Follow the Cash: The Commitment That Moves None of It

On 8 September two filings reached the Seoul market within two hours of each other. Read separately they are routine. Read together they describe the same transaction wearing two different accounting costumes — and only one of the costumes is visible in a cash flow statement.

The visible one. SFA Semicon, a semiconductor back-end packaging and test firm counting Samsung Electronics among its customers, disclosed a $75m loan to SFA Semicon Philippines Corp., a wholly owned subsidiary. In won that is ₩100.7bn, or 21.3% of the parent's equity. The loan carries 4.6% interest and runs from 15 September 2026 to 14 September 2029, drawable in tranches over the coming year. Including it, the balance owed by the Philippine unit reaches ₩130.2bn, roughly 27.5% of equity. The unit posted ₩223.8bn of revenue and a ₩22.5bn net loss last year. The stated purpose is facility investment and working capital.

Every element of that is trackable. It appears as an investing outflow; it sits on the balance sheet as a receivable from a related party; its recovery is observable quarter by quarter. It is a large exposure, but it is a counted one.

The invisible one. The same afternoon, Hyosung Chemical's semiannual report disclosed that Ernst & Young Vietnam, while issuing an unqualified opinion on Hyosung Vina Chemicals, had flagged material uncertainty about the unit's ability to continue as a going concern. The cited grounds: accumulated losses of 20.798 trillion dong (about ₩1,218.8bn) and current liabilities exceeding current assets by 6.482 trillion dong. The unit began commercial polypropylene production in April 2020 and lost ₩231.1bn last year, though it turned a ₩45.2bn first-half profit this year on ₩675.2bn of revenue.

The parent's support runs to ₩861.8bn in debt guarantees and a ₩204.2bn long-term loan at 6.62%. In July it extended a $60m guarantee by another year. Add the two instruments — ₩861.8bn + ₩204.2bn — and the parent has ₩1,066.0bn riding on that single unit. Under a fifth of it has ever moved through a cash flow line.

Why the distinction is not cosmetic. Cash governance is usually taught as follow the money: trace what a company raised and see where it went. That works when money moves. A guarantee is the case where it does not. It creates no outflow, so it never enters operating, investing or financing activities. It lives in the commitments-and-contingencies note as a contingent liability.

But the obligation is real, and it has a property the loan does not: the parent does not control its timing. A loan can be rolled, extended or restructured on the lender's own calendar. A guarantee converts to cash on the day the borrower fails to pay. The larger half of this exposure is therefore both the least visible and the least controllable — and a governance score built only on statements will systematically understate it.

A third form, for contrast. Also on 8 September, Bloomberg reported China's Ministry of Finance will issue special treasury bonds to inject 300 billion yuan into eight state financial institutions — Agricultural Bank raising up to 160bn (130bn subscribed by the ministry), ICBC 100bn (70bn), PICC's entire 15bn, and five others — bringing injections since early 2025 to 500bn yuan. Here cash genuinely moves, and the exchange filings name the amount, the subscriber and the instrument. The cost simply relocates: Bloomberg Intelligence estimated annualised EPS dilution of 3.5% at ICBC and 6.3% at Agricultural Bank.

Guarantee, loan, equity injection. These are not a ladder of commitment strength. They are a map of where the cost surfaces — in the notes, in the cash flow statement, or in earnings per share. Read only one of the three and you will conclude the wrong company is exposed.

The academic frame. Jensen's free cash flow argument (1986, American Economic Review 76(2)) explains why managers prefer deploying cash internally over returning it; Johnson, La Porta, Lopez-de-Silanes and Shleifer's "Tunneling" (2000, American Economic Review 90(2)) documents how intragroup transfers move value in ways minority shareholders struggle to observe. Neither paper is about these two companies, and nothing here suggests either did anything improper. What the literature establishes is narrower and sufficient: intragroup resource transfers are exactly the class of transaction where the observable record and the economic reality come apart most easily.

Which raises the reader's question. When you last checked a company's cash position, did you open the notes — or only the statements?

General structural observations from public filings. Not investment advice.

#RaymondsRisk #RelationalRisk #CorporateGovernance #ContingentLiabilities #IntragroupFunding #OffshoreSubsidiaries