Stock analysis series · SGX: O39
Oversea-Chinese Banking Corporation (OCBC)
A four-part look at Singapore's second-largest bank: how good the business is, how it has come through past crashes, what the price chart says now, and a Qimen timing view. Written in plain language, as the output of our framework.
Share price S$29.14 · 21 Oct 2026Analysis date 21 Oct 2026Latest results 1H 2026
For education and information purposes only. This is not financial advice, and the Strong / Neutral / Weak results are not a recommendation to buy, sell or hold any investment.
Part 1 of 4 · Fundamental analysis
The answer in one minute
12-month target S$27.50
- Singapore's second-largest bank. It earns from lending (63% of income), fees (16%), trading (12%) and life insurance through Great Eastern (7%).
- Business is better than ever: a record S$4.19b profit in the first half of 2026 (+13%), even as interest margins fell, because wealth fees and insurance grew strongly.
- Very safe: bad loans are only 0.9% of loans and the CET1 capital ratio is 15.7%.
- The problem is price: the shares nearly doubled in a year and now cost 2.18x book value, against an average of 1.24x since 2010.
- At today's price you would earn about 3.8% a year in dividends. Our models value the shares at S$24-29. Confidence: Medium.
How healthy is the business overall?
Good business, full price: 65 out of 100 (the six scores below add up to 39 of 60).
Think of it like a school report card: OCBC scores top marks for quality and safety, but loses points because the share price already reflects much of the good news.
Strong business and balance sheet; valuation is the weak spot.
Green is strong, amber is fair, red is weak. Valuation scores only 3 because you pay about twice the usual price for each dollar the bank owns, like buying a good house at the top of the market.
Where does every dollar come from?
Mostly lending, but fees, trading and insurance now drive the growth. 55 cents of every dollar of profit is earned in Singapore.
By income source, FY2025 (total S$14.6b)
By geography, FY2025 profit before tax
A bank borrows cheaply from depositors and lends at a higher rate; the gap is its main income. When interest rates fall, that gap shrinks, which is why OCBC's fee and insurance businesses matter more each year.
Does the business keep making more money?
Yes, strongly until 2024. 2025 paused as interest rates fell, then the first half of 2026 set a new profit record.
Total income (S$b)
Profit per share (S$)
Dividend per share (S$)
Profit per share rose 52% in four years and dividends rose 87%. The small dip in 2025 came from lower interest rates, not from bad loans: the share of loans gone bad fell from 1.5% to 0.9% over the same years.
Is it cheap compared with its neighbours?
No: you pay almost twice its usual price for each dollar of book value.
Price / book value (lower = cheaper)
Dividend yield (higher = cheaper)
DBS costs even more but earns a higher return on equity (17.5% vs 13.7%), so it earns its premium. UOB looks cheaper. OCBC sits above the price its own returns would justify.
What could the price be in a year?
Most likely around S$28.50, a little below today; a bad year could mean S$18.15.
Today S$29.14 · probability-weighted target S$27.50 · dividends of about S$1.12 come on top
We weight three stories: bad (25%), normal (50%) and good (25%). The possible fall is larger than the possible rise from today's price, which is why the result is Neutral rather than Strong.
What could go wrong?
A fall in the price investors will pay for its book value is the biggest danger.
- Valuation falls back toward 1.5-1.7x book
- Interest margin keeps shrinking
- Economic slowdown and bad loans (Greater China property, energy shock)
- Wealth and trading income fall with markets
- Geopolitics: US-China, tariffs, Middle East
- Competition from DBS, UOB, private and digital banks
- Regulation: capital rules, dividend caps in a crisis
- Weaker regional currencies (MYR, IDR, HKD)
- Technology outages, scams and cyber attacks
Items in the red squares deserve the most attention. The business risks are real but well covered: OCBC holds S$1.63 of reserves for every S$1 of bad loans.
Strengths and concerns
Strengths
- Record 1H 2026 profit of S$4.19b (+13%) despite a lower interest margin
- Wealth income S$3.29b (41% of income); client assets S$350b (+13%)
- Great Eastern insurance profit up 49%
- Bad loans 0.9%, unchanged for about nine quarters; reserves 163% of bad loans
- Interim dividend up 15%; S$2.5b capital-return programme; FY2026 dividend could reach about S$1.10
Concerns
- 2.18x book value vs a 1.24x average since 2010
- Dividend yield only about 0.7 points above 10-year Singapore government bonds
- Interest margin fell from 1.92% to 1.70% in a year, with more pressure expected
- 2Q 2026 profit was flattered by record trading (S$695m) and insurance gains
- Fully phased-in capital ratio fell 1.2 points in one quarter to 14.0%
What to watch next
- 3Q 2026 results on 6 Nov 2026: analysts expect about S$1.97b, roughly flat year on year
- Net interest margin: below 1.65% would be a negative sign
- Credit costs above 25 basis points, and any new Greater China property bad loans
- Fully phased-in CET1 below 13.5% would mean fewer special dividends
- FY2026 final and special dividend (Feb 2027); a price below about 1.8x book would be more attractive
Mini glossary
- Net interest margin (NIM)
- The gap between what the bank earns on loans and pays on deposits, as a % of its loans and assets.
- Book value / P/B
- What the bank owns minus what it owes, per share; P/B compares the share price with it.
- Return on equity (ROE)
- Yearly profit as a % of shareholders' money in the bank.
- CET1 ratio
- The bank's core capital cushion against losses; regulators set a minimum.
- Bad loans (NPL)
- Non-performing loans: loans that are not being repaid on time.
- Credit cost
- The year's charge for expected bad loans, in basis points of total loans (100 bp = 1%).
- Special dividend
- A one-off extra dividend on top of the normal payout; it may not repeat.
Part 2 of 4 · Resilience test
The answer in one minute
2 pass, 1 borderline
- Profit per share grew about 7.4% a year for 20 years, and book value per share about 6.8% a year.
- S$10,000 invested 20 years ago became about S$97,800 with dividends reinvested (12.1% a year).
- It has crashed five times since 2000, by 26% to 59%. The worst was -59% in the 2008 financial crisis.
- With dividends reinvested, every crash was regained within 9 to 18 months of the bottom.
- Today it is 10% below its September high. The biggest risk is the high starting price, not the business.
How did it do on the three tests?
Two passes (growth and recovery) and one borderline (how deep it falls).
Compounding
PassProfit per share +7.4% a year over 20 years and +6.4% over 15; total return 18.5% a year over 10 years vs 10.6% for the STI ETF. New shares diluted holders by about 1.6% a year.Crash depth
BorderlineWorst fall -59% in the 2008 financial crisis; the median of five crashes was -44%. None caused lasting damage to the business.Recovery time
PassWith dividends reinvested, all five crashes recovered within 9-18 months of the bottom (median 14). On price alone, two took 6-7 years from the peak.Three health checks: does it grow, how hard does it fall, how fast does it get up again? OCBC grows and recovers well, but like all bank shares it can fall hard in a panic.
Did S$10,000 grow?
Yes, strongly: to about S$97,800 over 20 years with dividends reinvested.
S$10,000 invested in Jan 2008, by Oct 2026
S$10,000 with dividends reinvested, to 21 Oct 2026
Before tax and costs. The STI ETF benchmark is only available from 2008.
Dividends doubled the result. But the last 12 months alone added about 80%, mostly because investors are now willing to pay more for each dollar of book value. A fair long-run expectation from history is about 10-12% a year, not 18%.
Did the business get bigger, and did each share's slice get bigger?
The pie grew about 5.7 times; each slice grew about 4.2 times.
The pie: total net profit
Your slice: profit per share
Total profit grew 9.1% a year; profit per share 7.4% a year, because the number of shares rose about 37% (mainly the 2014 rights issue for Wing Hang Bank).
Like a pizza that grows while a few more friends join the table: each slice still grew a lot. That is moderate dilution, not value destruction.
How often has it crashed, and why?
Five times in 26 years, always in market-wide or sector storms.
Dot-com bust, 9/11, SARS
-48%back in 1.5 years with dividendsGlobal Financial Crisis
-59%back in 0.8 years with dividendsEuro debt crisis
-26%back in 1.2 years with dividendsOil & gas loans, China slowdown
-32%back in 1.2 years with dividendsTrade war, then COVID-19
-44%back in 1.1 years with dividendsEach time, the bank kept earning and its book value per share fell in only two years out of twenty. Most of the falls came from investors suddenly paying less for each dollar of book value: price storms, not broken businesses.
How deep did it fall and how long did recovery take?
Deep falls, quick recoveries with dividends; on price alone, two took six to seven years.
Depth of fall (price, peak to bottom)
Months from bottom to recovery (with dividends)
Our pass mark is 36 months (3 years). On price alone, the 2000 and 2018 peaks took 85 and 72 months to regain.
An investor who bought at the May 2018 peak waited about 3 years to be whole with dividends reinvested, but almost 6 years on price alone. Dividends matter most in slow recoveries.
Where are we now?
A 10% dip, not yet a crash.
Today S$29.14 is about 10% below the 52-week high (S$32.57 intraday, 4 Sep 2026).
If investors priced OCBC at its long-run average of about 1.16x book value, the price would be about S$15.52, 47% below today. Treat that as a valuation stress test, not a forecast and not a chart support level.
What to watch next
- Price-to-book above 2.0x while profit-per-share growth stays below 5%: raises the risk of a deep fall
- Credit costs above 30 basis points or bad loans above 1.3% of loans: would signal real business damage
- Net interest margin below 1.6%: would weaken profit growth
- Fully phased-in CET1 below 13.5%: limits dividends and raises the chance of new shares
- Any rights issue, or share count growing more than 2% a year
Mini glossary
- Compounding
- Growth that builds on itself year after year, like interest on interest.
- Price return / total return
- Price only, or price plus dividends reinvested.
- Crash / drawdown
- A drawdown is how far the price is below its last high; a crash is a drawdown of 20% or more.
- Book value per share
- The bank's net worth divided by the number of shares.
- Dilution
- More shares sharing the same profit, so each share gets less.
- Stress price
- The price if valuations fell back to long-run or past crash-low levels.
Part 3 of 4 · Technical analysis
The answer in one minute
for a new entry at S$29.14
- The long-term trend is still up: the price is 18.6% above its rising 12-month average.
- The short-term trend turned down on 7-8 October: a 10% fall in two days on about four times normal volume, after a broker downgrade and a jump in US bond yields.
- The price now rests on the first floor at S$28.53-28.93. Selling has faded, but there is no buying surge yet.
- Ceilings are close (S$29.78-30.26, then S$30.66-31.46); the next solid floor is about 13% lower at S$25.26-25.55.
- Big event soon: 3Q 2026 results on 6 Nov 2026. A daily close below S$28.50 turns the chart unfavourable; above S$31.46 on rising volume, favourable.
Where does the trend point on each timescale?
Up long term, pausing in the middle, down short term.
Up
Higher highs and lows since 2000; 10.5% below the S$32.57 all-time highPausing
Below the 10-week average, but 14% above the rising 40-week average (S$25.55)Down, stabilising
Lower high on 6 Oct, broke the September floor; holding S$28.5-28.9Like weather: the climate (monthly) is warm, but this week's forecast (daily) is stormy. When timescales disagree, the longer one usually wins, but timing should respect the short one.
What has the price done over the last year?
Up 73% in 12 months, far ahead of the market, but down 9% in the last month.
OCBC price change to 21 Oct 2026
STI ETF price change, same periods
Price only, not including dividends. Bars use the same scale (73.5% = full width).
OCBC led the three Singapore banks over the past year (DBS +38%, UOB +21%). In October it fell further than the market: the leadership is being tested, not yet lost.
Where are the floors and ceilings, and what yield would I lock in?
Each floor lower gives a higher dividend yield: 3.9% at S1, 4.4-4.6% at S3-S4.
Think of the yield as the rent you collect on your money (based on about S$1.12 of dividends over the next 12 months). Below S1 there is a thin "air pocket" down to S3. *The stress level comes from Part 2: a valuation stress test, not a chart floor.
What could happen next?
Most likely back-and-forth between S$28.5 and S$30.3 into the 6 November results.
Daily close above S$30.26, then above S$31.46 (50-day average) on volume above 1.2x average
S$31.85-32.57 (R3), then about S$33.4Price holds S$28.5-30.3 through the results
Sideways for 2-6 weeks while the 20- and 50-day averages flattenDaily close below S$28.50 on volume above 1.2x average
S$26.4-26.7 (S2), then S$25.3-25.6 (S3) and S$24.3-25.0 (S4)No percentages are given: chart judgement alone cannot produce reliable odds.
An earnings miss, a further rise in bond yields or a market sell-off could gap the price through S$28.50. A close back above about S$31 would put the daily chart back in line with the long-term uptrend.
So what does our framework suggest for a long-term investor?
Part 1 is Neutral and the chart is Neutral: wait for deeper floors and build in steps.
| Part 1 result | Chart favourable | Chart neutral (today) | Chart unfavourable |
|---|---|---|---|
| Strong | Act now | Staged entry, first step at S1 | Wait for S3/S4 |
| Neutral (today) | Small starter at S1 | Wait for S3/S4, build in steps | Stand aside |
| Weak | Stand aside | Stand aside | Stand aside |
Short-term trades fail our 2-to-1 reward-to-risk test right now: buying the dip at S1 offers only 0.8 to 1 to the first ceiling, and a breakout needs a new all-time high above S$32.71. Expect that some steps may never fill.
What to watch next
- OCBC 3Q 2026 results on 6 Nov 2026 (DBS reports on 5 Nov)
- A daily close below S$28.50 (chart unfavourable) or above S$31.46 on rising volume (favourable)
- US 10-year Treasury yield: it reached 5.32% on 7 Oct and has pressured bank valuations
- A weekly close below the 40-week average (about S$25.6) would mean the intermediate trend has turned
- Next review of this chart: after the 6 Nov results
Mini glossary
- Support / floor
- A price area where buyers have stepped in before and stopped a fall.
- Resistance / ceiling
- A price area where sellers have taken over and stopped a rise.
- Zone
- Levels are price ranges, not exact lines.
- Moving average
- The average closing price over recent days, weeks or months.
- Reward-to-risk
- Possible gain to a target divided by the loss if the stop is hit; we want at least 2 to 1.
- Staged entry
- Building a position in slices at lower floors instead of all at once.
Part 4 of 4 · Qimen forecasting
The answer in one minute
Direction: Caution
- The window supports buying, but in three stages, not all at once.
- The buyer's position reads strong, and the longer-term profit picture has backing through this year: steady and dependable rather than spectacular.
- The weeks to 31 December look slow and noisy: mixed and sometimes misleading messages, and an older concern may be raised again.
- Let the 3Q results on 6 Nov 2026 and any dips decide when to add the second and third parts.
- Confidence: moderate. The opportunity is unlikely to slip away before year-end, so patience costs little.
The question for this reading
"Should I buy OCBC stock, and at what prices, before 31 December 2026?"
This part uses Hour Qimen (Chaibu method) to read timing for this one question, cast on 21 Oct 2026 at 8.35 pm. It does not replace Parts 1 to 3: the fundamentals decide whether the business is sound, the chart decides where the levels are, and the Qimen reading adjusts when to act and how much.
How does the timing look across the window?
Slow now, with room to add on weakness before year-end.
Buy a little only
At most the first of three parts, near the recent price. The price is likely to drift rather than climb.Check before adding
Read OCBC's own results announcement first. If a significant new or returning problem appears, put the later parts on hold.Add on weakness
Add the second and third parts only if prices come down or the results support it. Holding one or two parts by year-end is fine.Like crossing a busy road in light rain: you can start walking, but you take it step by step and check at each kerb. Here, the kerb is the 6 November results.
What does the reading suggest doing?
- Split your amount into three equal parts and buy no more than one part at a time.
- Write your reference prices down before you place any order: about S$28.80, S$27.40 and S$25.90 (the 20 Oct close, then about 5% and 10% lower). Change them only for a clear reason, not because of a headline.
- Before adding the second part, read OCBC's own results announcement due on 6 November 2026, and check whether any older issue has come back.
- Treat this as a steady, longer-term holding for the bank's business strength and dividends, not a quick trade to close before 31 December. No borrowing, and no money you may need soon.
How does it fit with Parts 1 to 3?
The reading agrees with Parts 1 to 3 on the main point: build a position in stages, not all at once. It is a little more willing to start now: its reference prices (about S$28.80, S$27.40 and S$25.90) sit above Part 3's long-term steps (S$25.40, S$24.60 and S$21.30). A cautious way to combine them is to keep any first part small and near the S1 floor (S$28.53-28.93), and let the deeper parts lean toward Part 3's levels. The reading does not change the verdicts of Parts 1 to 3; it only adjusts timing and size.
Disclaimer. This Qimen reading is my own opinion, shared for information and education only. It adds another dimension of detail, a traditional Chinese view of timing, alongside the analysis in Parts 1 to 3. It is not financial advice and not a recommendation to buy, sell or hold OCBC or any other investment. Markets can move against any reading, so please do your own research and speak to a licensed financial adviser before you invest.
Putting the four parts together
Great bank, full price: reading the results side by side
The four results look at different things, so they do not contradict each other. Part 1 (Neutral) rates the business and its value today. Part 2 (compounder with crash risk) shows how the shares have rewarded holders over 20 years. Part 3 (Neutral) is about entry timing over the next few weeks, and Part 4 (Conditional Yes) adds a timing view for the period to 31 December.
A modest fit today
About 3.8% forward yield (4.4-4.6% at the S3-S4 floors), the lowest of the three local banks; dividends per share grew 8.9% a year over 20 years.A good fit, at the right price
Profit per share grew 7.4% a year for 20 years and S$10,000 became about S$97,800 with dividends, but 2.18x book leaves little room for disappointment.S$15.52-18.15
Part 1 bear case S$18.15 in 12 months; Part 2 stress test S$15.52 (about 47% below today); past crashes of -26% to -59%.Valuation and the 6 Nov results
Shares cost 2.18x book vs a 1.24x average, and the interest margin is still falling. All figures are estimates and may change with the 3Q results.Important: This post is for education only and is not financial advice. The Strong / Neutral / Weak assessments, scores, targets and probabilities are the output of our framework, not a recommendation to buy, sell or hold. The Qimen reading in Part 4 is a traditional timing view and my personal opinion. Past performance is not a reliable guide to future returns. Consult a licensed financial adviser before making investment decisions.
Data: share prices to 21 Oct 2026; fundamentals to the 1H 2026 results (30 Jun 2026). Sources: OCBC results releases, 2025 Annual Report and Historical Financial Highlights; Yahoo Finance price and dividend data; DBS Research; Beansprout; The Kopi Notes; Financial Horse; JournalArta; Smartkarma; SmartWealth (SingStat inflation). Full source lists are in each detailed report.