Hong Kong Digital Bank

Hong Kong digital banks at year six: three at breakeven, five still losing

All eight Hong Kong digital banks are past their fifth full year of operation. The FY2025 results published in early 2026 give the first clean cohort comparison. Three operators have reached at or near breakeven: ZA Bank, WeLab Bank and Mox. Five remain loss-making: Livi, Airstar, Fusion, Ant Bank Hong Kong and PAObank. The variable that separates the two groups is deposit-side scale tied to a captive distribution channel. Lending economics matter at the margin, but most of the variance lives in funding.

The HKMA renamed the segment from “virtual banks” to “digital banks” in 2024. The sector remains under 0.3% of system loans and deposits. The HKMA has ruled out issuing further licences, citing the profitability picture above.

Zhong An Bank is the cleanest case

ZA Bank posted its first half-year of profitability in H1 2025 at HK$49 million, becoming the second Hong Kong digital bank to break even after WeLab. Its first full-year profit followed in FY2025 at approximately HK$70 million, reported in ZhongAn’s 19 March 2026 annual results filing. Deposits grew 14.7% to HK$22.2 billion. The bank’s H1 2025 NIM was 2.38%, which is bank-typical, but its underlying advantage is the ZhongAn customer funnel: a Chinese insurtech with a Hong Kong listing that channels retail users into onboarding. The Web3 virtual-asset trading platform corridor is the second pillar, with ZA Bank acting as a regulated fiat on-ramp for licensed VATPs.

Mox crossed the line in Q1 2026

Mox is the Standard Chartered joint venture with HKT, PCCW and Trip.com. Mox remained loss-making for the FY2025 calendar year. The Ventures division at Standard Chartered, which includes Mox, posted a $59 million charge for 2025, down $14 million year on year as delinquency rates improved. Mox then reached operational breakeven in Q1 2026 and is targeting full-year breakeven for FY2026. The accelerator was a 2025 transfer of a $1 billion personal loan portfolio from Standard Chartered Hong Kong into Mox’s balance sheet. The Mox case is the cleanest illustration of the captive-distribution thesis: the parent moved real balance-sheet exposure into the digital bank to compress its time to profit.

WeLab is the counterexample

WeLab Bank reached its first monthly breakeven in December 2024 and stayed profitable through Q1 2025 on annual revenue of HK$750 million. WeLab has no large parent ecosystem. Its profitability comes from a high-margin unsecured retail lending book with a 2024 net interest margin of 9.4% and an H1 2025 NIM around 10.7%, several times the sector average. The bank’s funding base is smaller than ZA’s and its loan book sits at around HK$5.5 billion. WeLab demonstrates that lending economics can substitute for deposit scale, with a trade-off in concentration risk on a credit segment whose performance is sensitive to the Hong Kong consumer cycle.

The five still losing money

Livi narrowed losses to HK$259 million in FY2024 and has not yet released its FY2025 audited number publicly. Airstar, Fusion and Ant Bank Hong Kong filed FY2025 disclosures on the HKMA register without consolidated public summaries that show profitability. PAObank, now branded Ping An Digital Bank, reported net interest income up over 60% to more than HK$220 million in its 30 April 2026 release, with no profit declared and a HK$700 million capital injection from Lufax in 2025. The five share a pattern of parent support that funds growth without yet generating standalone earnings.

What the HKMA’s no-new-licences signal means

The HKMA’s decision to halt new licences acknowledges the cohort math. Eight banks chase under 0.3% of the system, and only those with captive customer-acquisition channels or a high-NIM specialty have closed the gap to profitability. The next test for the segment is whether the three near-breakeven banks can compound earnings without further parental balance-sheet transfers, and whether the five still losing money will face write-downs from their parent shareholders. The cohort is now mature enough for that comparison to carry weight.

Like what you’re reading? Sign up for our newsletter.

Like what you’re reading? Sign up for our newsletter.