AI Megatrend; Blessing or Bubble?

To wrap up our 2025 November investor webinar, we're delighted to have Rudi Filipek van Dijk, editor and founder of FN Arena, a website of financial business and economic news analysis and data.

And Rudi will be talking about the AI megatrend and whether or not it's a blessing or a bubble.

Great to have you, Rudi.

Over to you.

It's a great pleasure to be here.

And the irony of this conversation, of course, is that we are sort of celebrating three years in the mega trend of AI.

And instead of throwing parties and commemorating the whole event, there's actually a lot of anxiety that has crept into the share market, both locally here in Australia and overseas as well.

So I thought maybe a little bit of context and of explanations could go a long way to address those anxiety and see whether we actually are, are we living?

the early days of the Nasdaq meltdown in 2000 or of the big sell-off in 1987 or, heaven forbid, the late 1920s and the early 1930s.

I'm quite a little bit biased in this one.

I do believe that the answer to all three questions is no.

But I also believe that clearly people can be wrong in these things and often they're easier to assess in hindsight.

And the other thing also which I think is equally important is that

Bull markets and trends, even mega trends, they don't unfold and develop in straight lines.

And even I'm old enough to have from close distance experienced the 1990s bull market with the whole internet that came into a foyer.

Even in those days, the market did on occasion have a stumble and have a sell off.

and then continue this path again.

So it wasn't a straight line between 1995 and the year 2000.

And there's no reason to expect that this time will be different in that regard.

So maybe the big challenge for investors is to distinguish the fact that the bull market can run too hot a little bit on occasion, but that doesn't necessarily mean that we have seen the end and the fall of the AI megatrend and it all can disappear.

in a heap, even though in the current context, a lot of people throw them into one basket.

I think it's probably very well for investors, very good to distinguish between the two.

And if we do get a share price weakness, we obviously have some weakness in Australia, post the August reporting season, but in the US markets have continued to travel on, so to speak.

If we do get that weakness at some point, and it may look scary in the short term, I think it's very important to distinguish between the two, and it doesn't necessarily mean that there's no money to be made anymore, or we shouldn't have any exposure anymore to the air mega trend.

Before I dig into the finer details of the local market, let's have a step back and go to the United States where essentially the epicenter of AI, where it's born and where it lives more than it does in Australia.

What we've seen over the past number of years is that AI, I mean, not only in the share market, but also in the economy has become very, very important.

And if today,

commentators and investors are still wondering why the US economy has been so resilient, even with tariffs and other headwinds that have come towards it.

I would point towards AI.

On some estimates, AI-related spending is responsible for 90%, if not more, of all capex spending in the United States over the past few years.

If that is correct, then it doesn't take too much imagination to see how much that has supported the economy, of course.

Because AI has been such a strong force, it has led to a market concentration.

And the market concentration, by all means, it is quite extreme.

But we also should take into account that when we invest in the share market, we invest essentially in corporate profits.

and corporate profits have been enormous amongst those companies who, A, have benefited from AI, but on the other end is also the ones who are financing AI.

So the concentration is high, but the concentration has always been high post the GFC, and that's already more than 10 years ago, that's 15 years ago.

And so it has been technology-led.

Those technology companies have been dominated the market over the past decade and a half.

They are still doing it today.

And we should take into account that that is an important distinguishment between the situation today and a comparison that is often made, which is the late 90s, early 2000s, when a lot of those companies trading on very high valuations essentially had no earnings.

They had revenues, but no earnings.

Today, the situation is different.

It is carried by earnings.

And while most people are very, very anxious

that the eternal rise of companies like Google, Alphabet, Microsoft, Nvidia and others can be quite scary.

It needs to be pointed out that even in today's market, the forecast for Nvidia is still to grow something along the lines of 56%, which is well above that most companies are growing at.

And it's not like we're putting that share price on 80 times next year's profits or something along those lines.

It's much more moderate.

In an ironic way, it can be argued that a general anxiety amongst investors might actually be the best recipe to prevent that we are actually going into a situation where we were in the Nasdaq meltdown in March 2000, for example.

Now, let's take that back to Australia, and the local market is not as dominated by AI as the US.

but AI is still important locally, and the anxiety is not less in Australia.

And I think there's two elements to this, and as I pointed out earlier, they are better distinguished from each other.

One is that the share market itself is at an all-time high, and the average PE ratio is at a level of about 20.

It's a level we normally don't see in Australia.

That's a level we normally see in the United States.

So you can see why investors are getting a little bit anxious about this.

The other element, of course, is all the doubts that are creeping in with the circularity and the magnitude of AI, of the investments itself made.

Now, let's stick to the first point for the time being.

Yes, the local market is at an all-time high.

It has a very high P-E ratio.

But the anxiety on my observation also comes from the fact that people are just simply looking at those P-E ratios and at the level of the index.

They don't look at what actually makes up underneath the surface those high P-E ratios.

And a simple way to point out is that

12% of the index locally now is made up by Commonwealth Bank shares.

Commonwealth Bank shares, because foreign investors are putting their money here, is trading at a PE that we've never seen in our lifetime.

Now, is Commonwealth Bank in a bubble?

That remains to be seen, but even if Commonwealth Bank is in a bubble, does that mean that the rest of the market is in a bubble?

And that's obviously, I would argue, no, it isn't.

So if you reduce or you take into account that the Commonwealth Bank, of course, on a P of 27 or something along those lines, well above average, it pushes the average up.

The other element, which I think a lot of investors haven't caught up with either, is that we increasingly in the slipstream of technological changes in the United States, technology has become increasingly important for the local market as well.

One simple example is that at the latest index re-weightings, Technology One has now joined the top 50 here locally.

Now, Technology One, one of my favorites, it's a stock that trades at something like 48 times next year's profits.

You can see with Technology One joining the likes of Y-Stack, of Xero, of ResMed, of REA Corp, of Promedicus,

that the average PE ratio for the market logically goes up over time and has been higher in recent times than it has been in the past.

I have a chart which we can show in a few seconds.

And what this chart doesn't know or doesn't do is correcting for everything I've just mentioned.

But it does correct for the resources sector, which has also a very disturbing impact on the average PE.

And what we can see from this chart is that valuations post-COVID have been higher than the years before, but they have not necessarily been higher than they are today.

So

Saying that valuations today are cheap would be warm, but saying that they are excessively valued also seems incorrect, because the valuations today are not that different from two or three years ago.

Valuations are higher than the past, and understandably for investors who haven't accustomed to that, that creates a lot of anxiety.

In particular, the market like Australia, which is a typical value seeker market,

also because of all the commodities and energy stocks.

Now, let's go to AI now, because AI is creating a lot of anxiety.

And I think one of the major reasons for this is that the numbers involved are absolutely ginormous.

I mean, ever since we had the GFC, the numbers that are being bandied around

is probably best summarized by a trillion here and a trillion there, and you throw a trillion on top.

And before you know it, we are talking real money.

And the situation is not that different for AI.

I mean, the numbers that are churned out and invested by the likes of Microsoft, Alphabet and Nvidia and others, they are truly, truly tremendous.

But here, too, we need to see this in context.

And some

Economists have already pointed out that in a relative sense, because what we're now seeing is that those companies also are much more profitable.

They generate much more cash flows.

Economies in general are much larger than they were in the past.

If you combine all those numbers now, we are still investing, or not we, but they are, less than 1% of GDP in the US.

Historically, that's not that enormous.

That number has at times been superseded in the past.

So if you put it into context, yes, the numbers are tremendous, they are enormous, but they're not that enormous, tremendous when put in the right context.

And it's good to remember

that the financing is done by companies who are ginormous, very, very profitable, they have very big balance sheets, and they churn out cash like there's no tomorrow, which also is a big difference between the disasters we've seen in the past.

Now, another element, oh, by the way, I happen to have a chart for this one as well.

And we can see this one comparing

with, for example, the stimulus that the American company put in place after the Second World War to get the European economy back on track.

What you see in this chart is that today's investments, with those investments back then, so important, absolutely dwarfed by the numbers that we are throwing at AI today.

But the companies doing it will assure us that this is necessary to get the infrastructure up and to get the system going.

Coming back to the technology itself, we are back in science fiction.

I mean, people are basically making forecasts that by tomorrow, none of us has a job anymore, and AI is going to create robots, and we're all going to be extinct before we know it, and we're just going to rule the world.

On the other end, we have a lot of skeptics.

I mean, it's not working.

It's a mirage.

The progress is not fast enough.

I think the truth is somewhere in between.

And for example, we are users of AI here at Efren Arena, not with the aim that we will have one day a website that runs itself and we don't need humans anymore.

We're just using it as a tool for us and to make our life easier, to make things quicker, sort of like an electronic slave, so to speak.

And we've seen

the progress that has been made over the past year, year and a half or so.

And that progress already has been quite enormous.

And I think in those terms is what investors need to see it for the time being.

I've been lucky enough an investor is good to highlight here.

We have one company on the stock exchange, which I just mentioned, Technology One.

They are amongst the first in the world to develop a whole suite of products built around and embedded inside AI.

I've seen the product, I've seen the presentations, I've seen the enthusiasm, and it's not going to replace us as humans, but it's going to make life much, much easier of all the users, the customers, the clients they have for their software.

And essentially what it does, it does dumb things very, very efficiently.

And who doesn't want to have dumb things done in his daily tasks at the office and outside the office in a very, very efficient way, which I think is why their product will resonate with their customers and why it will be successful.

And I think this is how investors need to assess AI at this point in time.

It's not going to shoot to the moon, but it will become very, very popular amongst people like you and me.

On some estimates, half of the internet is already written through AI.

And I see people completely complaining already on social media that too many posts are written by AI.

What AI will do, of course, it will create winners and losers.

but it also goes through phases.

And the most important thing to remember here is that I still remember, I'm old enough, when the first mobile phones came out and they looked like a brick.

I mean, if you put it in your jacket, you were leaning towards where the mobile phone was.

If you were skeptical in those days, you could have never predicted that

20 years down the track, look where we are now with the mobile phone.

We have smartphones and we do banking and everything through it.

You would have never predicted it in those days.

Similarly, when I went to university in the 80s, we had these Olivetti desktops.

They were absolutely atrocious and you would not expect in similar fashion that 20, 30, 40 years down the track, we have all internet, we have internet banking, we have everything that happens over the desktop.

the same thing is going to happen with AI.

And I think the best way to compare AI in a historical context is not to compare it with the PC or the laptop or the mobile phone, but to compare it with electricity, with power.

And that was the big driver underneath the

the bull market in the 1920s.

So the comparisons with canals in the UK and with trains in the United States, I think they are one.

I think it's a better comparison to make with electricity, with power.

And when we got the first power connections in the United States, we would have never predicted how powerful the tools and applications would become a century later that are developed on the back of that.

There is, that's a long winded way to say that we actually cannot predict accurately today where we are going with this new mega trend.

And yes, that makes some people a little bit uncomfortable because we are spending literally hundreds of millions of dollars and they will ultimately become trillions if you add it all up.

And there needs to be a return mate.

Well, the return mate, that's also a very moot point.

And I've seen quite some people making some analyses and predictions that don't really make much sense if you know better, so to speak.

And one of the problems also that Australian investors have to deal with is that

They've never dealt with technology stocks because they didn't have to.

But now technology stocks become ever more important part of the market and they will have to get their head around in how you value those technology stocks.

And it's not done in the same fashion as a cold producer or as BHP.

In the same respect, data centres are not a technology company.

I mean, I think the ASX has done the likes of NextTC a disservice by embedding them in the local technology index.

It's better to see them as an infrastructure play.

And AI helped me, by the way.

I asked a question recently to AI.

I said, when did

a Transurban list on the stock exchange and how long did it take before Transurban recorded its first profits?

The answers are 1995 and 19 years.

And I never heard anyone in those 19 years say that Transurban was in a bubble or was too highly valued.

Certainly not on that basis.

Now, the alternative view, if we

if we don't take the doom and gloom scenarios for granted, the alternative view is that just like in every other bull market, we will see pullbacks at some point, and we may well see one at some point.

But this technology is going to change the world, no matter whether we like it or not.

And as investors, we just have to take into account that even though the anxiety after three years is now almost reaching a crescendo,

is that there is a scenario which is equally as plausible is that 2025 today might be the equivalent of 1995 or might be the equivalent of 1925, which means that we're only halfway maximum in the development of this trend.

And that means that any pullbacks, any retreats, any corrections in the share market

on that premise basically would provide us with opportunities to get on board of a trend that has many more years in its tracks.

And I think that's a scenario that is equally important to consider for investors.

Dial down the social media, I would say, maybe because there's a lot of noise over there.

How then, as an investor,

How should we treat this?

Because we don't have ready to make answers.

Just waiting for us to be unveiled by someone.

No one's ringing the bell when the opportunity is there or when the end of the trend is coming.

I think the answer is diversification.

Don't put all your money in one basket.

And it's very good.

to have a diversified both geographically and over mega trends and other trends, and maybe even spread it out that not everything is AI related.

And let's just hope for all of us that we are not reliving the late 1990s or the late 1920s.

Because if I remember correctly, when these things came to their downfall,

There wasn't much to hide and definitely not in the share market.

So I would like to keep a more positive echo, echo in voice here and believe that whatever comes will ultimately end up in a positive scenario.

Thank you very much for that, Rudy.

You've pointed out quite a few technologies that I can remember also.

So it's a bit of a trip down memory lane, particularly that mobile phone into the pocket, tipping you left to right, depending which side it went into.

For our first question, Rudy, I would like to ask you, so as you pointed out, some investors remain unconvinced.

Maybe they're sitting on the sideline, maybe they're waiting.

What would you say to them about the potential for investing into AI-related industries and companies?

I would say to those investors, keep an open mind and don't necessarily go down the tunnel of it's all bad, it's all a hype, it's all a mega hype and it's going to collapse in a heap.

The other thing to remember also is there's not one way of playing AI.

There's multiple directions.

It doesn't have to be

in the chip manufacturer.

It doesn't have to be in the consumer of the chips.

It doesn't have to be in the builder of the infrastructure.

It doesn't have to be in the...

If AI develops as we currently think it will,

It will have many, many influences.

It will spread around and ultimately affect every part of the economy, every country.

So at the same time, it's almost like the old thing that people say about politics.

If you're not interested in politics, politics will be interested in you.

And I think if you try to ignore AI, AI will not ignore you.

But equally important is that even if you don't

want to invest directly in AI, is AI will ultimately will come to you.

Because as this technology develops, it will create winners and losers in the share market.

And in particular for those investors who are always attracted to buying beaten down, cheaply valued companies, it may well be that at some point the market has decided that

those companies are going to be losers because of AI.

And that means that cheap will be cheap for a very good reason, and you maybe don't want to own them.

So I think it will be inescapable, irrespective of what happens in the short term.

I think ultimately, we will all have our focus on AI regardless.

So don't, don't put your head in the sand, I think is, don't be the ostrich.

A second question for you, Rudy.

Obviously, making forecasts is difficult.

But if you had to take an educated guess, how do you see AI developing maybe in the next one to two years?

What do you think is coming just next?

That's quite interesting because I think it's very good to understand for people also, hence why I said you don't have to invest directly in what is important in the AI sphere right now, because this technology will develop.

And at this point in time,

the building of the infrastructure is all important because everyone wants to be on board, in particular everyone who counts in today's world and those are obviously the mega companies we have and also we should not forget the governments want to be involved as well.

So now it's all about building out the infrastructure but

That's not going to be eternally the epicenter of this development.

It will go to platforms and to technology companies.

I do believe those companies and those sectors will be the first to successfully

develop products and be able to sell them to their customers and therefore make revenues and profits out of it.

But ultimately we will go through sectors of the economy and as things stand now, expectations are very high that AI will prove its value for financial sectors.

We have to think not just asset managers but also banks and insurance companies.

It will clearly also go for those platform companies like REA Group, car sales, comparison websites and even the likes of Hub24 and NetWealth, that type of company.

And then ultimately it will probably go to healthcare companies and also to companies that do a lot of testing, data and testing.

So you have to think about the likes of IMDEX and ALS Limited in the market as well.

And I think it will just continue.

As I said earlier, every company ultimately is probably going to be forced to have some kind of investment in AI.

Not everyone will be successful at doing it.

At this point in time, a lot of companies are not that successful just yet.

but then everyone is also not yet on board.

The interesting thing is I picked up is that even BHP and Rio, and they are commodity companies, even they are already very much engaged with AI and BHP is about to roll out some of their AI solutions and products they have across their

they're all the businesses they have globally, so there is something happening.

In my presentation earlier, I did mention that I have another chart which I picked from a recent presentation and that actually stems from the August reporting season and that also harks in with my observation that we are more likely to be at the beginning stage of AI than at the end stage because August in Australia

was the first corporate results season where companies actively started talking about AI and how they employ it and everything.

And I've just put together a short list of all the companies and it's incomplete, but it gives people an idea how varied already the sectors and the companies are that started talking about AI in August.

And my prediction is in February next year, in August next year, that number will only increase

And if we go back to February this year or August last year, the numbers were very, very small.

So that's what I think is only getting traction now in Australia.

And that feeds also in the view that there's a lot to come yet and it's early days yet.

Thank you very much for your insights today, Rudi.

That was absolutely fascinating.

Thanks for having me, Dara.