James Learmonth argues that growth should resume in hardware, though not on the same trajectory
Last week, some of the best-performing names on global stock markets lost over $1 trillion (USD) in market capitalization. Semiconductor or ‘chip’ stocks including Nvidia, and leading memory names SK Hyniz, Samsung Electronics, and Micron, each lost hundreds of billions. That sell-off came in the wake of a legendary upward run for the sector. ETFs tracking the Philadelphia semiconductor index are still up almost 80 per cent year to date and well over 100 per cent in the past 12 months, despite the recent correction. That upswing was driven by the huge demand for semiconductors driven by the ongoing AI buildout. The downswing was driven by expectation.
James Learmonth, Co-CIO and Portfolio Manager at Harvest ETFs in Oakville, notes that many of these chip stocks saw their stocks drop even after they reported earnings that beat the published expectations. The problem, he says, is that as their stock prices grew ‘whisper numbers’ began to emerge. Those ‘whisper numbers,’ the unpublished but consensus higher expected earnings, were what the companies failed to beat. What followed, he says, was profit taking by investors who had already enjoyed a significant run up in semiconductor stocks.
“Ultimately that is healthy for markets. For the long term potential of any bull market, you want to have these resets. They never feel good and it’s never easy to just say ‘everything’s still fine, keep calm and carry on.’ But I think that’s what the case is here,” Learmonth says. “We saw saw Google and Meta raise their capex forecasts again. Microsoft and Amazon took a little more measured approach and got rewarded for that by the market, but the expectation that spending is going to continue to increase is still very much there. I don’t take that as, as an indication that we’re at peak spending growth yet. I think that spending trajectory is still very much in place and I think the AI infrastructure build out is likely still relatively early, mid to early cycle.”
Can chip stocks resume their growth trend?
With memory stocks among the biggest losers in last week’s pullback, Learmonth acknowledges that the memory manufacturing industry’s historical cyclicality may be in the minds of some investors. However, he argues that the long-term supply and demand imbalances for memory are so extreme, that memory manufacturers may have broken out of their cyclicality and any real shift will take far longer to emerge than it might have in the past. The three companies that manufacture the vast majority of the memory chips needed by AI still have huge pricing power and the scale of production required to meet or exceed growing demand is years away.
Broadly speaking, Learmonth sees that fundamental picture holding across the semiconductor space as the AI data center buildout continues. He acknowledges that markets may trade on sentiment to some degree, but that the underlying strength of earnings in the hardware and semiconductor space should push it back into growth mode. He acknowledges, though, that the likely growth of this sector will not be as steep as it was before. He believes that a slightly slower growth rate should be healthier for the sector overall.
Moreover, within the technology sector Learmonth doesn’t see any reasonable replacements for leadership. Software, with the notable exception of Palantir, has shown itself to still struggle with risk from AI disruptions rather than strength from AI adoption.
Coping with volatility and AI’s ‘inevitability’
The correction in chip stocks may come with additional lessons for advisors and investors beyond market narratives. The power of FOMO and the tendency to chase returns was writ large once again on some of the later investors in this sector. FOMO that has only been exacerbated by the wider popular discussion of AI as ‘inevitable,’ its buildout as ‘generational,’ and its promise for investors as ‘transformational.’
Amid all those noisy forces telling clients to go all in on today’s big AI winners, Learmonth says this recent pullback is another object lesson in diversification. Leading momentum names, he says, can correct sharply while previous laggards rebound. Diversification and balance can help with that. He argues, as well, for more equal weight allocations if investors are using index funds for this sector. The market cap run-up in some leading names has been so significant that cap-weighted index funds can be far more concentrated than they appear. Equal weighted funds, he says, offer a truer diversification.
From there, Learmonth says that advisors need to keep their clients focus on the long-term. Maintaining that vision isn’t easy, but it can go a long way in ensuring clients are well served over time.
“That’s probably the hardest thing for most people to do, to disassociate themselves from the short term gyrations in the market and stay focused on the longer term,” Learmonth says. “We’ve liked the technology sector for a long time, even before AI became the dominant theme. And we think there are still many themes within the technology sector. A lot of them link to AI, but AI becomes an enabling factor for things like autonomous vehicles, continued build out of communication networks, data storage and cloud based infrastructure, and cybersecurity And so from that fundamental perspective, I think maintaining that long term perspective is very important.”
