The Zombie Pattern: Why Reinvestment Intensity Stops Telling You Anything the Day the Plan Lands

Late on 16 September, a KOSPI-listed company filed the arithmetic of its own rehabilitation. STX disclosed a capital reduction of 99.13%. Paid-in capital drops from KRW 426.4bn to KRW 3.7bn. Every 103 shares become one. The controlling shareholder's 11,995,657 common shares, together with 5,701,959 shares previously issued on a debt-for-equity conversion to a related party, are cancelled outright and without compensation. The record date is 6 October.

Two further filings complete the picture. To convert rehabilitation claims, 139,528,385 new common shares will be issued to third parties at KRW 2,500 each, payable on the 21st. Separately, 27.2 million shares will be issued to improve the capital structure through an M&A, raising KRW 68.0bn — of which KRW 67.0bn goes to repaying debt. Payment on 12 October; listing on 6 November.

What reinvestment intensity actually assumes

Reinvestment Intensity Index reads three things: the reinvestment rate, the coefficient of variation in capex, and the investment gap. Only the first matters here.

The reinvestment rate asks what proportion of available capital went back into the operating asset base. A low reading is treated as a warning, and reasonably so. But that reading rests on an assumption that is almost never stated: that the allocation was a decision the company made. The number is informative because it is the residue of a choice.

A rehabilitation plan dissolves that assumption. In the filings above, the allocation is set by a court-approved plan and the creditors standing behind it. The reduction ratio, the share consolidation, the conversion price, the use of proceeds — all of it is determined outside the company. The reinvestment rate will print near zero. It is no longer measuring a decision. It is recording that there is none left to measure.

This is not a criticism of the metric. It is a statement about its window. The years in which reinvestment intensity carries information are the years when capital allocation still sits inside the firm — which is to say, the quiet years before anything is filed under this heading at all.

The same absence, in a different language

The point travels. On 14 July 2026, Inotiv, Inc. (NASDAQ: NOTV) announced that the U.S. Bankruptcy Court for the Southern District of Texas had confirmed its Plan of Reorganization. The release describes a "significantly strengthened balance sheet," existing stakeholders who "collectively manage more than $60 billion in assets" and who "provided additional capital," and operations continuing "as usual."

Four claims, and every one of them concerns the right-hand side of the balance sheet or the continuity of operations. There is no figure for what the additional capital funds inside the business. The Korean filing writes "debt repayment" into the use-of-proceeds line; the American release does not fill in a use-of-proceeds line at all. Both documents leave the reinvestment cell empty. The language of restructuring is the language of claims, not of operations.

A third filing from the same Korean disclosure roundup shows the order in which these things get recorded. A KOSDAQ-listed holding company reported that a subsidiary — an optical-film maker for displays — had resolved to dissolve at an extraordinary general meeting. Its total equity at the end of last year was already negative KRW 3.835bn. The dissolution is the entry in the record. The years in which reinvestment stopped are not an entry anywhere.

Korea parallel

This is why RaymondsIndex builds its readings on relational and allocation signals across the listed universe rather than on the terminal filings. Terminal filings are precise, timely, and almost useless as leading indicators, because by construction they arrive after the decision rights have moved. The useful window is upstream, in the years when capex is drifting, when the investment gap is widening quietly, and when nothing has yet been named.

Academic frame

The mechanism has a literature. Caballero, Hoshi and Kashyap (2008), "Zombie Lending and Depressed Restructuring in Japan," American Economic Review 98(5), documented how firms kept alive by forbearance depress investment and restructuring across an entire sector — the damage accumulates in the years before any formal proceeding. Banerjee and Hofmann (2018), "The rise of zombie firms: causes and consequences," BIS Quarterly Review (September 2018), traced the same pattern across advanced economies and found that zombie firms' defining trait is persistence, not sudden failure. Both point the same direction: the informative interval is long, quiet, and prior.

For the individual holder

When a rehabilitation plan is published, the numbers finally become easy to read. The reduction ratio is exact. The conversion price is exact. The use of proceeds is exact. That legibility is what makes the moment feel like information.

It is worth asking what the same company's reinvestment rate looked like in the years when nobody was filing anything dramatic — and whether that was ever available to you at the time.

#RaymondsRisk #RelationalRisk #CorporateGovernance #CapitalReduction #DebtForEquity #ZombieFirms