Open in interactive viewer → charts, metric popovers & call review

Discount Bank's Quarter of One-Offs Is Really a Bet on Mercantile — and a CAL Sale That Won't Close

A 10.4% headline ROE obscures 1,100 heads leaving, a failed divestiture, and a bank deliberately shrinking its risk density to lift returns
DSCT.TA · Earnings Call · 2026-11-18

The headline is noise; the org chart is the signal

Israel Discount Bank reported a 10.4% return on equity for the third quarter of 2026 — a number management spent most of the call explaining away. Strip out ILS 200 million of merger provisions and roughly ILS 130 million of merger-related costs and the quarter's ROE was 11.9%; strip out the special bank tax too and it was 12.6%. Over nine months the same pattern holds: 11.8% reported versus 13.4% clean, on ILS 3 billion of net income. That gap is the story. This is a bank that is voluntarily absorbing pain now — firing people, merging a subsidiary, and holding excess provisions — in exchange for a structurally cheaper franchise later. The keyword record confirms the reset was total, not incremental. Discount's entire top-of-list vocabulary this quarter is fresh: merger, Mercantile, IDB New York and RWA density all registered their sharpest momentum readings in the same period. When a bank's language rotates wholesale in a single quarter, the operating model has changed, not just the quarter.

Mercantile: synergies that arrive in 2028

In October the board approved the merger with Mercantile Bank. The arithmetic: 500–600 full-time employees out by the end of 2028, roughly ILS 200 million of annual cost savings from that alone, and about ILS 250 million in total annual run-rate savings once overlapping branches, governance and support functions go. It also hands Discount deeper penetration of what it calls Israel's fastest-growing segments — the Arab community and the ultra-Orthodox sector — where the bank already claims a leading share. The timing is the catch. Approval from the Bank of Israel is expected by the end of the first half of 2027, the headcount reduction begins in the second half, and the process runs roughly 18 months from there. Morris Dorfman was explicit that “we're entering the merger period with Mercantile, which will be probably a period of 1 to 2 years” — Morris Dorfman, Likely an Executive, possibly CEO or CFO · 2026-11-18 and that costs rise before they fall. Meanwhile the earlier standalone programme — 600 FTEs out by end-2026, worth ILS 270 million fully loaded — is already 90% achieved, which is why staff costs fell 1.1% quarter-on-quarter and 1.3% year-on-year. The bigger frustration for anyone underwriting the story is disclosure. Management has deferred its long-term targets: “as soon as we finalize negotiations with both Mercantile and Discount labor unions... we plan to present the market with an updated version of our long-term strategy and targets” — Morris Dorfman, Likely an Executive, possibly CEO or CFO · 2026-11-18. Until then, investors are asked to value a restructured bank on a restructuring calendar.

The CAL overhang that keeps paying for the capital ratio

Discount's capital story rests on a divestiture that has already failed once. The bank must exit its CAL credit-card subsidiary under law, and the sale to Union Investments and Harel was terminated after the Competition Commissioner's ruling. Management is now seeking an extension and pursuing alternatives including an IPO.

Following the Competition Commissioner's ruling and the termination of the sale of CAL to Union Investments and Harel, the bank has requested an extension to allow the process to be completed in an orderly manner. At the same time, we continue to pursue all possible ways to complete the divestiture ASAP, including a possible IPO and the sale of CAL to potential buyers.

Morris Dorfman, Likely an Executive, possibly CEO or CFO · 2026-11-18
The prize is real: about 60 basis points of CET1 from lower risk-weighted assets, plus roughly 4.5 points of LCR and 3.5 points of NSFR. That would push CET1 toward 11%, well clear of the Bank of Israel's requirement. But note that CAL itself earned ILS 99 million at a 13.2% ROE last quarter — it is a good business being sold for regulatory reasons, and the CAL as required framing is a reminder that this is an obligation on a clock, not a strategic choice. Every capital-ratio chart presented without the divestiture is showing a contingent number.

Where the money actually leaked

Revenue fell 5% quarter-on-quarter: interest income down 4.2% on lower Bank of Israel rates and subdued CPI, fees down 3.2% on weaker capital-market and credit-handling income, partly offset by ILS 38 million of real-estate gains and 7.5% growth in financial income from regular operations. The culprit is the low interest rate environment colliding with genuine deposit competition. Dorfman did not soften it: “there is a strong competition in Israel. I assume that this competition will continue” — Morris Dorfman, Likely an Executive, possibly CEO or CFO · 2026-11-18. That is the same core deposits pressure showing up across the global banking keyword set this quarter — a sector-wide cost-of-funding squeeze rather than a Discount-specific miss. Israeli peers reporting in the same window told the same story: credit growth and expense control at Mizrahi, cost of risk and lower rates at Hapoalim, a rising loan portfolio at Leumi. Discount's answer is mix. Its RWA density is falling because it is deliberately tilting toward retail — mortgages plus 7.4% and consumer credit plus 5.4% year-to-date, versus corporate at 12.7% — while acknowledging it cannot out-price competitors on the commercial side. Management was also quietly correctional on rate sensitivity, telling an analyst it had "remained the same more or less" even as the disclosures showed otherwise.

Credit quality is the quiet asset

Against all that soft revenue, the loan book behaved impeccably. Non-performing loans eased to 0.64% with coverage at 189%; problematic loans fell to 1.4% of gross loans, and in real estate from 3.9% a few years ago to 1.4% today. Loan losses ran 16 basis points in the quarter, 17 for nine months — more than two-thirds of which was collective provisioning the bank chose to take. Total quarterly loan growth of 1.2% was slower than prior periods, and Dorfman pushed back on reading anything into it: “I don't feel comfortable with 15%, 16%, or 18% growth at the moment in Israel” — Morris Dorfman, Likely an Executive, possibly CEO or CFO · 2026-11-18, targeting instead a rate a little above nominal GDP. The subsidiary picture cuts both ways. IDB New York chopped nearly 100 staff year-to-date and drove its cost/income ratio from 72% to 59%, yet still earned only $27 million at a 7.3% ROE and generated most of the quarter's specific provisions. Mercantile, ex one-offs, made ILS 208 million at a 14.7% ROE with loans up 8.9%. IDB New York is the fixable problem; Mercantile is the thesis.

What to watch

Three markers decide whether this quarter was a trough or a plateau. First, whether the Bank of Israel approves the merger in the first half of 2027 — the entire ILS 250 million synergy case is hostage to that date. Second, whether CAL gets sold or floated; a 60-basis-point capital uplift and a 40% payout ratio both lean on it. Third, whether deposit competition compresses net interest margin faster than the retail mix shift can offset it. For now the franchise is buying its own future with today's earnings — reported ROE 10.4%, real earning power somewhere in the low teens — and asking shareholders to wait until 2028 for the bill to arrive in reverse.