> ## Content Index > Fetch the complete content index at: https://tokyobrainie.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # Deep Cut 007｜120兆円は危機の証明ではない——地銀の不動産融資と「バブル期の2倍」の読み方｜¥120 Trillion Is Not a Crisis Verdict - URL: https://tokyobrainie.ghost.io/news-cutting-room-3fdbc361-dae9f036-en/ - Published: 2026-10-22T15:00:00.000Z - Updated: 2026-11-11T00:15:02.000Z - Description: ¥120 Trillion Is Not a Crisis Verdict Nikkei reported that Japanese banks’ outstanding loans to the real-estate industry reached ¥120.7588 trillion at the end of Jun - Author: TokyoBrainie - Tags: News Cutting Room, Tokyo Dispatch, English, #Import 2026-11-11 08:54 [English](https://tokyobrainie.ghost.io/news-cutting-room-3fdbc361-dae9f036-en/) · [日本語](https://tokyobrainie.ghost.io/news-cutting-room-3fdbc361-dae9f036-ja/) ## ¥120 Trillion Is Not a Crisis Verdict Nikkei reported that Japanese banks’ outstanding loans to the real-estate industry reached ¥120.7588 trillion at the end of June 2026, up 8.2% from a year earlier and equal to 18% of total bank lending. Regional banks added roughly ¥3.1 trillion over twelve months, the largest increase in the available series. Some lenders now show unusually high property exposure, while banks headquartered outside Tokyo finance a large share of projects in the capital. Those are important facts. They justify supervision. They do not, by themselves, establish a property bubble, deteriorating underwriting or an imminent banking crisis. ### What the article gets right - **Concentration is rising.** A national aggregate can look safe while a few banks, geographies or property types become fragile. - **Cross-regional lending creates an information problem.** A regional bank searching for yield in central Tokyo may know less about the borrower, exit liquidity and local micro-market than a lender operating there continuously. - **Higher rates matter with a delay.** Refinancing risk appears when loans reset or mature, not necessarily when the policy rate first rises. - **Low defaults are not a leading indicator.** A 0.39% real-estate default rate is reassuring about present credit performance, but it cannot rule out a future correction. - **Property funds can transmit a shock quickly.** Foreign-fund selling, cap-rate expansion and thinner transaction liquidity can weaken collateral values before bank defaults appear. ### Where the framing overreaches The headline comparison—18% of lending today versus roughly 10% in the bubble era—is arithmetically valid but economically incomplete. First, the denominator has changed. Japan’s industrial structure, corporate cash holdings, manufacturing loan demand, bank regulation and capital-market financing are not what they were around 1990\. A larger share of a different loan book is not automatically a larger loss risk. Second, outstanding loans are a stock, not an underwriting score. The stock can rise because land and construction costs rose, transactions expanded, refinancing volumes increased, or credit standards weakened. Only the last explanation directly supports a “loose lending” thesis. Third, the ratio reveals exposure, not loss severity. For that, investors need origination LTV, DSCR and interest-coverage distributions, maturity walls, fixed-versus-floating rates, borrower equity, collateral haircuts, provisioning and bank capital. **Eighteen percent is an exposure ratio, not a loss ratio.** ### The most important evidence is buried inside the warning The article itself reports that Japan’s financial authorities do not currently see an immediate systemic threat. The Bank of Japan’s April 2026 Financial System Report says the system is stable overall, capital and funding bases are robust, and banks remain above regulatory capital requirements even in a severe property-price stress resembling the pre-Lehman “mini-bubble” collapse. It also says real-estate lending has grown faster than lending to other industries and therefore requires composition-level monitoring.[Bank of Japan, Financial System Report, April 2026](https://www.boj.or.jp/research/brp/fsr/fsr260421.htm?ref=tokyobrainie.ghost.io) The article also reports a July real-estate default rate of 0.39%, below the 1.22% rate for all sectors. Current credit performance is therefore not consistent with a crisis already underway. At the same time, low defaults can coexist with excessive new risk because defaults lag lending conditions. This makes the correct conclusion narrower than the headline atmosphere: > **Japan has a concentration and underwriting-monitoring problem, not evidence of a present systemic property crisis.** ### Why “lending is growing” is not an acquittal It is tempting to laugh at crisis warnings while credit and transaction volumes are expanding. The laughter is useful only if it targets bad inference. Credit booms, by definition, often look healthy before they turn. More lending can support transactions and push prices higher; higher collateral values can then justify still more lending. That feedback loop is precisely why supervisors watch growth. But growth alone is ambiguous. The correct chain is not “loans rose, therefore safe,” or “loans rose, therefore bubble.” It is: **loan growth → underwriting quality → property cash flow → refinancing terms → collateral value → expected loss → bank capital** If a report jumps from the first box to the last, it has described a quantity, not measured a crisis. ### The 1990 comparison needs five normalizations | Comparison | Why it matters | | ---------------------------------- | ------------------------------------------------------------------------------------- | | Nominal vs real loan balance | Inflation and construction costs raise required financing even without more leverage. | | Loan share vs GDP / property value | A share of bank lending is not the same as economy-wide leverage. | | Origination LTV and DSCR | Loss risk depends on borrower equity and property cash flow. | | Bank capital and provisioning | Modern capital, liquidity and supervisory rules differ sharply from 1990\. | | Borrower and property mix | Rental housing, developers, REITs, offices, hotels and logistics respond differently. | ### What would turn a warning into a crisis thesis? A serious bearish case requires several conditions to converge: 1. Cap rates rise faster than rents and net operating income. 2. Refinancing occurs at materially higher rates. 3. DSCR falls below 1 and borrower equity is thin. 4. Transaction volumes collapse, making appraisals stale. 5. Watch-list loans, arrears and provisions rise across multiple banks. 6. Losses become large relative to common equity—not merely large in yen. Without these links, ¥120.8 trillion is a large number searching for a probability. ### Editorial audit **Score: 82/100 — a strong warning article with an overpowered headline.** The reporting is substantially better than pure fear journalism: it includes low default data, banks’ risk controls and the BOJ’s conclusion that there is no immediate systemic danger. Its weakness is hierarchy. The most alarming historical comparison is placed at the top; the strongest resilience evidence arrives later. A fair headline would have been: > **Regional-bank property exposure hits a record; concentration rises, but defaults and capital buffers still point away from an immediate crisis.** ### What to monitor - Loan growth and property share by individual bank, not only system totals - Origination LTV, DSCR and interest-rate stress tests - Fixed/floating mix and the 12–36 month refinancing wall - Tokyo office vacancy, rents and cap-rate spread over JGB yields - Central-condominium transaction volume, discounts and inventory - Watch-list loans, restructurings, provisions and default rates - Geographic and property-type concentration - Deposit outflows and wholesale-funding dependence ### Final cut **Loan growth is neither a certificate of health nor proof of a crash.** It is fuel. Whether the fuel becomes productive investment or a fire depends on underwriting, cash flow, refinancing and capital. The size of the powder magazine matters. But it does not tell us whether the fuse is wet, dry—or even lit. --- At 15:00 JST on 23 October, the Bank of Japan released its October 2026 Financial System Report. The report does not declare a banking crisis. It says Japan’s financial system remains stable overall, financial intermediation is functioning smoothly, and capital ratios remain above regulatory requirements under all tested stress scenarios. What changed is the transmission map. The BOJ now explicitly combines an erosion of expected AI-investment profitability with higher oil prices, higher overseas interest rates and portfolio adjustment by non-bank financial institutions. Under that “overseas rate rise plus” scenario, losses increase in overseas private-credit funds and data-centre lending, while lower equity and fund prices and wider corporate-bond spreads weaken banks’ securities portfolios. The cross-border numbers are material. Japanese banks’ lending and investment exposure to overseas non-banks rose from JPY 8 trillion in March 2008 to JPY 126 trillion in June 2026\. Overseas funds’ investment in Japanese securities rose from JPY 111 trillion to JPY 408 trillion over the same comparison. The BOJ therefore says Japan may have become more sensitive to global market swings and overseas-fund behaviour. The property conclusion is also more precise than a simple bubble headline. Eleven of fourteen financial-imbalance indicators remain within their green trend ranges, and current activity shows no major imbalance. Yet regional banks’ property-fund lending includes substantial out-of-area exposure and some mezzanine loans. Tokyo commercial-property yield gaps have narrowed as property prices and JGB yields rose. Defaults remain low, and for rental businesses higher interest expense has so far been offset by higher operating profit, but that balance is not guaranteed. **Previous judgment:** JPY 120.8 trillion of property lending is a concentration signal, not proof of a systemic crisis. **Update:** The judgment is maintained and strengthened. The BOJ’s new stress test confirms system-wide capital resilience, while identifying the more credible fault lines: overseas non-bank leverage, AI and data-centre credit, fund redemptions, regional-bank cross-border property exposure, mezzanine risk and deposit stickiness. Tokyo’s cash market also rejected a one-way panic reading. Reuters reported the Nikkei down 1.3% at 10:50 JST, but it closed only 0.3% lower and ended the week down more than 4%. This intraday recovery is evidence of dip-buying resilience, not proof that AI-credit risk has disappeared. --- [Original Notion archive](https://app.notion.com/p/3fdbc361ccd28138aa02c3d8dae9f036?pvs=204&ref=tokyobrainie.ghost.io)