There’s never been a better time to buy the Magnificent 7.
I don’t necessarily believe that, but you could make the case if you really wanted to.
With the caveat that Nvidia’s yet to report, and while acknowledging that Apple’s results left something to be desired, earnings season went ok for the mega-caps and importantly, full-year EPS estimates for the vaunted septet are more or less unchanged compared to a downward revision of 4% for the rest of the S&P 500.
The figure on the left, below, shows that Mag7 YoY EPS growth outstripped estimates by 16ppt, which Goldman’s David Kostin noted was the most pronounced beat in nearly four years.
The figure on the right shows that the valuation premium for the group’s now “just” 43%, down from a silly 100% at the height of 2021’s “stimmy”-fueled “everything bubble.”
Over the course of the furious rebound from the post-“Liberation Day” lows on Wall Street, the Mag7’s outperformed the index, rallying almost 30% versus the 19% bounce for the S&P. And yet, the group’s still lower on the year.
Between the big earnings beat, the relatively modest valuation premium (28x on a forward multiple versus 20x for the market) and the fact that the mega-caps are underperforming for 2025 by ~9ppt, you might argue they’re a bargain, absurd as that might sound.
One argument against the Mag7 says their earnings growth advantage versus the rest of the market’s set to shrink to nothing by end-2026.
The figure above, from the same Goldman note, is a cautionary tale vis-à-vis the shrinking advantage narrative.
“As evidenced by Q1 results, consensus estimates showing a diminishing earnings growth premium for the Magnificent 7 may be proven wrong,” Kostin remarked, dryly.
For what it’s worth, the “median” Mag7 stock trades cheap to his valuation model, which takes into account a slew of “fundamental attributes,” including balance sheet strength.
I’m not the first to point this out, but several of the Mag 7 are moving in the direction of asset-heavy, declining ROIC businesses. Referring of course to the massive AI capex vs minor AI revenue vs even more minor AI profits of the hyperscalers.
Your comment made me wonder if the massive AI investments are being amortized thus impacting EPS materially yet, or are the builds considered WIP (thus not yet hitting expense) and essentially acting as a repository of what used to be free cash flow, but is now sequestered on the BS as Fixed Assets. And would they need to write-down these fixed assets if the ROI doesn’t pan out as they hoped in their business cases. So far app level ROI has proven elusive.
The AI datacenter investment is capex, hits the income statement through higher depreciation and associated operating costs. Some of the hyperscalers have extended the useful lives of their servers, partly offsetting the depreciation increase.
Thanks JL. Given how fast the values of yesterday’s “must have” NVIDIA chips plunge, doesn’t it make sense to depreciate them to zero up front?
Or do CFOs want to wait to write them down until the promised AI earnings explosion distracts analyst attention from the write-downs?
Seems squishy to me too. The stated rationale is that “old” AI chips will still be useful. Seems more like an economic than technical issue – sure the H100 circa 2023 will work for a long time, but given its power/performance will GOOG etc have any profitable use for it?
Silicon and everything that goes with it has, in many ways, been the engine of global growth for 25 years or more. It has replaced the energy sector that has reigned since coal became useful. It seems structural that the Mag7 will continue to prosper if the global economy grows, since they are everywhere. They question for me is how long before new companies overtake any one of them. A good thought case is Apple. Will they come out with some market dominating new products and services or are their best years behind them? The rest of the market feels like a derivative.
Every company and every industry has a life cycle. They generally end in some form of market saturation, technological advance or major productivity advance. I thought Apple’s was about to end with the millennium but Jobs lucked into the i-phone so a new cycle was off and running. Now these phones are Cracker Jack toys to be given away. What’s next? Nothing I can think of. So far, at least, nothing they can think of. The company is not really set up for anything but incremental changes to existing products and they are not creating any new moats that can’t be instantly copied. As to the seven. Apple is running out of creative gas and has been wounded by Trump’s trade war. Meta, or whatever it is, is mainly a rolling high school year book. There are limits to the growth in the customer base. Virtual reality is not yet getting much attention from competitors because demand is not mushrooming. The “product” is mostly useless. Then there is Tesla. It’s future is not in taxis without drivers. That idea was destroyed for me, at least, when I saw the original movie “Total Recall” with Arnold. They had these funky cabs driven by silly cartoon robots. Couldn’t take Tesla seriously after that. There is no useful business model here, and the CEO is beginning to lose his “cleverness” gene, and public support. Sales and share are down. Still no major promises fulfilled. The size of the global auto market is hardly growing and the life cycle on EVs may be peaking. So, in my view, at least, three of the 7 are either past their peak or soon will be.
I like your assessment of Meta: ” … mainly a rolling high school year book. There are limits to growth in the customer base.” I’d add that it has been a lab for human behavior, and effectively monetized.
FB still has some room for growth via the Facebook groups product — so many private groups, clubs, etc. have given up on maintaining their own websites and went to FB, which provides mobile-friendly web page setups at much lower (time) cost. I wish this weren’t so, but having some experience in dealing with HTML via setting up web pages, I can’t blame “normal” (non-programming) people for wanting to take the easy path.