Taiwan central bank holds rates and raises second-home mortgage LTV cap to 70%
Published · Updated
What happened
The Central Bank of the Republic of China (Taiwan) kept its policy rate unchanged and announced that the maximum loan-to-value ceiling for a natural person buying a second home would rise to 70% from 60%. The adjustment changed one of the bank's selective credit-control settings for housing finance.
Why it matters
A central bank can leave the benchmark rate unchanged while still changing financial conditions through targeted credit rules. Easing a second-home mortgage cap can increase housing transaction capacity and bank lending in that segment, while also testing whether earlier property curbs still anchor borrower and market behavior.
On September 18, the Central Bank of the Republic of China (Taiwan) left policy rates unchanged and separately eased a selective housing credit rule by lifting the maximum loan-to-value ratio for a natural person’s second-home mortgage from 60% to 70%. The move produced an immediate market reaction because the rate hold was expected, while the housing-credit adjustment was not. For Taiwan's finance and banking authorities, the live issue is whether the change supports sentiment and credit demand without weakening macroprudential discipline in the property market.
Watch for
Central Bank of the Republic of China (Taiwan) release on 2026-09-22 or later clarifying eligibility, regional scope, or implementation details for the 70% second-home LTV cap Financial Supervisory Commission statement by 2026-09-22 on banks' underwriting expectations for second-home mortgages after the central bank change Taiwan Stock Exchange trading data on 2026-09-19 to 2026-09-22 for listed financial shares and property developers relative to the TAIEX close Major state-linked banks including Bank of Taiwan or Land Bank of Taiwan posting updated second-home mortgage rate sheets or application terms by 2026-09-22
Sources
- 央行理監事會利率連2凍 房市管制鬆綁第2戶貸款成數升至7成 · cna.com.tw
- 本行理監事聯席會議決議 · cbc.gov.tw
Researched and written by AI from the sources listed, under published rules. How V³ Aware is made