

Even a profitable company can trade below 1x P/B. If earnings remain trapped inside and the market distrusts who controls the decision, prices are held down by a trust problem rather than profitability. Fixing the business alone cannot solve it.

P/B below 1x fell to 50% → 27%, roughly halving; ROE below 8% moved to 47% → 43%, barely changing. The improvement differs by roughly 6x.

There is one interpretation. Over the past 4 years, Japanese companies did not become dramatically better earners. Earning power barely changed, but the market began seeing the same companies differently. The second row of the table — Trust — moved.

Cross-held shares mean more than their ownership percentage. They are shareholders who will not sell, whatever management does. For managers, they were a shield against anyone taking over the company, however low the price.

When the shield dissolves, the calculation changes. Low P/B becomes an M&A target signal rather than a badge of shame. Managing the share price becomes a survival issue rather than a matter of face. Disclosure pressure extends to the use of sale proceeds. Tokio Marine answered: growth M&A.

Cross-shareholding unwinding did not begin in the 2020s. Operating companies' share fell from its FY1990 level to 30.1% → 17.7%, and financial institutions to 43.0% → 27.3%, over 35 years. The past 3 years' KRW 91 trillion of sales are the final phase of that long decline, accelerated by institutions and inflation.

Retail investors did not fill the gap. Foreign corporations rose from 4.7% to 34.7%, gaining 30 percentage points. This is a survey record. They first exceeded 30% in FY2013.

In deflation, cash was the best asset. Even deposits yielding 0% earn a positive real return when prices fall. No campaign could move household money in that world: there was no reason to move.

Inflation changed the conditions for the first time.

Keep the sequence straight. The new engine was not built because yen carry was blocked. Inflation produced both ① BOJ rate normalization (ending the old engine) and ② household money movement (igniting the new engine) at once: two effects of one cause.

If this is true, government behavior should change. Under the old engine, yen weakness was a subsidy. Under the new engine, it is a loss of household purchasing power. Which side gets protected is revealed by money, rather than words.

The key is not the level, but Speed. Nomura sees a 1.5% BOJ policy rate; markets price 2%. Close that gap at once and the old engine stops abruptly.

Last fiscal year's trade surplus of +JPY 1.3631 trillion was led by semiconductor-equipment exports. The first two months of the new fiscal year maintained a surplus of JPY +402.6 billion, but that cannot compare with primary income (over the same two months, +JPY 8.4856 trillion). Statistically, trade's era of feeding the country is over. The current account is now driven by Primary income.

Korea moved too. The share of companies reducing their share count rose to 5.7%(2019) → 29.8%(2025)(US 41.7%), while annual buybacks rose from KRW 4.2 trillion to KRW 14.8 trillion. Yet total payout is 1.60%, last of five forests.