Hapoalim: What the 2025 to 2026 Issuance Wave Really Says About the Funding Stack
The main article flagged funding as one of Hapoalim’s key post-rate-peak tests. This follow-up shows why: deposits still grew, but the mix moved toward costlier and less stable layers, liquid assets fell by about NIS 21 billion, and the bank is already using both local and international debt markets as a standing part of balance-sheet management.
The Follow-Up: The Funding Stack Is No Longer Running on Autopilot
The main article argued that Hapoalim’s post-rate-peak challenge would not show up first in reported earnings. It would show up in the quality of the liabilities funding growth. This follow-up isolates only that question: is the 2025 to 2026 issuance wave just opportunistic execution by a strong bank, or is it already evidence that liability management has become a structural and permanent part of the model.
The short answer is that it is structural, but not distressed. There is no liquidity event here. Deposits from the public still rose to NIS 592.7 billion, LCR still stood at 130%, NSFR still stood at 116%, and both local and international debt markets remained open. But under that headline, the picture changed. Demand deposits fell, the share of interest-bearing deposits kept rising, liquid assets fell, and regulatory funding ratios were preserved partly through active liability-side work. In other words, Hapoalim is still well funded, but it can no longer assume that the balance sheet will fund itself.
That is why the point matters now. The bank did not open just one funding channel. It opened several in parallel: commercial paper, local senior bonds, subordinated loss-absorbing instruments (COCO), credit-linked notes (CLN), and then dollar bonds abroad in January 2026. When management explicitly says it is examining additional bond issuance shortly after the financial statements and from time to time according to ongoing liquidity needs, that should be read exactly as written: not as theoretical backup, but as a routine part of the liability machine.
Deposits Are Still Growing, but They Are Weakening Where It Matters
The yellow flag is not the size of the deposit base. It is the quality of that base. Deposits from the public rose 3.2% in 2025 to NIS 592.7 billion, but demand deposits fell 1.9% to NIS 296.3 billion, while time deposits rose 8.8% to NIS 296.4 billion. Within the demand layer, non-interest-bearing deposits fell to NIS 147.3 billion from NIS 149.6 billion, and interest-bearing demand deposits also fell, to NIS 149.0 billion from NIS 152.3 billion. The cheap and flexible part of funding did not collapse, but it stopped growing exactly when credit kept expanding quickly.
This is not only a deposit-pricing story. It is also an asset-allocation story on the customer side. Off-balance-sheet customer securities holdings jumped 13.8% to NIS 1,283.5 billion, and the bank explicitly says that the increase was driven partly by customer migration from deposits and current accounts into capital markets. So the deterioration in funding mix is not just about paying more for deposits. It is also about customers parking more money outside the deposit base.
The economic meaning goes beyond a few basis points of funding cost. Demand deposits do two things at once for a bank: they lower the cost of funding and they provide a relatively stable, elastic source of balance-sheet support. When they weaken, the bank does not just pay more. It also has to rebuild part of that flexibility through other layers of funding.
| Public deposit layer | 2024 | 2025 | Change |
|---|---|---|---|
| Non-interest-bearing current accounts | NIS 149.6 billion | NIS 147.3 billion | Down 1.6% |
| Interest-bearing demand deposits | NIS 152.3 billion | NIS 149.0 billion | Down 2.2% |
| Total demand deposits | NIS 301.9 billion | NIS 296.3 billion | Down 1.9% |
| Time deposits | NIS 272.3 billion | NIS 296.4 billion | Up 8.8% |
| Total public deposits | NIS 574.3 billion | NIS 592.7 billion | Up 3.2% |
Liquidity Is Still High, but the Warehouse Is Smaller
The easy mistake is to look at a 130% liquidity coverage ratio and conclude there is no story here at all. That is wrong. The story is not the fact that the ratio stayed high. It is the way it stayed high. In the fourth quarter of 2025, average liquid assets used for liquidity-risk management fell to NIS 154.7 billion from NIS 175.9 billion in the comparable quarter. High-quality liquid assets (HQLA) fell by the same order of magnitude, to NIS 154.7 billion from NIS 175.9 billion. The bank explicitly says the main reason was credit growth exceeding growth in sources of funding.
Why did the ratio barely move despite the drop in the liquid inventory? Because the denominator became easier. Average net cash outflows fell to NIS 118.8 billion from NIS 134.1 billion. Within that, weighted outflows fell by about NIS 4.5 billion, mainly due to lower wholesale deposits, while weighted inflows rose by about NIS 10.9 billion, mainly because of higher cash inflows from performing credit exposures.
That leads to the sharper analytical point. A 130% headline ratio can look like an unchanged buffer. In reality, the bank is already using the balance sheet more aggressively. It has not lost control of liquidity, but it has already consumed part of the liquid inventory to let credit keep growing faster than deposits.
NSFR Tells the More Important Story
If LCR describes the next month, NSFR describes what is happening to the structure of funding itself. Here the picture is clearer. NSFR fell to 116% at year-end 2025 from 125% at year-end 2024. That is still above the 100% regulatory minimum and above the bank’s own 105% internal limit, but the margin is no longer comparable with where it stood a year earlier.
More important than the ratio itself is the decomposition behind it. Available stable funding rose only 3.0% to NIS 453.9 billion. Required stable funding, by contrast, jumped 10.7% to NIS 390.9 billion. In other words, the uses side grew far faster than the stable-funding side.
The internal composition of funding is even more revealing. Within available stable funding, retail and small-business deposits fell to NIS 242.5 billion from NIS 251.2 billion, while wholesale funding rose to NIS 90.0 billion from NIS 79.7 billion. That is exactly the part the ordinary deposits table does not tell on its own. Even when total deposits rise, the most regulator-friendly and stable funding layer is weakening, and the system starts leaning more heavily on less comfortable layers in the model.
| Stable-funding component | 2024 | 2025 | Change | Why it matters |
|---|---|---|---|---|
| Retail and small-business deposits | NIS 251.2 billion | NIS 242.5 billion | Down 3.5% | The core stable layer weakened |
| Wholesale funding | NIS 79.7 billion | NIS 90.0 billion | Up 12.9% | More weight moved to less stable sources |
| Total available stable funding | NIS 440.5 billion | NIS 453.9 billion | Up 3.0% | Stable sources did not keep up with uses |
| Total required stable funding | NIS 353.2 billion | NIS 390.9 billion | Up 10.7% | Credit and balance-sheet uses kept expanding |
| NSFR | 125% | 116% | Down 9 points | Still high, but with less room |
This is the center of the thesis. The question is not whether Hapoalim has access to funding. It does. The question is whether the stable-funding layer is keeping pace with the uses side. As of the end of 2025, the answer is: not fully.
The Issuance Wave Shows Routine Management, Not One-Off Timing
This is where the 2025 to 2026 sequence really matters. If this had been a single transaction, it could still be read as opportunistic execution into a good market window. But Hapoalim effectively built a full funding ladder: short commercial paper, local senior bonds, COCO instruments, CLNs, and then offshore dollar bonds.
| Funding layer | What the bank actually did | Amount disclosed in the filings | What it says |
|---|---|---|---|
| Commercial paper | Issued in March 2025 and again in June 2025, both with one-year maturity | About NIS 2.03 billion and another NIS 1.47 billion | A short-end bridge that fills liquidity gaps without locking long maturity |
| COCO | Expanded in March and May 2025 and issued again in August 2025 | About NIS 1.16 billion, NIS 1.15 billion, and NIS 1.64 billion | Reinforces the loss-absorbing and capital-linked layer |
| Local senior bonds | Non-indexed issuance in June, CPI-linked issuance in August, and two series in December | About NIS 1.58 billion, NIS 3.42 billion, and NIS 4.12 billion | Locks medium-to-long local funding |
| CLN | Transactions in March and June, additional transactions in August, plus a partial early redemption in July | About NIS 1.75 billion and another NIS 1.4 billion, against an early redemption of about NIS 85 million | Diversifies the funding stack through credit-linked instruments |
| Liability redemptions | Early redemption of Series 9 in April 2025 | About NIS 1.2 billion | This was not just issuance. It was active reshaping of the liability stock |
| Offshore and continued local pipeline | Pricing on January 7, 2026 and closing on January 14, 2026 of two dollar tranches, then top ratings for a new local Series 104 on March 1, 2026 | $2 billion, and later up to NIS 500 million for Series 104 | Evidence that both international and local channels are being kept open in parallel |
The immediate report on the international pricing is a key moment. The bank raised $2 billion after roughly $6.9 billion of demand, in two $1 billion tranches maturing in 2029 and 2033 with 4.722% and 5.252% coupons. That tells two things at once. First, market access is excellent. Second, that access is already being used. This is no longer just optional capacity sitting on the shelf.
Early March 2026 points in the same direction. Midroog assigned Aaa.il with stable outlook to the new local Series 104 up to NIS 500 million nominal value, and S&P Maalot assigned ilAAA to the same series at the same size. Together with management’s statement that it is examining additional issuance according to ongoing liquidity needs, this maps a continuous operating model: the local market stays open, the international market stays open, and the bank wants to keep both channels alive.
Conclusion: Not a Funding Crisis, but Permanent Funding Discipline
Hapoalim is not entering 2026 with a liquidity problem, and not with critical capital-markets dependence either. The bank itself says issuance accounts for only about 5.5% of total sources of funding, so debt markets are still not its primary source. But the opposite reading, as if capital markets remain just a backup layer that can be activated occasionally, no longer fits the numbers.
The more precise reading is sharper. Hapoalim has moved into a phase where liability management is a permanent part of operations. Deposits are no longer moving in the right direction on quality, the liquid inventory is already being used to support growth, and NSFR is already compressing despite very strong market access. As long as the bank keeps both local and international channels open, this is not a stress story. But if credit keeps growing faster than stable funding, liability management will stop being a tactical choice and become one of the main variables determining the quality of Hapoalim’s earnings.
That also defines what matters next. The key question is not only whether the bank issues again. It is why, at what cost, and which deposit layer replaces what is being lost. At a bank like Hapoalim, the important question is never just whether another source of funding exists. It is how often that source already needs to be used to keep the machine running at the same speed.
Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.
The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.
The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.