Credit Mispricing In Japan: Repurposing the Zombie-Firm Literature for the Nikkei 225
Series
SSRN Working Paper Series; 6813479
Type
working paper
Date Issued
2026
Author(s)
Abstract
The academic zombie-firm literature counts distressed firms and measures their macroeconomic costs. It pays far less attention to what happens next. On the Tokyo Stock Exchange, classified firms rarely exit, and most recover. This paper turns that recovery gap into an investment strategy. We convert ten published zombie classifications into continuous credit-quality scores and form equally weighted portfolios of the most zombie-like decile of the Nikkei 225, rebalanced annually under a strict point-in-time protocol. Over 2000–2025, the three headline classifications deliver compound annual returns of 7.2 to 10.2 percent against 3.8 percent for the benchmark, net of a 50 basis point transaction-cost charge. Spanning tests against identically constructed value portfolios isolate the credit component: the coverage-based classifications retain 2.3 to 3.0 percentage points per annum over the value placebos, with a joint estimate of 2.7 points (t = 1.8). This premium is economically meaningful, statistically moderate on its own, and doubly concentrated: it sits precisely where classified firms show no elevated exit risk, and it is confined to the bank-connected blue-chip segment. On the broad TOPIX universe the same deciles are fully spanned by generic value. Both patterns are the signature of credit mispricing rather than compensation for default. The signal remains investable even in years when binary zombie counts are low.
Language
English
Keywords
zombie firms
credit mispricing
value investing
Japanese equities
TOPIX
Nikkei 225
Carhart factors
Schumpeter
HSG Classification
contribution to scientific community
Refereed
No
Official URL
Subject(s)
Division(s)
Additional Information
JEL classification: G11, G12, G14, G32
File(s)![Thumbnail Image]()
Name
Credit_Mispricing_Japan_Nikkei225_v5.8.pdf
Type
Main Article
Size
877.75 KB
Format
Adobe PDF
Checksum (MD5)
d23af10761983856602d7b4ad57b1e16