Call it the Nvidia NVDA-Q dilemma. Since the start of the year, shares of the chip maker have nearly tripled in value, propelled by growing excitement over how artificial intelligence will stoke demand for the company’s products.
A handful of other companies have also caught the AI wave in a big way. Thanks to that lift, Alphabet Inc. GOOGL-Q, Amazon.com Inc. AMZN-Q, Apple Inc. AAPL-Q, Meta Platforms Inc. META-Q, Microsoft Corp. MSFT-Q and Tesla Inc. TSLA-Q have joined Nvidia Corp. in a very small group of major market winners.
Their collective surge poses a difficult question for investors. How much – if anything – should we be willing to pay for these magnificent seven stocks? Are they worth 30 times projected earnings? Forty times? Or are they hopelessly overvalued at current prices?
The answer may depend on which historical parallel you choose to use.
It’s easy, for instance, to dismiss the current boom as a rerun of the 1990s dot-com silliness. If so, your best move would be to simply ignore today’s red-hot stocks.
That doesn’t seem quite right, though. Many of the most prominent dot-com flops were speculative fantasies that soared despite an absence of earnings.
The stocks getting a lift from AI today aren’t like that. They all have enormous revenues, a history of profits and entrenched market positions. You can question their share prices, but there is no doubt they are impressive businesses.
A better parallel might be with the Nifty Fifty stocks of the early 1970s. These represented companies with great track records and impressive prospects – think Pfizer Inc. or Xerox Corp. or Coca-Cola Co. Like today’s magnificent seven stocks, they traded for lofty valuations far above the broader market – until they hit hard times in the recessionary mid-1970s.
In a famous 1998 paper, Wharton finance professor Jeremy Siegel argued that the Nifty Fifty was no bubble.
He calculated that someone who bought an equally weighted basket of these 50 stocks at their most expensive price-to-earnings ratios in late 1972 would have reaped returns over the next 25 years that were nearly indistinguishable from the returns produced by the broad S&P 500 index.
Prof. Siegel took a couple of somewhat contradictory lessons away from his Nifty Fifty research.
One lesson was that investors shouldn’t rule out stocks simply because they happen to be a bit more expensive than average. The market is no fool and stocks with superior growth prospects will typically trade for higher multiples of earnings than more-mundane companies. Sometimes those higher multiples will be justified.
The second lesson was that investors should not be willing to pay any price for these supposedly superior growth stocks. Value still matters. If an investor had bought the cheaper half of the Nifty Fifty in 1972, they would have reaped returns that were roughly double what someone who bought the more expensive half of the list would have collected.
Those are fine conclusions. Let me add one observation of my own: Valuations can shift a lot over a short period.
Nvidia offers a case in point. Over the past five years, it has traded as low as 20 times forward earnings and as high as 73 times forward earnings. Investors would generally have fared better by waiting for the valuation to fall to a more modest level.
Investors who are eager to place a bet on the AI trend might want to keep this volatility in mind and wait for cheapness to emerge wherever possible.
The most reasonably priced members of the magnificent seven right now are Alphabet and Meta. Alphabet is valued at 21 times forward earnings, Meta at 22. This makes them somewhat more expensive than the broad market, but not by much.
Alphabet and Meta seem particularly reasonable when judged against their own valuation histories. Both are trading at price-to-earnings ratios below where they have typically traded since 2018.
Granted, their relative cheapness points to the challenges both companies now face. Meta is grappling with its desire to pivot away from its Facebook roots and toward an uncertain future in the metaverse. Alphabet may be losing ground in its AI competition with Microsoft.
On a more positive note, both Meta and Alphabet have the time and resources to get their houses in order. This doesn’t mean they will turn things around in short order, but it does offer hope for patient investors.
The opposite may be the case for some of the more expensive members of the magnificent seven. Aswath Damodaran, a professor of finance at New York University and an expert on stock valuation, recently published his analysis of Nvidia stock. He estimates its value to be around US$240 a share, far below the US$420 it is now trading around.
You can dispute Prof. Damodaran’s assumptions, but his deeper point is that valuations do matter, even when dealing with high-growth stocks.
To my eye, Meta and Alphabet seem like reasonable bets for investors who desperately want a piece of the AI magic.
Beyond that? The best idea might be to see if the Nifty Fifty timeline repeats itself and today’s high-flying stocks suddenly become a whole lot less expensive over the next year or so.