Eric Lascelles - Managing Director, Chief Economist and Head of Investment Strategy Research

Hello and welcome to our latest video #MacroMemo. We have lots to cover this week. We'll talk, as seemingly always in recent months, about the war with Iran and what's been happening there. We'll talk a little bit about a shifting monetary regime. We're going from rate cuts to maybe tentatively rate hikes over a period of time. We'll talk a bit about the semiconductor boom and the stock market and how we're feeling about that.

We'll continue on the AI front and talk about whether artificial intelligence is going to be a winner take all, with one or a few companies dominating, or if there's room for more. We'll also talk about artificial intelligence in the context of unemployment concerns. And lastly, we'll just give a nod of the head towards Canada's recently announced sovereign wealth fund.

That sounds like an awful lot so we better get going. We'll start just with Iran and won't spend a great deal of time here. If you're looking for good news, the ceasefire is still holding nearly a month after it first took hold. We'll certainly take that anytime a war has functionally stopped. That is good news. Usually that's the end of a story for a war, but we're not quite there on this one.

And that's because there is mixed news in the sense that negotiations for an actual peace treaty are proving elusive and difficult. It’s hard to say exactly when that will be reached, though we do still think there is scope for a deal. If you're a pessimist, if you're looking for the bad news, it is, unfortunately, that the Strait of Hormuz, at least as I record these words on May 4th, is still blocked.

Of course, that last one is of greatest relevance to energy prices, to the impact on the global economy and so on. It is still quite negative.

One way we're thinking about this is that this war arguably presented a buying opportunity for investors when stocks were down, let's say 10%, in late March, at the moment of maximum pessimism. It does seem like rather less of a steal right now with some stock markets having fully rebounded and even setting records in a North American context.

Certainly this is not to suggest those are inappropriate and not at all to downplay the contribution from rising tech sector profits and expectations, and so on. But it is fair to say there is still some pain to come from this war. The war still needs to be resolved. There is some lagged economic and inflation damage that we need to budget for.

There is likely going to be some spillover, via transportation costs, into other things. Even fiscal positions have been hurt as some subsidies come into place and it takes months to normalize. And so not to suggest we've swung in a particularly negative direction, but just whereas there seemed to be quite a buying opportunity for investors, let's say a little over a month ago, that buying opportunity arguably seems to have passed at this point in time.

And we are perhaps of a more neutral view as to how rapidly and well this is resolved relative to where the market is right now.

In terms of other subjects of related note, it is worth mentioning that we've actually seen a surprising degree of government support, specifically for consumers, in the context of this energy shock.

And it struck us that it was probably more widespread. It certainly was more widespread than we had initially realized, probably more widespread than the average investor realizes as well. And so just sorting through the consumer supports that governments have put on just to compensate consumers for the high energy prices, first of all. Not in the U.S., but it's really the exception here.

Almost everyone else of note has done something, and so it has reduced the hit to some extent. Many countries have cut their fuel tax. So Canada has done that, as well as Germany, Spain, Australia, Brazil, Turkey, India, Vietnam. Indonesia and Thailand have both the implemented fuel subsidies, which gets you to the same sort of outcome.

And then you actually have some countries that have gone even more strongly than that. And they have actually, to varying degrees and in varying ways, actually capped the price of fuel altogether for retail customers. Japan has done that. Mexico, Poland and China, notably, South Korea very notably as well. I mentioned that because the country that is doing this to varying degrees should see less inflation than you would think, because a lot of it's being eaten essentially by the government. It should therefore suffer less economic damage than you would think, because again, it's being eaten by the government.

They will suffer more fiscal damage, they will have bigger deficits than otherwise. And so it’s just worth being aware of what's happening and I would say on the net it’s diminishing the amount of global economic damage. But again, at the cost of more government borrowing.

Okay. Let's pivot over and let's talk about central banks for a moment. And so we have been or had been in a monetary easing regime for a number of years.

Call it 2024 through to early 2026. It seems like we are shifting out of that. Not a lot of central banks are cutting rates anymore. Some central banks are thinking about raising rates, some indeed are even raising rates. The Reserve Bank of Australia is a prominent developed world example. It does look as though we are shifting now into a monetary tightening period.

Not to overstate that, we don't think rate hikes are going to come fast and furious here or be particularly near term for most countries. Nevertheless, as we look forward, at least across the major developed world central banks, we have the U.S. Federal Reserve on hold, it should be noted. However, we have rate hikes from the Bank of Canada, from the European Central Bank, the Bank of England, the Bank of Japan, 2 or 3 hikes over the next year or so.

The basic logic here and kind of the shift in terms of what's happening is twofold, on the economy and the inflation side. On the economic side, growth is just holding up. Global Purchasing Managers’ Indices (PMIs) have gone up, not down, even through this war in recent months. Economic data itself is pretty good to quite good in some cases.

We still see the growth tailwinds we've been blabbing on about now for several months, in terms of fiscal support and AI investment and AI productivity and these sorts of things.

And then on the inflation side, of course, there has been an inflation shock from this energy shock. It is temporary. And so central banks quite rightly should look through most of that.

But now that it's lasting a little longer and we're talking about three months and perhaps even a bit beyond, that does increase the scope for some spillover via transportation costs into some other products. And we don't have oil prices coming all the way back down to where they were before. So you end up with central banks not able to quite completely ignore these consequences, even if they can look through most of them.

The takeaway is rate cuts don't seem as appropriate and maybe there will need to be a little bit of modest tightening in various markets.

From an investment standpoint, we had been talking about how over the past few years, it was actually a very favorable investment regime. Economic expansion plus rate cuts is historically the best time to be a stock market investor.

Quite a good time to be a bond market investor. If we are shifting into a regime of rate hiking paired, we think, with economic growth, historically, that's still a pretty good time to be a stock market investor. Historically, it's okay for bonds, but it's less attractive for bonds because of course, some interest rates are actually rising.

Okay. On to the next subject, the semiconductor boom. Certainly it's well appreciated the tech sector has been very strong in the stock market.

I think less appreciated is how heterogeneous the performance has been because, of course, beneath the surface, particularly in recent months, you have software, stocks at an out right down. You have computer chip stocks or semiconductor stocks that have been way, way up – to the point that they are now themselves 16.7% of the S&P 500. Just the computer chips, not the big hyperscaling companies that use them, not the software companies and so on.

Fundamentally, there is a very good story here. AI CapEx is rising incredibly. Much of those data centres, they're all relying on computer chips, but much of the cost is those semiconductors. And, to the extent that supply is constrained in the chip sector, the profit margins have been surging and so on.

So there's a very real story there and quite remarkable. We would warn – not with too much of a force – but we would warn that the technicals aren't great for this sector now. There is some vulnerability to a pullback given the signs that we're seeing. Fundamentally, valuations are quite stretched.

For instance, price-to-sales are now 17.4 times. That's price-to-sales not price-to-earnings. And so that's nearly three times normal for the sector – way higher than the software sector or hardware and equipment, the other bits of technology. The computer chip industry is famously very cyclical with big booms and sometimes big busts.

And so really the debate here is you do have profoundly strong fundamentals in AI as this incredible force for change. But, with challenging technicals, challenging valuations, some history of cyclicality. And so really the advice here isn't buy or isn't sell. It's more about managing risk in this environment because this sector is perhaps particularly vulnerable to volatility right now.

Risk management is key, we think, at this point.

Let's continue down the AI path. Is AI a winner take all? Many tech sectors are. You look at social media dominated by Meta, or search dominated by Google, or online retail dominated by Amazon. And it's very possible AI will be the same with just one or a handful of dominant players.

That's very much the way that the mega hyperscalers are behaving. Investing so much, feeling they must be first to the finish line. And there is some logic to that. There are such massive upfront investment costs for AI and being able to access user data, having a big user database first is very important. So it could be that this is another winner-take-all.

And we're seeing some of the classic mergers and acquisitions (M&A) associated with those sorts of outcomes. We would just say we think there's a fighting chance that there is room for multiple players, maybe for fast followers – that this could be actually a more diverse ecosystem than a lot of people are thinking. And, you can see some fast followers, like DeepSeek, that Chinese model that sort of rocked the world in early 2025.

It was built for $6 million. It was built for orders of magnitude less than the cost of many of these others and was almost as good. Fascinatingly, the performance difference between the truly top models has been converging, not diverging recently. They're more similar than they were a year ago. They're even more similar than they were two years ago, which doesn't seem like one party is just sailing away.

To the extent that everyone is using the same computer chips and building off of the same data, maybe it's not a complete shock that that's the case. You look at some of the cheaper open models, like DeepSeek, and they don't lag all that badly behind. They are orders of magnitude cheaper to produce, and you know laggards avoid missteps, laggards avoid having to pay for the most expensive computer chips.

Many experts have said that it's not clear that large language models will be best for everything. World models may be more useful. You will have different parties, perhaps, pursuing some of these things. If the bottom line is not that AI is of uncertain future, we think it's a very big deal indeed.

But it's not necessarily going to be monopolistic. It may be a diversified set of competitors, a better chance, at least, than you would think, just looking at other parts of the tech sector.

Okay, last high subject for me: unemployment concerns are very real. We've seen large layoffs from some big American tech companies recently as an example.

And so there's very real concern, not unjustified, because of course you could imagine AI replacing significant numbers of jobs. I do want to start just with an observation that at the top-down level, as we look at the U.S. job market, it still looks fine. Jobless claims are very low. The Challenger job loss index is very normal looking and not particularly rising right now.

The WARN index, which is formal notice companies must give to the government of large layoffs, looks pretty low and pretty normal as well. So I would say there's no evidence that something has turned in the last six months or anything like that. Though there are real concerns because, again, we can imagine AI replacing some jobs.

One thing we did was we worked our way through what sectors seemed more vulnerable. Of course, the general practice here is you say, well, who has the biggest potential for AI to replace workers. And so, of course, you can think of a variety of sectors that might be particularly vulnerable to that.

We did something else though, in addition to that. That was also looking at which sectors have the highest price elasticity of demand. The idea there is well, if AI is going to make products cheaper, for companies that embrace AI or are in a position to do so because their sector is in a position to do so, possibly demand rises for that product and you actually don't lose as many jobs as you think.

Or you even gain jobs because your business is growing because it's suddenly become cheaper. So we wanted to factor that in too, and to give you the punchline. I encourage you to take a peek at the written MacroMemo if you want the full story, because there's a big list of 30 sectors and high, medium, low vulnerabilities and so on.

We ended up with high vulnerability to job loss being advertising, marketing, consulting, wealth management – that's where I work. Software, IT, semiconductors, law, transportation, telecom, education, accounting, retail banking, medical care, pharmaceuticals as well.

And so high risk, but a number that were low risk, a number of medium risk, not all as you would expect, because in some cases, with a small price drop, if that product becomes cheaper due to AI it could actually result in more jobs because the sector itself is growing sufficiently.

So do keep that in mind as well. But on the net, we're certainly watching very closely. We did work in early March, in an early March #MacroMemo that spoke as well to the fundamental risks there. And we laid out different scenarios and we had relatively low probability that there was large scale, enduring job losses.

We had a number of scenarios that did affect the labor market, importantly, but fell short of that. There’s a relatively low chance, we think, that there will be structurally high sort of massive unemployment.

I get to finish now just with the briefest of comment here. Canada has announced a sovereign wealth fund, $25 billion being put towards it over the next three years.

We'll just note it's different than most sovereign wealth funds. Most use a government surplus and invest abroad. None of that is true here. You've got a government running a deficit, planning to invest purely domestically. So sort of violating both of the standard practices. And really some unusual bits to it.

For instance, promising to operate at arm's length, but equally, expected to support kind of government initiatives and infrastructure and resource plans and so on, expected to generate market returns despite doing those things for the government, which is going to be hard to pull off.

In particular, why wouldn't private firms have already been doing that? In theory, allowing retail investors in with a minimum risk of loss, which sounds hard to pull off as well. So it's a bit of a duckbill platypus in terms of the vision here. It’s not clear how it's going to work exactly. We'll see, to be honest, how it precisely settles.

I wouldn't be surprised if 1 or 2 of the objectives was set aside, just given how hard it is to achieve all of them at once. But for all of that nitpicking, I would say we do think this can be viewed broadly positively, just in that it's a further confirmation that the Canadian government does want to grow the economy, does want to get money invested into the economy, get infrastructure going, get resource investment going.

And certainly there are some positives to see in that.

Okay, I'll stop there and say thanks very much, as always for your time. Hope you found this useful and interesting. Please tune in again next time.