by Jurrien Timmer, Head of Global Macro, Fidelity Investments
Greetings from Shanghai, where the view (and people and food) never disappoints. I have now had several dozen meetings over the past 2 weeks in Korea, Japan, and China, with three more days coming up in Taiwan and Hong Kong before returning home in time for the July 4th fireworks.
As I take stock of the conversations, several major themes jump out for being consistently on the minds of the institutions visited. First and foremost, the same question that is on our minds stateside is equally prevalent here: is the AI boom sustainable or is it a bubble? How much is too much, and how can I simultaneously monetize this boom while also protecting myself if it’s a bubble? That topic was quickly followed by questions about the Fed under Kevin Warsh’s new leadership. Beyond that, gold was a prevalent, which is quite interesting. Finally, the conversation tended to evolve into a discussion of what a diversified portfolio looks like in a post-60/40 world.
Let’s review.
Branches on a tree
The first topic regarding the paradox of profiting from a boom while protecting from a bubble is of course an existential one and reflects the duality of the profit seeking polarity juxtaposed against loss eversion. Fortunately, in my view there is a solution to this challenge, illustrated by the decision “tree” below.
The chart measures the return of four major equity themes against the S&P 500 (vertical line). The “branches” that sprout to the left show the themes that have underperformed the S&P 500 since the melt-up started in April 2025, and the branches on the right are the winners. The size of the dots measures the theme’s forward P/E.
The binary options are clear: mega caps and AI vs equal-weighted and ex-AI. For me, a barbell of the two sides makes the most sense, with a focus on value and income as the hedge against the risk of the AI boom getting out of hand. Fortunately, within that value/income bucket is one group that has also sprouted to the right: Eurozone financials. Perhaps a barbell of AI and Eurobanks is the match made in heaven, especially since the latter trades at a P/E of 11x.
As for the AI trade, it came under pressure last week as questions continue to mount about the ROI on all this capex as well as the disruption risk to the LLMs. And as the AI names wobbled, the ex-AI names not only outperformed but rose in absolute terms as well.
And just like that, the S&P 500 equal-weighed index has caught up to the cap-weighted index, as breadth improved to 65%. It’s good to see that the broadening has not come too much at the expense of the index itself (which is down 3% from its high).
Of course, semiconductors are usually a cyclical business, so the modest P/E may not be that meaningful (peak earnings produce misleadingly low P/E’s). But perhaps semiconductors are now a secular industry as they fuel a generational shift in technology. But this question requires a leap of faith that it’s different this time, which in itself gives me pause.
As for the broader market, earnings are booming there too. Second quarter earnings season will be upon us in a few weeks, and we can see how strong the incoming tide is in terms if earnings estimates.
That rising tide is keeping overall valuations at bay, with the S&P 500 index trading at a 22x forward multiple.
Thinking about the markets in a holistic sense, it’s clear that today’s landscape has tails on both sides. The chart below shows the distribution for the trailing P/E ratio, earnings growth, and the S&P 500 annual return. We are at or near the tail on all counts.
What can go right? Clearly the earnings boom is lifting both valuations and returns, as it should. And if the AI boom creates a productivity miracle for all downstream sectors, then the market should benefit (but perhaps less so the cap-weighted index if the market broadens).
On the left tail are two models: the CAPE model and the Fed model. The CAPE model suggests that the next 5-10 years are going to produce less beta than the last 5-10 years (on the assumption that CAPE ratios mean-revert). And the Fed model is a risk in case the global bond bear market that started in 2021 continues and the US 10-year returns to the danger zone of 5%.
What could prompt yields to head into the danger zone? Inflation, for one. While TIPS break-evens are plummeting because of declining oil prices (and faith that Fed Chair Warsh will get inflation back down to its 2% target), I wonder if the bond market is declaring victory too soon.
The massive AI buildout may lead to deflation down the road, but it’s clearly inflationary for now. The Taylor rule below shows that all measures of this model are heading higher instead of lower, which suggests that the Fed eased more than it should have in 2025 and early 2026. Now the markets want the Fed to take a few of those cuts back. Meanwhile the TIPS forward curve is in a world of its own (the grey dots below).
That central bank hawkishness is also lifting the US dollar, as expectations mount that the Fed will have to undo the last few rate cuts. This looks like a clear breakout from a long base.
As for the questions about gold as a diversifier in this post 60/40 world, it hasn’t been much of one this year, which is why it is a point of discussion. Gold is now positively correlated to equities and bonds, which is a departure from the past.
I suspect that gold’s correlated behavior will be temporary as gold has fallen out of favor after it went vertical last year. The fast money used to be in Bitcoin, and then it left for gold, and now it’s chasing semiconductors. The chart below shows how gold’s real rate model fell apart in 2022 when gold became a proxy for global liquidity. With global M2 now slowing from a growth rate of 12% at the peak to 7%, gold is understandably weaker. But it’s too weak considering the modest deceleration in M2.
The boom was big, and it took gold well above its liquidity-based fair value. Now the pendulum has swung the other way, and gold is 13% cheap compared to where global liquidity is trending. With central banks turning hawkish and semiconductors stealing the show, there may not be much of a catalyst for gold to rally just yet.
That same logic applies to Bitcoin as well. It’s getting closer to its power law support line. Again, I don’t see the catalyst yet for a reversal.
1270482.1.0
This information is provided for educational purposes only and is not a recommendation or an offer or solicitation to buy or sell any security or for any investment advisory service. The views expressed are as of the date indicated, based on the information available at that time, and may change based on market or other conditions. Opinions discussed are those of the individual contributor, are subject to change, and do not necessarily represent the views of Fidelity. Fidelity does not assume any duty to update any of the information.
Copyright © Fidelity Investments
