We're on Bubble-watch. Stay active.

We are somewhere in a bubble

We really believe that we are somewhere in a bubble and we would definitively call it that. And of course with bubbles you can have mid cycle pull backs.

If we look at an analogy here and I like this chart because it's actually pretty simplistic.

What we're doing here is we're just plotting the S&P cumulative return, the tech bubble versus now.

So in the Navy blue line, you see the launch of Netscape actually in August 1995.

And similarly we have the pink line, which is the launch of ChatGPT in November 2022.

And you can see quite a similar trend here, strong upward cycle.

In both instances we've had mid cycle pullbacks, so back in ‘98 and of course last year with tariffs and now this year and in the first quarter as well with concerns around the Middle East.

That said, we fundamentally believe that there is further for this market to run and why is that?

Multiple reasons we can talk through, but one is actually the monetary stimulus is still feeding through with a transmission mechanism.

We have the rate cuts coming through from last year.

Secondly, as the fiscal spend coming through and the one big beautiful bill so-called, this has been genuinely quite supportive and the US is running a deficit of 6%, which we can discuss the sustainability of that, but is giving a genuine fiscal impulse to the economy on the demand side.

In addition, we are seeing elements of speculation and frenetic moves.

We do think we're seeing indications of retail hype, meme stocks for example doing pretty well again, profitless tech performing very well last year as well.

And the top 5% of beta, the really sort of more speculative growth, these stocks doing relatively better.

Fascinating example is Allbirds, the sneaker company that you know was very trendy back in 2021 in Silicon Valley and then had its time and kind of went out of fashion and was almost written off in investors minds.

And they then said a couple of weeks ago, look, we're pivoting. We are now going to be a compute company and we're going to buy GPU's. And the share price of that company was up 600% in one day. So clearly elements of speculation here and signs of a bubble.

Don’t leave the party early

Leaving the party can be just as bad as not going. And actually, if you look back to the tech bubble and the NASDAQ performance, over 50% of the four year total NASDAQ return occurred in the final year.

So what are we doing here?

We're very fortunate to have such a broad platform of analysts who have the time and the capacity to be on the ground meeting management teams, building bottom up models and forecasts.

And what we're doing is we're focusing on participating where the fundamentals really justify that is companies where we're seeing real earnings upgrades coming through, real pricing power management teams that we trust to allocate capital effectively throughout the cycle and market share gains in this competitive environment.

So we have to navigate this responsibly. And we're very fortunate to have the the platform to be looking at all these companies in this excitement.

Our bubble-watch framework in action

What we've done is we've said, look, we need a framework to make this systematic and not emotional.

So we've studied a lot of prior bubbles to look at the conditions for bubbles and also conditions around which bubbles do end.

And there are five key points that we would draw upon here.

And as you can see on the right hand side, we have sort of we're taking the temperature on each of these.

We've got a bit of a traffic light system; green looking more promising, red high risk, a little bit more flag for concern. On interest rates, bubbles do not pop with interest rates flat or going down.

Historically, they only tend to pop when rates go up.

Now of course, US, Iran does add a little bit of nuance here.

But that said, when we look at everything coming out of the Fed and we speak to our macro analysts, they will tell you that this is a supply side shock and hiking rates is not going to change the price of oil unfortunately. That is not how the transmission mechanism works.

In addition, the Fed does have a dual mandate and it has to focus on both employment and output as well as inflation.

And so we need to be cognizant that whilst unemployment is in a very good place right now, we may ultimately see some pressures from some of the cuts being put through thanks to the productivity gains.

So right now, we're feeling neutralist on rates, a little bit worse versus the start of the year.

But on our side, we're not expecting an aggressive hiking cycle right now in the US in particular.

What about the broader economy and employment?

Of course, bubbles do tend to pop when we see recessions and that is something to watch.

But as I mentioned, actually the global economy was in an encouraging position before this kicked off.

And we don't invest in economies, we invest in companies, in bottom up stocks.

And if you look at the earnings of these companies, they're still accelerating.

We're seeing the all country world index, we're seeing high teens expectations for the next year as in EM, we're seeing higher 28% growth expected on a one year view.

So we're still seeing encouraging signs coming out of companies as they report and no imminent signs of a recession.

What about the AI super cycle?

When we look at historical bubbles, of course, when the tech price does not come through in reality, that can cause disappointment.

Scott and I have spent a fair amount of time meeting the companies this year, but also on the ground in Silicon Valley.

And what is fascinating is actually this is a piece where we feel even more constructive than at the beginning of the year.

One example is if you look at the scaling laws, the models that are coming out are only getting better and better and more powerful.

And you can see that in Mythos, for example, the release from Anthropic, where it was deemed so powerful that actually they did not want it to be yet released to the general public.

And it's being used to shore up cybersecurity for some dominant companies.

So of course, we need to watch for risk of CapEx cuts.

But actually, this week, even with the MAG 7 results, we've seen further CapEx increases.

So one to watch, but an area we still feel very good on.

What about leverage? Leverage is #1, where again, if you see excess leverage and of course, a hiking rate cycle. #2, that tends to be bad for bubbles.

Net, we want to run a lot of data on this front. And sovereign debt is a place to monitor with higher fiscal debt to GDP. But private debt is still in a very good pace.

Households have deleveraged substantially. Post to GFC.

And whilst we're seeing pockets of stress in private credit, when we really dig into it, there's nothing we see that is genuinely systemic on that front.

And then very lastly, geopolitics. As I mentioned, we are running a lot of scenario analysis here, but ultimately this is a place where we feel worse unequivocally versus the start of the year.

So that is one reason why we have pared back the beta of the portfolio at the margin from 1.13 coming into the year versus our core benchmark to 1.08 at the end of the first quarter.

Well, it's remained on Wall Street and what we're seeing there with AI remaining the driver there, but increasing concerns around spending later to broader rotations and sharper swings.

To break it down, we are joined by Sam Ruiz from T. Rowe Price. Sam, welcome back to you.

Good morning, Andrew.

I was just asking you before you came on the show, how, how has your month been?

So you feel as though you're sort of getting ahead of the market at the moment?

Yeah, I mean, you’re effectively alluded to it.

There's only two things that matter right now.

It's how you play the AI trade that is driving everything, not just earnings, but economic growth and it's how you hedge inflation.

So what are you expecting comes from Iran, whether it's resolution or not?

I think that our positioning at the moment is reflecting that we need to have a little bit more sort of commodities in the portfolio to hedge that inflation risk.

But we've successfully transitioned to where we think the better opportunity in AI is right now.

And that's away from the company's spending on AI and it's more towards the companies that are receiving those dollars from those spenders.

All right. Because when we take a look at that AI CapEx, it remains watering, eye watering, doesn't it?

Because in fact, I was just talking about this with our previous guest in terms of that at least CapEx, PIMCO analyst, they're coming up.

They reckon that CapEx will absorb 94% of hyperscalers operating cash flows over the next two years.

Yeah, and that is the concern for us. So these hyperscalers have great businesses.

Why have they been so good over the last 15 years as companies, really strong free cash flow, really big net cash balance sheets and they've been able to recycle that free cash flow into high recurring revenues and profits.

Now there's this concern and this is really our view, it's something we debate a lot actually, is we think that the hyperscalers will continue to spend money whether they can monetize this successfully or not. And why is that?

Because it's now a competitive and defence imperative that if they don't spend the money, someone else is going to compete and take their share.

That means that the CapEx numbers, we think, stay stronger for longer, but it's at the cost of those companies spending money and reducing free cash flow and potentially now going out to the market and raising debt and raising equity.

Increasing leverage makes the quality of those businesses potentially a little lower, while the question marks around how they monetize it continue to grow.

Yeah, In fact, big tech new debt issuance now already around close to 1 1/2 billion dollars.

What you're saying is they're going to continue spending until what point?

So a number of things. So these hyperscaler businesses are really dependent on economic growth.

They get their revenue, they get their demand from areas like digital advertising and they have to receive that revenue to be able to recycle that free cash flow into the CapEx.

Well, it's not really free cash flow that's going into CapEx.

The other thing is what we call scaling laws.

So there's a there's models, we sort of think of Claude's model, open AI's model.

And right now, for every incremental dollar that's spent on training those models, the models are getting better.

If that stops, we realize we've reached the peak, then there's no incentive to spend as much money to keep training new models.

The other thing is that's really important is to see continued monetization of these models.

So what is, I just believe, truly staggering?

If you think of Anthropic, they really had no revenue in 2024.

They now look like they're going to exit this year at potentially $100 billion annual run rate of revenue.

This is a company with no revenue two years ago that's now going to have more revenue than the biggest sort of enterprise software player like Salesforce.

If any of those things slow down or stop, that is going to put a bump in the road in terms of whether these companies are incentivised to spend more.

All right.

So how you change your portfolio?

Actually, I'm just going to correct myself too, because I said I've missed the decimal point on that amount of new debt issue.

It's 135 billion this year so far.

How are you shifting your portfolio then as a result if you're concerned about that spin?

Yeah. So our, if you were to call it MAG 7, it's not really a MAG 7 anymore. It's more like a MAG 10.

We have reduced our weight to those companies.

Right now. We are more incrementally overweight.

Some of the hardware players now there's a couple of transitions here that are playing out.

It's the spenders on AI with great stocks and that's sort of already started to filter through to who's receiving those dollars.

But what we're realizing is there were sort of key suppliers such as NVIDIA that were in the box seat to receive that money because they effectively had a monopoly on the GPU or the cheap or what's needed.

What we're realizing now is that this CapEx that is being spent is such an enormous tidal wave of money to the point where this sort of even in the last few weeks, we've seen some of the CEOs of some of those largest companies in the US.

And they've told us that they would have spent even more than the surprising figures they announced if they could get access to more of the components, whether it's the connectors that are going in, whether it's the memory that's effectively going into the data centers.

What that means now is you know that more money is likely to come for the next two years.

But also, you might have said two years ago, we're only going to buy chips off of NVIDIA.

They're the best.

Now you're like, they can't give us enough.

We're going to filter to a second best.

So I'll give you a great example.

We actually really like a company called AMD.

AMD does not have as good of the GPU as NVIDIA, but if you can't get enough GPUs from NVIDIA, you're more likely to now go down to the next best and go acquire that from someone like an AMD.

That's the sort of thing that we're starting to figure out.

Companies that didn't have the likely expansion of growth are actually coming ahead of us now.

Sam, what about the space trade you're getting on that?

Space trade is a really interesting one. We were there. We've actually already opted out.

So in our portfolio we owned SpaceX privately, we actually bought it closer to a $30 billion $40 billion valuation and exited around 400 billion, which looks a little light now given that we're going to see it come on in, in mid June around 1 1/2 trillion.

The concern for us is there is definitely an economic model there, but the valuation surrounding that seems quite excessive to us.

So that's one on IPO, we'd be more willing to sit out and just wait to see how the market assesses whether that valuation is something that the company can grow into or not.

What about the other forthcoming IPOs, Open AI, Anthropic?

Yeah.

So these are great companies.

The concern we have is not necessarily there's not much information yet around the valuations that are coming into the market.

It's these are going to be multi trillion dollar companies coming into the market at IPO, which is astounding, raising records level of, of capital.

We haven't seen this amount of issuance before.

And this is going to have to suck capital from somewhere else in the market.

So I, I mentioned that we're a little bit more conservative around some of these MAG 7 type mega cap tech stocks because fundamentals are potentially weakening.

There's also going to be a trade here where investors and big funds need to say, well, if I'm going to have to own these stocks because they're going into the index.

And a little side point here, these IPOs don't normally get into the index day one.

Normally they have to have a number of quarters of showing financials before they make the cut.

It's likely we might see these come into the indexes immediately and that money has to come from somewhere, most likely the mega cap tech stocks.

Sam, what do you think is the greatest motivator at the moment on Wall Street?

Is it that the fear of missing out on this, on the AI boom, or is it the fear of losing looking back to the dot com bubble and thinking, hang on, we've been here before?

Yeah, I, I think it's a little bit of everything.

The one thing I'd really emphasize is a lot of this rally has actually been supported by really strong fundamentals.

So if we looked at the last earnings that just came out of the US market, call it Wall Street or whatever you will, this was one of the biggest upgrades and surprises to expectations.

We saw year over year earnings growth of 28%, which is enormous for the US market.

And it's no surprise following that we had a couple of the best months in April that we've seen in 20/25 years for various US indexes that are out there.

There is a bit of retail participation here, some speculation, a lot of leverage creep.

If you look at margin lending levels, they're actually the highest they've almost ever been, which are a little bit worrying to us.

And that speculation is it's a really important distinction.

We think there's an AI opportunity, but we're really trying to focus where we're invested with companies.

We think we'll have the cash flow are profitable today where the margins have a little bit more duration into the future and aren't something that's you know really profitable today, maybe loss making in two years time.