Key topics
- Quality stocks led by tech giants like Nvidia have outperformed.
- DeepSeek AI raises risks for dominant US tech companies.
- Tech valuations remain high, but disruption may cause a selloff.
Sign up for your early morning brew of the BizNews Insider to keep you up to speed with the content that matters. The newsletter will land in your inbox at 5:30am weekdays. Register here.
The seventh BizNews Conference, BNC#7, is to be held in Hermanus from March 11 to 13, 2025. The 2025 BizNews Conference is designed to provide an excellent opportunity for members of the BizNews community to interact directly with the keynote speakers, old (and new) friends from previous BNC events – and to interact with members of the BizNews team. Register for BNC#7 here.
If you prefer WhatsApp for updates, sign up for the BizNews channel here.
By Nir Kaissar
DeepSeek, a Chinese artificial intelligence startup, has developed a model that can apparently answer questions as well as any chatbot in the US. It might even help answer a long-running question on Wall Street without being asked.
For many years, so-called quality stocks in the US, which are essentially shares of the most highly profitable companies, have been among the best performing investments. The MSCI USA Quality Index has returned an enviable 15% a year since 2015, including dividends, nearly 2 percentage points a year better than the S&P 500 Index.
Technology giants such as Nvidia Corp., Microsoft Corp. and Alphabet Inc., recently catapulted by AI’s promise, have been a big part of quality’s success. And here lies the question: In a mostly efficient market like that of the US, where risk and return are supposed to be closely related, why should investors earn above market returns for owning the most stable, dominant and profitable companies?

Monday’s tech rout in which about $590 billion was erased from Nvidia’s market capitalization has brought this question back to the fore.
It’s true that higher profitability translates into a higher payoff for investors, all things equal, but investors usually pay a premium to own more profitable companies. I’ve included in my analysis the Russell 1000 Growth Index as a stand-in for quality in parts because there’s significant overlap between the two, particularly at the top. Both include tech heavyweights Nvidia, Microsoft, Alphabet, Apple Inc. and Meta Platforms Inc., though growth has a higher weighting toward them at 43% than quality’s 26%. Growth, however, offers a longer history and a useful contrast with value, whereas there isn’t a ready-made basket of stocks that represents the other side of quality.
The growth index is expected to achieve a return on equity of 34% this year, but at a price of 34 times forward earnings. The quality index offers only a slightly better tradeoff, with an expected ROE of 32% in exchange for 28 times earnings. Compare that with the Russell 1000 Value Index, which is expected to produce an ROE of just 12% but can be had for nearly half the earnings multiple.

As those numbers show, after accounting for both profitability and valuation, investors mostly get what they pay for, so it’s not entirely clear which basket of stocks is the best bet. Yet in the tech-fueled era, the quality index has beaten the S&P 500 by 2 percentage points a year since 1994 through December, the longest period for which its performance is available, and every time over rolling 10-year periods.
It may seem as if quality is a sure winner, but the emergence of DeepSeek, and its obvious potential disruption of US tech giants, suggests otherwise. It hints that quality investors are paid to take more risk, much the way stock investors are paid a premium relative to safe bonds or bond investors relative to stable cash. In both cases, investors are compensated for incremental risk.
The naked problem with abnormally high profits is that they’re not sustainable — most people wouldn’t recognize the names of the most profitable companies from 50 or 100 years ago. There are many pitfalls awaiting fat profits, and here are two: They tend to attract regulators, as Microsoft learned in the 1990s and Alphabet has experienced more recently. DeepSeek highlights a second, and potentially more lethal one, which is that high profits — or the potential for it from innovations such as AI — attract competitors eager to take market share.
The risk lies in not knowing when the party will end, which makes it difficult to get out in time. Once disruption is evident, stocks can reprice quickly to reflect the expected loss of profits well into the future. That’s particularly true for high-valuation companies such as Nvidia where much of the price is dependent on rosy future growth. The declines can be sudden and steep.
When the dust settled on Monday, the growth index was down about 3%. Quality was down, too, although not as much because it’s not as concentrated in the big AI names. Those are not drastic moves, but they show a market wrestling with the impact DeepSeek will have on the US’s biggest, most profitable companies that have loudly hung their future prosperity on AI.
If the losses among tech stocks continue, some perspective will help. Since 1995, growth’s forward price-earnings ratio has been about 46% higher than value’s on average. Even after the DeepSeek selloff on Monday, growth trades at a 72% premium to value, well higher than the historical average and comparable to the premium just before the tech selloff in 2022. When disruptions get serious, growth has even been known to trade at a discount to value, as it did during the 2008 financial crisis.
So, it’s too soon to say if DeepSeek will dethrone Big Tech. But it’s a good time to be reminded that the outsized returns investors have enjoyed from quality stocks are probably not a free lunch, even if the risk is hard to see and even harder to anticipate.
Read also:
- 🔒 ASML orders beat estimates as AI investments drive demand
- 🔒 The Economist: DeepSeek disrupts Nvidia, competition heats up in AI chip war
- 🔒 DeepSeek vs Nvidia: Can Blackwell reclaim the AI lead?
© 2025 Bloomberg L.P.