Published: 15 May 2026 | Last data refresh: 15 May 2026 | SGX: O39
At ~S$22.92 (14 May 2026), OCBC trades at 1.71 times book value. Run that multiple through the Reverse P/B formula — Implied ROE = P/B × (Ke − g) + g — and the answer is 9.95%. That is what the market is currently pricing as OCBC’s long-run return on equity: a level of profitability the bank has not delivered in recent memory. FY2025 came in at 12.6% despite a 29-basis-point NIM compression and a Pillar Two minimum tax hit. Q1 2026 is running at an annualised 13.0%. The implicit bet embedded in today’s price is that something — rates, credit quality, fee income sustainability — will grind ROE down by roughly 2.65 percentage points on a permanent basis.
This article investigates whether that implied deterioration is reasonable. I approach it through three frameworks — a Dividend Discount Model, a Justified P/B analysis, and the Reverse P/B calculation above — and they produce a wide range of answers. The DDM gives S$19–21 (slight downside from current price). The P/B framework gives S$25–47 (meaningful to substantial upside). The gap between the two is not a modelling error. Understanding why they disagree is, in my view, the most analytically useful thing this piece can offer. This is not a directional call; I hold a position and I will say what I think at the end, but the frameworks are the substance.
The Rate Headwind Is Real — But Part of It Is a Choice
OCBC’s NIM has fallen 54 basis points from its FY2023 peak. The trajectory is stark: approximately 2.30% in FY2023 → 2.20% in FY2024 → 1.91% in FY2025 → 1.76% in Q1 2026, with compression still ongoing. Over this downcycle, OCBC’s NIM compression has been among the sharpest of Singapore’s three listed banks, a point that has attracted some pointed commentary. The obvious culprit is rate normalisation — as the Federal Reserve cut rates and SGD, HKD, and USD benchmark rates followed, loan yields compressed and assets repriced downward.
That explanation is real but incomplete.
Per OCBC’s Q1 2026 results, management made “a strategic increase in income-accretive high-quality treasury assets by an average of 7%” during the quarter. Rather than deploying excess liquidity into higher-yielding commercial lending — which would have supported reported NIM — OCBC consciously repositioned the balance sheet toward lower-yielding but higher-quality instruments. The trade-off is clear: some NIM drag now, in exchange for a better-quality asset base if credit conditions deteriorate.
This partly explains OCBC’s sharper NIM compression relative to peers. It is a feature of a deliberate balance sheet strategy, not a symptom of competitive weakness. A bank that stretches into riskier lending to protect headline NIM is not doing something admirable — it is borrowing future credit quality to improve present optics. The S$191 million in precautionary provisions booked in Q1 2026 — overlays for Middle East tail risks — signals that management is already hedging against a credit downturn that banks chasing NIM through riskier loans may not be as well-positioned to absorb.
The partial offset to NIM compression shows up in volume. Average interest-earning assets grew 10% QoQ in Q1 2026. The CASA ratio stood at 50.2%, providing a relatively stable, lower-cost funding base. Net interest income was S$2.22 billion in Q1 2026, down 5% year-on-year — a meaningful decline, but one absorbed without any deterioration in credit quality: the NPL ratio held at 0.9% for the eighth consecutive quarter.
Where the NIM story goes from here turns primarily on the interest rate path. NIM stabilisation somewhere above 1.75% is the base case underpinning my model. Below that level, the DDM values deteriorate materially — Table 1 in the valuation section makes this visible.
The Wealth Pivot — Structural, Not Cyclical
The more interesting story in FY2025 and Q1 2026 is not NIM; it is what is happening on the other side of the income statement.
Non-interest income grew 15.8% year-on-year in FY2025 and a further 23% in Q1 2026, hitting a record S$1.61 billion in a single quarter — more than 40% of total income. Wealth management fees grew 33% in FY2025 and 34% in Q1 2026 to S$422 million. AUM reached S$343 billion, up 15% year-on-year. The progression of non-interest income’s share of total revenue tells the story cleanly: 28.6% in FY2023 (the actual low, the peak NIM year, when high rates crowded out fee income) → 32.6% in FY2024 → 37.4% in FY2025 → 42.0% in Q1 2026.
The critical analytical question is whether this is a cyclical fee recovery — markets rally, product sales increase, fees go up — or a structural re-rating of OCBC’s revenue mix. The statistic I watch most closely is net new money: S$27 billion of inflows in FY2025. Net new money is not a market-appreciation figure. It represents clients actively bringing assets to OCBC, which is structurally distinct from passive AUM growth through mark-to-market gains. That number suggests the wealth engine is attracting capital, not merely riding it.
OCBC’s position in wealth management is differentiated from DBS through Bank of Singapore, its dedicated private banking franchise. Announced days before the Q1 2026 results, the acquisition of HSBC’s Indonesia wealth business — Tan Teck Long’s first strategic deal since taking over as CEO in January 2026 — builds on OCBC’s existing Indonesia franchise and adds S$6.6 billion of AUM from a portfolio HSBC is divesting. The strategic fit is clear; the execution question is real, and I flag it in the risks section.
There is also a structural demand driver that quarterly data alone does not capture. Singapore officially became a super-aged society in 2026, with more than 21% of the population aged 65 or older — a demographic milestone that OCBC explicitly anticipated and positioned around. Then-CEO Helen Wong launched the OCBC SeniorCare programme in March 2025, with S$2 million committed and more than 180,000 seniors targeted. At the launch, she said: “By next year, Singapore will likely become ‘super-aged’. OCBC is hence committing more than S$2 million to the OCBC SeniorCare programme.” This is not a CSR initiative attached to the wealth business. It is a deposit-gathering and intergenerational wealth transfer capture strategy. The 65-and-above demographic holds the majority of Singapore’s investable assets; a bank that captures those primary relationships is also positioned to capture the downstream wealth transfer to the next generation. Demographic tailwinds of this kind are slow-moving and do not show up cleanly in any single quarter — which is precisely why they tend to be underpriced in bank multiples.
I think the distinction between cyclical recovery and structural re-rating matters enormously for how the bank should be valued. If the wealth engine is structural, the P/B framework is the right anchor. If it proves cyclical, the DDM is closer to correct. This tension drives the valuation section.
Capital Allocation — Over-Capitalised, But the Return Is Quantified
OCBC’s capital position is comfortably above any regulatory requirement. CET1 stood at 17.0% on a transitional basis in Q1 2026, or 15.2% on the fully phased-in (1 January 2029) basis. After the FY2025 final dividend payment on 8 May 2026 and the anticipated impact of the HSBC Indonesia acquisition, the fully phased-in ratio is expected to settle around 14.0%.
The MAS D-SIB minimum CET1 requirement under MAS Notice 637 is 9.0%. Singapore banks have broadly managed toward a 12.5–13.5% fully phased-in range.
In February 2025, OCBC announced a S$2.5 billion capital return plan: S$1.5 billion in special dividends and S$1.0 billion in share buybacks. The special dividends tranche is largely complete — S$0.16 per share was paid on 8 May 2026 as part of the plan. Of the S$1.0 billion buyback tranche, only approximately 20% (~S$200 million) has been executed. Management confirmed at the 2026 AGM: if buybacks are not completed by FY2026 year-end, the remaining balance will be paid as a special dividend — approximately 18 cents per share.
This is the Double Dividend picture for FY2026: ordinary DPS of S$0.83 plus up to S$0.18 in capital return completion equals up to S$1.01 per share, a potential total cash yield of 4.4% at current prices. The DDM I use below captures only the ordinary dividend; the capital return kicker is separate, finite, and quantified — and I note its exclusion explicitly in the valuation section.
Valuation — Three Lenses and Why They Disagree
Setting up the cost of equity
CAPM inputs: risk-free rate 2.10% (SGS 10-year yield, 15 May 2026), equity risk premium 5.0% (standard Singapore ERP), beta 0.95 (OCBC, large-cap). Cost of equity: Ke = 6.85%.[^rf]
DDM: the dividend-yield anchor
The Two-Stage DDM uses FY2026E ordinary DPS (D1) of S$0.83 — held flat from FY2025, conservative given the Q1 2026 annualised run-rate implying S$0.88. Phase 1 (FY2027–FY2030) grows at 5.0% per annum, anchored to Q1 2026’s 5% year-on-year profit growth. Terminal growth is 2.5%, consistent with long-run nominal GDP.
Phase 1 projection:
| Year | FY2026 | FY2027 | FY2028 | FY2029 | FY2030 |
|---|---|---|---|---|---|
| Ordinary DPS (S$) | 0.83 | 0.87 | 0.92 | 0.96 | 1.01 |
| Discount factor | 0.9359 | 0.8759 | 0.8197 | 0.7672 | 0.7180 |
| PV of DPS (S$) | 0.78 | 0.76 | 0.75 | 0.74 | 0.72 |
Sum of PV (Phase 1): S$3.75. Terminal DPS (FY2031): S$1.03. Terminal value: S$23.77, discounted at 0.7180 to give PV of S$17.07. Two-Stage DDM intrinsic value: S$20.82 — implying 9.2% downside from S$22.92.
The Gordon Growth Model produces a scenario range under varying terminal growth assumptions (D1 = S$0.83, Ke = 6.85%):
| Scenario | Terminal growth (g) | Fair value | vs S$22.92 |
|---|---|---|---|
| Bear | 2.0% | S$17.11 | −25.3% |
| Base | 2.5% | S$19.08 | −16.8% |
| Bull | 4.0% | S$29.12 | +27.1% |
One limitation of the DDM deserves explicit mention before moving on: it models only the recurring ordinary dividend stream. A holder in FY2026 may also receive up to ~S$0.18 per share from the completion of the S$2.5 billion capital return plan — the remaining ~S$800 million of the buyback tranche, payable as specials if buybacks are not executed by FY2026 year-end. That is an additional ~2.2% yield on top of the ordinary 3.6%. The DDM fair value of S$20.82 should be read as a valuation anchored to recurring dividends only — it can move lower still under more pessimistic growth assumptions, as Table 1 makes clear. The capital return kicker is a separate, finite, and quantified top-up that the DDM does not capture.
P/B framework: the earnings-power view
The Justified P/B formula is (ROE − g) / (Ke − g). Multiplied by book value per share (S$13.38), this gives:
| Scenario | Sustainable ROE | g | Justified P/B | Fair value | vs S$22.92 |
|---|---|---|---|---|---|
| Bear | 11.0% | 2.0% | 1.86x | S$24.83 | +8.3% |
| Base | 12.5% | 2.5% | 2.30x | S$30.76 | +34.2% |
| Bull | 14.0% | 4.0% | 3.51x | S$46.95 | +104.8% |
Even the bear case — ROE of 11.0%, well below both FY2025 actual (12.6%) and Q1 2026 annualised (13.0%) — produces a fair value above today’s price.
Why the DDM and P/B framework disagree — and why that matters
The two frameworks produce radically different answers from broadly similar underlying assumptions. This is the intellectual core of the analysis, and it should not be glossed over.
The DDM values only the cash actually distributed to shareholders. It is fully correct when the payout ratio equals 1 − g/ROE — that is, when the bank retains only what is needed to fund growth at exactly the cost of equity rate, making retained and distributed earnings equivalent in value. Under that specific condition, DDM and P/B converge.
OCBC retains approximately 49% of earnings (50.9% payout ratio in FY2025). Its ROE is 12.6%. If those retained earnings are deployed at 12.6% — the same ROE as the existing business — they are creating value at roughly 5.75 percentage points above the cost of equity (6.85%). The DDM, by modelling only the dividend stream, implicitly assigns zero value to that retained-and-compounded amount. It treats S$1 retained by a bank earning 12.6% ROE as worth no more than S$1 retained by a bank earning exactly its cost of capital.
The formal reconciliation condition is: the two frameworks agree when g = ROE × retention rate. At OCBC’s current parameters: 12.6% × 49% = 6.2%. My model uses g = 2.5%. The gap — 3.7 percentage points — is precisely what explains the DDM/P/B divergence. It is not an error; it is the model’s way of saying the DDM implicitly assumes OCBC cannot redeploy retained earnings productively, while the P/B framework assumes it can.
Which framework is correct depends on one empirical question: does OCBC actually compound retained capital at near-ROE rates, or does it dissipate the value through bad acquisitions, excess costs, or misallocated capital? I think the wealth management build-out — demonstrated through net new money inflows, not just market appreciation — is the evidence that retained capital is being deployed productively. But this is a judgement call, not a certainty, and readers should form their own view.
Reverse P/B: what the market is saying
The sharpest edge of the analysis is the Reverse P/B calculation. At S$22.92 (P/B 1.71x), using Ke = 6.85% and g = 2.5%:
Market-implied long-run ROE = P/B × (Ke − g) + g = 1.71 × 4.35% + 2.5% = 9.95%
That is 2.65 percentage points below FY2025 actual ROE of 12.6%, and 2.15 points below the five-year average of approximately 12.1%. The market is not pricing modest deterioration — it is pricing ROE settling permanently at a level OCBC has not delivered in recent memory. Whether that is a reasonable forecast of the bank’s future, or a pessimistic anchor that the wealth pivot is gradually rendering obsolete, is the investment question this analysis poses.
Sensitivity analysis
Table 1 — Gordon Growth DDM: Intrinsic value (S$) under varying Ke and g
D1 = S$0.83. Values above S$22.92 in bold.
| Ke ↓ / g → | 1.5% | 2.0% | 2.5% | 3.0% | 3.5% | 4.0% |
|---|---|---|---|---|---|---|
| 6.0% | 20.75 | 23.71 | 27.67 | 33.20 | 41.50 | 55.33 |
| 6.5% | 18.44 | 20.75 | 23.71 | 27.67 | 33.20 | 41.50 |
| 7.0% | 16.60 | 18.44 | 20.75 | 23.71 | 27.67 | 33.20 |
| 7.5% | 15.09 | 16.60 | 18.44 | 20.75 | 23.71 | 27.67 |
| 8.0% | 13.83 | 15.09 | 16.60 | 18.44 | 20.75 | 23.71 |
| 8.5% | 12.77 | 13.83 | 15.09 | 16.60 | 18.44 | 20.75 |
At the base case (Ke = 6.85%, g = 2.5%), the DDM value sits between the 6.5% and 7.0% Ke rows at S$19.08–S$23.71, consistent with the model’s Two-Stage result of S$20.82. The current price requires terminal growth of at least 3.0% at Ke = 6.85% to be justified on a pure dividend basis.
Table 2 — P/B Framework: Fair value (S$) under varying ROE and Ke
g fixed at 2.5%. Book value per share S$13.38. Values above S$22.92 in bold.
| ROE ↓ / Ke → | 6.0% | 6.5% | 7.0% | 7.5% | 8.0% | 8.5% |
|---|---|---|---|---|---|---|
| 9.0% | 21.74 | 19.33 | 17.39 | 15.81 | 14.50 | 13.38 |
| 10.0% | 25.09 | 22.30 | 20.07 | 18.25 | 16.73 | 15.44 |
| 11.0% | 28.43 | 25.27 | 22.75 | 20.68 | 18.96 | 17.50 |
| 12.0% | 31.78 | 28.25 | 25.42 | 23.11 | 21.19 | 19.56 |
| 12.5% | 33.45 | 29.73 | 26.76 | 24.33 | 22.30 | 20.58 |
| 13.0% | 35.12 | 31.22 | 28.10 | 25.54 | 23.42 | 21.61 |
| 14.0% | 38.47 | 34.19 | 30.77 | 27.98 | 25.65 | 23.67 |
At Ke = 6.85% (between the 6.5% and 7.0% columns), current ROE of 12.6% implies a fair value in the range of S$27–29. Even if ROE falls permanently to 11.0% — below any level OCBC has delivered in recent years — the P/B framework gives S$22.75–S$25.27 at plausible Ke assumptions, straddling the current price rather than clearly below it.
Risks — What Would Change My View
1. NIM stabilises above 1.80% by H2 2026. If NIM finds a floor earlier than expected — driven by a slower Fed cutting cycle or MAS exchange rate policy — DDM values rise materially and the bull case becomes the base. Watch: OCBC’s NIM guidance at Q2 2026 results, expected 7 August 2026.
2. WM fees reverse. If global markets sell off — Middle East escalation, trade war reignition, a credit event — AUM shrinks, product sales fall, and non-interest income retreats toward ~29–31% of total income (the FY2022–2023 range, the last time peak NIM crowded out fee income). At that level, the wealth pivot thesis is unproven and ROE compression would look more structural than cyclical. Watch: non-interest income as a share of total income in Q2 2026 results.
3. ROE structurally falls below 11%. The P/B bear case assumes 11%. If ROE falls below that level and stays there for two or more consecutive halves without a clear recovery path — rather than reflecting a one-quarter rate or tax effect — the P/B base case collapses toward S$24–25 and the market’s 9.95% implied ROE starts to look less like pessimism and more like accuracy. Signal: annualised ROE trend across H1 and H2 2026.
4. Capital return plan not renewed. The Double Dividend kicker is a finite, one-cycle event. Once the ~S$800 million is distributed — whether as buybacks or specials — the recurring income picture reverts to ordinary DPS of S$0.83, and the DDM floor reasserts. If OCBC does not announce a successor capital return programme, total yield guidance deteriorates noticeably. Watch: any commentary on capital policy beyond the current plan at FY2026 full-year results.
5. HSBC Indonesia acquisition mispriced. If AUM transfer disappoints or integration costs escalate, it undermines both the near-term earnings story and the management’s execution credibility — two things the market will be watching closely. Watch: any investor day or integration update post-completion, expected approximately mid-2027.
The numbers bracket a wide range because the core question remains genuinely unresolved: does a bank with a 50% dividend payout ratio and 12–13% ROE deserve to be valued on its dividend stream or on its total earnings power? If the wealth engine is as durable as Q1 2026 suggests — and if Singapore’s demographic shift toward a super-aged society provides the structural tailwind OCBC is explicitly positioning for — the P/B framework is the right anchor, and at 1.71x book, this bank looks mispriced under most plausible ROE assumptions.
If NIM continues falling and WM fees prove cyclical rather than structural, the DDM is closer to correct, and the current price looks roughly full — with a finite kicker from the capital return plan that evaporates once FY2026 closes.
The Reverse P/B finding is the sharpest way to frame the choice. OCBC has not delivered sub-10% ROE in recent memory. The market is pricing in permanent impairment. I hold a position because I think the impairment case is harder to make than S$22.92 implies — but I hold it with the five conditions above clearly in view. If any of them materialise, I would revise.
[^rf]: This article uses Rf = 2.10% (SGS 10-year bond yield, 15 May 2026, sourced from Trading Economics). This differs from the DBS analysis published on this site on 14 May 2026, where Rf = 3.2% was used in error. The SGS 10-year yield has traded in the 1.7–2.2% range since late 2025; the 3.2% figure was not reflective of prevailing rates. A correction note has been published on the DBS article. All figures in this OCBC analysis use the correct rate.
Disclosure: At time of writing, the author holds a long position in OCBC (SGX: O39) of approximately 1,100 shares. This holding was disclosed prior to and during the research process. The analysis reflects the author’s independent view and should not be read as a recommendation.
Disclaimer: This article reflects the author’s personal analysis and opinions, written in a strictly personal capacity. It is not financial advice and does not take into account any individual reader’s financial situation, investment objectives, or risk tolerance. Information is sourced from publicly available filings as of the date noted but accuracy cannot be guaranteed. The author may hold positions in securities discussed (see disclosure above). Readers should conduct their own research and consider consulting a licensed financial adviser before making any investment decisions. The author is not a licensed financial adviser under the Financial Advisers Act of Singapore.
An ordinary man with extraordinary inspiration