The Zombie Pattern: How Distressed Companies Drain Before They Fall

On 1 July, Korea's exchange switched on a tougher delisting regime. A KOSDAQ company whose market value stays under ₩20bn for 30 consecutive trading sessions goes on the watchlist; fail to hold the line for 45 sessions out of the next 90 and the delisting process begins. Kim Seong-cheon of the Korea Exchange told the KOSDAQ 30th-anniversary event on 2 July that roughly 50 companies are expected to become delisting candidates on the market-cap test alone, with the first designation likely in August. The threshold rises again in January, to ₩30bn.

The reform is sensible, and the design detail is the interesting part: the exchange deliberately made it harder to leave the watchlist than to enter it. As Kim put it, the entry bar is similar to before, but the exit bar is much higher. That is an explicit attempt to stop distressed issuers from parking in purgatory for years.

Still, notice what the trigger actually observes. Price.

What the price knew last, and what the filings knew first

A Korea Capital Market Institute report published on 30 June puts a number on how long the evidence sat in public. A quarter of KOSDAQ companies posted R&D at effectively 0% of sales continuously from 2000 to 2020. In 2025 that same cohort reached 0.02%. Over the same window the top decile moved the opposite way: 7.41% of sales in 2000, 9.49% in 2015, 11.85% in 2020, 20.54% in 2025.

Profitability tracks the split. The bottom decile's average operating margin fell from -31.08% (2000–2004) to -48.45% (2021–2025), while the top decile edged up from 15.94% to 16.75%. KCMI's conclusion is that KOSDAQ has become two markets wearing one name, and that a segment system — premium, standard, watchlist — is the reasonable response.

Read that as a timing problem and it becomes something sharper. The delisting screen looks at 30 sessions. The disclosure that these companies had stopped investing in themselves was continuously available for two decades.

What RII actually reads

Reinvestment Intensity Index is often described as "does the company reinvest." That is not quite it, and the imprecision matters.

RII reads three things together: the reinvestment rate, the coefficient of variation in capex, and the divergence between investment and the operating base that should be driving it. The most useful signal is rarely a zero. It is incoherence.

Two patterns recur. First, capex persistently below depreciation. That is not capital discipline; it is liquidation on an instalment plan, and it is nearly invisible in the headline numbers, because depreciation is a non-cash charge that props up operating cash flow exactly while the asset base shrinks. A company can post improving operating cash flow for years while quietly consuming itself.

Second, a low reinvestment rate paired with a high capex coefficient of variation. Spending that arrives in isolated lumps — around a project announcement, or shortly after a capital raise — and then stops is event-driven, not operational. Money goes in and does not come back as capacity.

The same lag, a different market

The US offers the control case. Large-company bankruptcies reached 372 in the first half of 2026 on S&P Global Market Intelligence data — the highest January-June total since 2010, a fourth consecutive annual increase, and more in six months than in all of 2022 (317). Industrials led with 50 filings, consumer discretionary 35, healthcare 26. ABI and Epiq AACER put total filings at 310,550, up 12%, with commercial Chapter 11 up 28%.

And credit markets barely moved. Ample private capital absorbed the losses, so the price signal never fired.

Why this is an old finding

The academic literature settled this a while ago. Caballero, Hoshi and Kashyap (American Economic Review, 2008) showed that Japan's zombie firms did not merely survive — they depressed restructuring and investment across their entire sector. Acharya, Eisert, Eufinger and Hirsch (Review of Financial Studies, 2019) found the same congestion effect in Europe when cheap credit kept impaired borrowers alive. Banerjee and Hofmann (BIS Quarterly Review, 2018) documented the secular rise in zombie share across advanced economies and linked it to falling productivity.

The common thread is that the damage is done during the survival, not at the exit. By the time a price threshold or a bankruptcy docket registers the failure, the capital has already been spent — or not spent, which turns out to be the same thing.

So the practical question for an individual investor isn't which names will trip the ₩20bn line in August. That list will be public. It's this: among the companies that clear it comfortably today, which ones have already stopped reinvesting?

General market observation based on published filings and research; not investment advice, and no individual issuer is assessed here.

#RaymondsRisk #RelationalRisk #CorporateGovernance #ZombieFirms #KOSDAQ #ReinvestmentIntensity