Stocks

OCBC is cheap (undervalued) at $32? – Why I disagree despite being a shareholder

September 12, 2026

Okay so this one is a little different.

Over the weekend, an article popped up on my Google feed with the title “OCBC is cheap (undervalued) at $32.27. Not Kidding.”

Now OCBC is my biggest bank position, so naturally I wanted it to be right.

I mean, why would I not want this position to double from here.

But wanting something to be true is not the same as it being true, so I sat down and ran the numbers myself.

And I wanted to share the numbers with you guys – whether OCBC is truly undervalued.

In this article, I wanted to discuss:

What does the article actually argue?
What does an objective, first-principles look at the numbers say?
Would I add to OCBC at S$32, and what am I doing with my own position?

Table of Contents

What the article argues – 6 reasons OCBC is still undervalued at S$32

To be fair to the author, it is a well-argued piece.

The starting point is the price action – OCBC has gone from a 52-week low of S$16.19 to S$32.27, so is anyone buying today simply too late?

The article says no, for 6 reasons.

One – the valuation models still say upside. A discounted cash flow model puts fair value at S$39, and the article cites a broker consensus target of S$34.

Two – OCBC is becoming less of a bank. Non-interest income jumped 51% in Q2 2026, and the bank is rolling out AI-powered wealth platforms and hiring 600 more relationship managers.

Three – the re-rating is OCBC catching up to DBS, not irrationality. OCBC started 2026 at 1.5x book against DBS at 2.4x, and the gap was never justified given OCBC’s asset quality.

Four – there is a century of Singapore property sitting on the balance sheet at cost. OCBC’s 2018 annual report showed OCBC Centre at 65 Chulia Street with a carrying value of about S$20 million against a market value of about S$1.1 billion – plus shophouses across the CBD.

Five – the dividend story isn’t finished. A 50% payout on growing earnings means the dividend grows on its own, and there is still about S$800 million of the capital return programme to come.

Six – OCBC isn’t just a Singapore story. Greater China and Malaysia profits grew 31% and 21% in FY2025, and Bank of Ningbo contributes over S$1 billion a year.

The author does qualify the article – the yield has compressed, margins are easing, and analyst targets sit below the price.

But the verdict is that he / she is still adding at S$32.27, because after 7 years of dividends the effective cost per share is only S$6.29 (personally I’m not a big fan of this style of thinking, to me the “cost price” of a stock is the market price today, because that is the opportunity cost I am giving up by not selling).

But for what it’s worth, I agree with a good chunk of the article – OCBC is a better business today than the market gave it credit for at S$16.

The question is whether that is already in the price.

Let’s take a deeper dive.

For transparency before we go further – OCBC is my biggest bank position, followed by a smaller position in UOB, and I do not hold DBS today.

A bank is worth its book value times a multiple – and the multiple is the whole game

Let’s reason from first principles.

Strip a bank down and only 2 numbers matter for the share price: its book value, and the multiple the market pays for it – and that multiple is set by how much the bank earns on its equity, its ROE.

A bank earning 11% on its equity should not trade at the same multiple as one earning 17%, and it doesn’t – that’s why UOB is at 1.3x book, OCBC at 2.3x, and DBS at 3.2x.

Here’s the chart for reference.

Chart 1 — OCBC price-to-book, year-end 2016 to 2025 and September 2026.

In plain English – OCBC has never been this expensive in at least a decade.

The 10-year median is about 1.1x book, and today it is at 2.3x – up from 1.4x at the start of the year, and roughly where DBS itself started 2026.

Now a high multiple is not automatically wrong, if the ROE has moved up to match.

OCBC’s ROE was 13.7% in the first half of 2026 – a good number, but the same 13.7% it earned in FY2023 and FY2024, when the stock traded at 1.0 to 1.2x book.

In plain English – the market is now paying double the multiple for the same ROE.

The price is not saying “OCBC earns 14%” – it is saying “OCBC will earn a lot more than 14%, for a long time”.

How much does OCBC need to earn to justify 2.3x book?

To pin this down, I pulled the numbers for 14 banks across Singapore, Asia, Australia and the US, and plotted ROE against price-to-book.

Chart 2 — Price-to-book against return on equity for 14 banks, September 2026.

You can see the pattern – the more a bank earns on its equity, the more the market pays for it.

The banks above the line get a premium beyond what ROE alone justifies – and OCBC is now one of them, alongside DBS and Commonwealth Bank of Australia.

To sit on the line at 2.3x book, a bank needs to be earning close to 19% on its equity.

OCBC earns 13.7% – which the line says is worth about 1.9x book, not 2.3x.

In plain English – at S$32, you are paying for DBS-level returns from a bank that has never delivered them for a full year (its best was 17.9% in 2012, and it has averaged about 12% since 2010).

That doesn’t mean OCBC can’t get there.

But it does mean the price already assumes it will.

And interestingly – note where UOB sits on the chart above.

Testing the 6 reasons against the data

So with that framework, let’s go through the 6 reasons one by one.

The valuation models. I could not find the S$34 consensus target anywhere – the consensus figures I can verify sit between S$29.65 and S$32.30, at or below the current price. The S$39 discounted cash flow number comes from a model that was showing S$31 in January – it chased the price up. And CGS International downgraded OCBC to Hold in July on valuation alone, with a S$29.80 target.

In plain English – the models say fairly valued, not cheap.

The transformation. This one is real – non-interest income is now 44% of total income, up from 36% a year ago, and wealth fees were up 39% in the first half. That said, the Q2 step-up was carried by trading income (a record S$695 million, up 60% on the previous quarter) and insurance – the kind of income that comes and goes with markets. The durable part, fee income, grew 26%.

Very good – and largely why the stock doubled. But to a large extent it’s already priced in.

Catching up to DBS. As discussed above – OCBC trades at 74% of DBS’s multiple, and earns 78% of DBS’s ROE. The discount is already proportionate to the earnings gap – to close the rest of it, OCBC’s ROE has to close first, not sentiment.

The hidden property. The most interesting argument, so I ran the numbers. OCBC carries about S$4.0 billion of property, plant and equipment and S$0.7 billion of investment property at cost. Suppose the true value is S$5 billion higher – a generous assumption. That is 3.5% of a S$145 billion market cap, and it takes price-to-book from 2.35x to about 2.2x. And OCBC plans to redevelop the Chulia Street site, not sell it.

Real, but too small to matter at this market cap.

The dividend. 50% of a growing pie is a growing slice, agreed. But the yield at S$32 is about 2.9% on the ordinary dividend, or 3.5% counting the one-off 18 cent special expected with the full-year results – the lowest of the 3 banks, against about 4.1% for DBS and 3.8% for UOB.

The dividend is safe. But I would hardly say it’s spectacular at this price.

Asia growth. Greater China is 21% of operating profit and Malaysia 14%, so the regional story is genuine. But the HSBC Indonesia acquisition only completes in mid-2027, and Bank of Ningbo is a mature 20% stake the market itself values at 0.8x book. Growth, yes – a game changer inside 3 years, maybe not.

The S$6.29 effective cost. I understand the psychology – years of dividends do change how a position feels.

But really, I’m not a fan at this kind of thinking.

To me, for every day you hold a stock, you’re actively making a decision not to sell at the market price – and therefore giving up the opportunity cost of that money.

Viewed this way, the true cost of a stock – is the market price it trades at today.

Holding onto your position at S$32 today earns exactly the same return as for someone buying their first share.

But that’s just how I see it, and I know many investors disagree.

Long story short – 4 of the 6 arguments are true but to a certain extent already priced in, 1 is too small to matter, and the valuation-model argument is 50/50.

Never miss a post! Follow Financial Horse on WhatsApp or Telegram — one tap, and every new article reaches you the moment it’s published.

What has to go right for OCBC at S$32 – and what history says about it

Here’s the thing about buying a bank at a 10-year-high multiple – the business can do fine and you can still lose money, because the multiple does the heavy lifting either way.

Just look at how AI stocks have been trading of late to understand how this can go.

So I ran a simple 3-year exercise: OCBC keeps growing book value at about 6.5% a year (what a 13–14% ROE with a 50% payout delivers), pays about S$3.40 of dividends including the special, and you sell in 2029 at a given multiple.

This is what the return looks like:

Chart 3 — 3-year total return on OCBC bought at S$32.27, at each exit price-to-book.

In plain English – if the multiple holds at 2.3x, you make about 9–10% a year, which is a perfectly decent outcome.

If it drifts to 2.0x – roughly where CGS puts fair value – you make about 4% a year, which is basically the dividend.

And if it goes back to 1.4x, where OCBC ended 2025, you lose money over 3 years despite the dividends.

The operating case barely moves the answer – this is a bet on the multiple.

So do re-ratings like this stick? It depends on whether the ROE keeps going up.

DBS re-rated from about 1.2x book in 2020 to 3.2x today, and it has stuck – because ROE went from 10% to 17.5% along the way.

Commonwealth Bank of Australia hit 3.3x book on a 13.9% ROE in mid-2025, and fell 20% in the second half of the year when earnings flattened.

And OCBC’s own previous peak of 1.24x book in 2017 took 7 years to be seen again.

Premiums stick when ROE keeps rising, and revert when it plateaus.

Which is why the next 2 or 3 quarters matter, and the variables I’m watching are:

Net interest margin – 1.70% in Q2, down 22 basis points year on year; management guides for it to stabilise in the second half. That needs to happen.

Wealth fees – up 26% in the first half. If that holds without the trading windfall, the ROE plateau case weakens.

Credit – OCBC’s NPL ratio is a clean 0.9%, but UOB’s Greater China NPL ratio jumped from 3.5% to 4.8% in the same quarter on a Hong Kong commercial property account.

Capital – fully phased CET1 dropped to 14.0%, management’s operating floor, which caps loan growth and buybacks at the same time.

In plain English – if you think OCBC is now a 15–17% ROE bank, S$32 is fair and the article is right.

If you think 13–14% is the run-rate, the price already embeds the best case.

Why I’m not selling either

That said – none of this is a reason to sell.

The stock is in a powerful uptrend – at S$32.27 it was about 38% above its 200-day moving average (as at 4 Sep 2026) – earnings beat in Q2, guidance was raised, and analysts are still revising estimates up.

My historical mistake has been selling winners too early on valuation – a stock at a record high on record earnings can stay there a lot longer than the valuation math suggests.

I can run all the valuation analysis I want, but the trend is up and the earnings keep delivering, so a more practical view is required.

Momentum is a reason to hold, not a reason to add.

What I’m doing with my own money

As shared above – OCBC is my biggest bank position, followed by a smaller position in UOB, and I do not hold DBS today.

As much I would like for OCBC to be undervalued, after looking at the numbers I can’t help but think the more objective conclusion is that it is fairly valued at best, overvalued at worst.

Personally I am holding onto my position, but I am not adding in a big way at these prices – at 2.3x book, the risk-reward for fresh capital is nothing like what it was at S$16 to S$20.

The trend is what I’m watching, and the Q3 results in early November are the next real test – if NIM is stabilising and fee growth holds, the ROE plateau argument gets weaker, and I’ll revisit.

Long story short – if you ask me the article gets the business right and the price wrong.

OCBC is a better bank than it was a few years ago there’s no doubt about that – but at S$32, the market already knows that, and a good part is already priced in.

But hey that’s just me, I would love to hear what you think – OCBC has doubled in a year. Are you adding at S$32, holding, or taking profit?

This article was written on 8 Sep 2026. It will not be updated going forward.

My live views on OCBC — including the price at which I would buy — are updated weekly on FH Premium, alongside my full watchlist and personal portfolio.