Korean real estate trust firms have lent 9.5 trillion won of their own money to troubled building projects

Set against the loan line, the reserve line makes the trust firms' first half look worse than the loan growth alone suggests. The balance rose by 543.9 billion won ^[1] and reserves against it rose by 327.1 billion won ^[2]. For every 100 won of new lending, reserves grew by about 60 won ^[3]. They now equal about 30.8% of the reported balance, up from about 29.0% at the end of last year ^[4].

The second-quarter loss was a swing of 234.5 billion won from the first quarter's profit ^[5]. The report attributes it to money tied up for long periods as the firms paid construction costs through delays and unsold units, mostly at land trust projects with completion guarantees ^[4].

Builders are passing losses along because their cash is draining faster than their reported profits show. At 12 mid-sized builders ranked 30th to 100th, cash fell 13.2% to 1.1057 trillion won in a year while short-term borrowings rose 4.4% to 840.3 billion won ^[10]. Cash net of short-term debt went from 468.0 billion won to 265.4 billion won, a 43% drop ^[7]. Unbilled receivables rose 12.2% to 787 billion won ^[12]. Taewang E&C, the extreme case, had a few hundred million won of cash left by the end of last month, according to reports ^[9].

The step from a builders' slump to a financial-sector credit problem has less evidence behind it. The report does not include the trust firms' capital, how they fund their own lending or who owns them. On the figures given, the losses have moved from builders onto 14 companies whose combined first-half deficit was 82.7 billion won, against a loan book of 9.5 trillion won ^[3][1].

The first-half figures fit more than one outcome. Provincial presales could recover, letting sites complete and the trust loans be repaid, with some reserves written back. Sites could stay stuck while the firms keep lending. At the first-half pace the balance passes 10 trillion won before the year is out, because the 483.5 billion won still needed is less than the 543.9 billion won added in the first six months ^[8]. Or the firms could stop funding weak sites and book the losses at once, pushing more projects and more builders toward receivership.

I think the second path is the likeliest. The construction industry expects clearing troubled sites to take considerable time, with provincial housing and non-apartment markets recovering slowly and construction and financing costs still high ^[13]. The report places all of it in the lingering aftermath of the project-finance troubles that surfaced after the 2022 Legoland default ^[14].

The same reserve line supports the opposite reading. Reserves grew roughly twice as fast as loans, 12.6% against 6.1% ^[9], and a lender whose reserves outgrow its loans is recognising losses early. The view is wrong if the second-half figures show the loan balance flat or falling and reserve growth dropping below loan growth.