Keppel stock is up 77% in 6 months and pays a 3.4% dividend yield – Will I buy this “AI” stock?

So I received a question on Keppel stock recently.

I won’t replicate the whole question for confidentiality reasons.

But this person essentially bought Keppel earlier this year, is up a good amount, and is thinking whether it would make sense to sell (or buy more).

Keppel stock is up 77% in 6 months

Here’s the Keppel chart for reference.

Keppel went as low as $5.70 in April earlier this year.

This week it closed at $10.09.

Which is a 77% increase in slightly over half a year.

Yes I know there are stocks on the NASDAQ that do multiples of this.

But considering this is a blue chip stock listed on the SGX, it’s still pretty impressive.

Will I buy Keppel Stock?

But a chart is just a chart.

To truly understand where the stock is going next, we need to understand more about the business of Keppel.

Understanding Keppel’s business today

Keppel’s business today is split into 3 parts:

1. Infrastructure
2. Real Estate
3. Connectivity (eg. Telco like M1)

Infrastructure is the money maker

Infrastructure is the biggest money maker – and the reason why the stock is up so much this year.

What exactly does Keppel’s infrastructure business do?

Keppel’s Infrastructure business develops, owns and operates essential energy, environmental and water assets, and sells integrated power and decarbonisation solutions.

It runs power generation & retail, district cooling, waste-to-energy, and desalination/NEWater operations, and packages these as asset-light “infrastructure-as-a-service.”

Why is Keppel stock is up 77% in 6 months? Keppel as an “AI” stock?

But the reason why Keppel stock is up so much?

Is because whether rightly or wrongly, the market seems to think of Keppel as an “AI” stock.

The same is playing out in the US with electricity generation companies.

The thinking is that AI data centres consume a lot more power than traditional data centres.

With the amount of new AI data centres being constructed, that is demand for a lot of additional power.

Benefits whoever owns and can produce more electricity – including power generation companies.

How much AI upside potential is there for Keppel in reality?

I suppose you will ask just how much AI upside potential is there for Keppel in reality?

Yes no doubt that Keppel has won a couple of data centre contracts.

But if you look at the net profits from infrastructure – it’s generally flat vs last year.

So if there is a lot of AI upside potential, this isn’t really showing up in earnings yet.

Personally I have my doubts on how much Keppel will benefit from the AI build out.

A lot of the build out has been taking place in US so far.

How much AI data center buildout will we see in Singapore?

And how much of this Keppel can capture?

There are a lot of unknowns at the moment.

The stock has run a lot in a short time, and whether that will be supported by earnings I have much more mixed views.

Followed by real estate – a distant second

Then we have the real estate business.

This is the plain vanilla real estate business so nothing much to talk about here.

Yes real estate delivered nice profits vs last year, but frankly real estate in today’s market is just no longer viewed as a “sexy” business.

Sale of M1 is good for Keppel

And then we have connectivity, which is mainly the Telco business.

Connectivity will be much smaller going forward with the sale of M1 to Simba.

What I would say though, is that I think the sale of M1 is a good move for Keppel.

This is highly competitive business with not a lot of growth and not a lot of pricing power.

The fact that they can free up almost $1 billion on their balance sheet, and use that money to focus on their core infrastructure business.

That to me is a big win.

What is the ultimate vision for Keppel?

Big picture – what does Keppel want to be in 2030?

In one sentence – a leading global asset manager + operator – focussing on fast growing sectors like digital infrastructure.

What does that mean in plain English?

I suppose they want to do what CapitaLand is doing.

In going asset light, focussing on the fund management business to generate strong recurring income.

And hitting ROE above 15%.

But focussing primarily on infrastructure (or other fast growing infrastructure sectors).

Now is that a vision I can get behind?

H*** yes.

Remember how I said real estate is not sexy this decade?

Well… infrastructure is.

Just look at how all the big PE funds like Apollo and Blackstone are going all-in into infrastructure assets like power generation and ports.

Which is why Keppel executing this asset light strategy for infrastructure has a chart that looks like this.

And CapitaLand doing the same for real estate looks like this:

But… some realism is required – lots of work needs to be done for Keppel’s transformation

That being said.

Keppel is still in the process of transformation, and a lot more work needs to be done.

Looking at the hard numbers – you can see how the bulk of the funds under management continues to be old school real estate via REITs.

Yes Keppel wants to change this over time to focus on sexy high growth areas like digital infrastructure, but this is a process that will take time to play out.

Valuations of Keppel stock – is Keppel “cheap”?

At today’s price of ~S$10.09, Keppel trades around:

• 18.3x TTM P/E
• ~26x EV/EBITDA
• ~1.8x P/B
• ~3.4% trailing dividend yield

Whether that is cheap or expensive depends very much on how you see Keppel today.

If you see Keppel as a hybrid infrastructure / real estate player, I think those valuations are at best fair value, at worst expensive.

But if you see Keppel as an AI power generation play, that is well placed to ride the massive wave of capex coming into digital infrastructure in the years ahead.

Then Keppel actually looks cheap at these prices.

It’s all a matter of narrative.

Technical Analysis – wait for a pullback?

That said, the charts show Keppel in a strong uptrend.

Very strong in fact, to the point where it looks overextended here relative to the 50 day moving average.

From a technical analysis point of view looking at this chart.

It may make sense to wait for a pullback before adding, if you were so inclined to add.

Keppel stock is up 77% in 6 months and pays a 3.4% dividend yield – Will I buy more of this “AI” stock?

Bottom line for me.

If you buy Keppel stock at these valuations today, you pretty much believe that Keppel will benefit from the capex boom into AI data centres.

Okay, you don’t actually need to believe it.

You just need to believe that the market believes it.

And so far at least, that belief would have made you money in 2025.

But look at the chart for the AI stocks, and you can see why many people are calling AI a “bubble” that will burst eventually.

Is AI a bubble that will burst?

My view – whenever we have a transformative technology like AI.

Humanity almost always overbuilds capacity.

Whether it was railroads during the industrial revolution.

Or internet bandwidth during the dot com bubble.

We always went into the mania phase, investors overbid, stocks went crazy, too much capacity was built, and eventually everything came crashing down.

The difficulty is knowing where you are in the cycle.

Sitting out the AI story in 2025 would have been a painful mistake, and I’m not ashamed to say that a good chunk of my 2025 portfolio returns similarly came from riding this “bubble”.

But where are we in the bubble – how much more room is there to go.

Boy that is a really hard question, and I’m not ashamed to say I have no clue.

What I have been doing though.

Is that from a risk management perspective – I have been steadily taking profits on the way up, and I have been rotating into “cheaper” names with hopefully less downside when s*** hits the fan.

But like I said, that’s a big if.

This AI bubble can end tomorrow, or it can go on for another 12 months.

Who knows.

So… will I buy Keppel stock?

I think bottom line.

Even if I want to buy, I will wait for a pullback.

The chart looks overextended here, and I would be wary buying into Keppel stock like this.

But if we get a pullback – would I buy?

I think it really depends on where we are in the AI cycle.

If it’s clear that the AI bubble has burst, then probably not.

But if the AI mania is still in full swing, then hey Keppel could be an alternative way to play the AI story.

Love to hear what you think though!