CITY
Rp 2,084 trillion of Indonesian household credit, rendered as a night city. Each district is a lending product. Each building is a slice of the loan book. And the windows are lit at the exact rate the loans perform — the red ones have stopped paying.
One window in thirty has gone dark.
NPL — a non-performing loan — is credit classified kurang lancar, diragukan or macet: in plain language, a loan that has stopped performing. At April 2026 ratios, that is one mortgage window in thirty. Personal lending, the calmest district, keeps one in fifty-three dark.
The city below is built from OJK's own portal data. Building volume tracks the size of each book; the share of red windows in every district is its NPL ratio. Nothing is decorative.
Every consumer district runs cleaner than SME.
Consumer credit is in far better shape than business lending: every product below sits under 4.2%, against 4.86% for SME credit in the companion report. But four movements deserve attention, and the sparklines carry them.
The tall towers are the weak spot.
Flats and apartments rose every single month — a full point above house mortgages. The house book itself (Rp 780 trillion, by far the largest consumer product in Indonesia) holds stable around 3.2%. Property stress lives in the towers, not the streets.
A district that shrinks while it sours.
The automotive book fell from Rp 131.2 to 128.6 trillion over four months while its ratio climbed 2.48% → 2.59%. And the rise is not just arithmetic from the shrinking denominator: the stock of bad loans itself grew, by Rp 78 miliar.
A slow, steady creep.
2.37 → 2.50%, on a slowly shrinking book. Not alarming; consistently one direction. Card receivables sit at issuers' head offices — 98% booked in Jakarta — so this district exists only as a national figure. No province can be blamed, or cleared.
The calm giant of the west side.
The largest district in the city — payroll loans and multiguna mass — runs below 2% with only mild drift. Bank paylater, its small neon annex (Rp 28 trillion), is growing fast and actually improving in quality: 2.27% in January, 2.12% by April.
Fourteen provinces, one red book.
Every mortgage book of Rp 10 trillion or more, ranked by the share that has gone bad. The right-hand column is the size of each book — Jawa Barat's 4.02% sits on Rp 180.5 trillion, which is what makes it the finding.
Property is the problem. Only property.
Jawa Barat holds the worst large mortgage book in the country — 4.02% on Rp 180.5 trillion, the second-largest book there is. Its flats run 5.49%. Its ruko run 5.98%, nearly two points above national. And yet its personal lending and its locally-booked auto slice are cleaner than the national ratio.
Set against the companion SME report, one consistent risk profile emerges for West Java: stress concentrates in SME credit (6.44%, worst of the big provinces) and in housing — both tied to the same industrial belt and its property market. That reads like a regional income-and-employment problem feeding through to businesses and homeowners, not undisciplined consumer lending.
One growth flag: Jabar's bank-paylater book grew 22% in four months (Rp 3.8 → 4.7 T). Quality is fine today at 2.38%; books growing that fast deserve a look every quarter.
Four districts, all standing. Watch the towers.
All four consumer products are healthier than SME credit — mortgage 3.28%, automotive 2.59%, credit card 2.50%, personal 1.89%, against SME's 4.86%.
The deteriorating cells: flats and apartments (4.16%, climbing every month), automotive (rising ratio on a shrinking book), credit card (+13 bps). House mortgages and personal lending are stable.
And Jawa Barat's consumer problem is property, specifically — the worst large mortgage book in the country, while its households otherwise pay on time.