UAE banks’ profits to rise in 2025, Moody’s warns returns to stay under pressure

The UAE’s biggest banks are enjoying stronger profits in 2025, but Moody’s Ratings warns that the good news comes with limits: returns on assets are set to remain under pressure as lower interest rates, rising provisioning charges and heavier tax burdens weigh on efficiency.

The four largest lenders — First Abu Dhabi Bank, Emirates NBD, Abu Dhabi Commercial Bank and Dubai Islamic Bank — posted a combined net profit of Dh32 billion in the first half of the year, up 6% year-on-year. The uplift was largely driven by robust non-funded income and steady loan growth, which offset softer margins and rising costs.

But beneath the headline profit growth, returns are beginning to soften. Aggregate return on assets fell to 1.8% in the first half of 2025, down from 2.0% in the same period last year. According to Moody’s, this erosion will persist.

“Going forward, we expect UAE banks' 2025 net profits to continue to expand yet returns on assets will remain constrained by lower interest rates, higher provisioning charges and increased tax bill,” said Badis Shubailat, Assistant Vice President-Analyst at Moody’s Ratings.

Part of the squeeze comes from the UAE’s new corporate tax framework, which has sharply lifted banks’ effective tax bills. The top four lenders collectively paid Dh6.6 billion in taxes in H1 2025, equal to 17% of pre-tax earnings — more than double the effective rate seen before the levy was introduced.

At the same time, loan-loss provisioning is normalising after a period of strong recoveries and reversals. This has lifted the sector’s cost of risk, while digital transformation spending continues to push operating expenses higher.

For now, the underlying economic backdrop remains supportive, with robust credit demand and strong fee income generation. Yet the message from Moody’s is clear: UAE banks can keep growing their bottom line, but the ability to squeeze as much return from their asset base as in previous years is fading.