The Magnificent Seven Fell 40% While the S&P Dropped 18%: Here’s How Much Tech Concentration You Can Actually Afford

I’ve been studying portfolio concentration for years, and one stat from Schwab’s On Investing podcast keeps echoing in my head. In 2022, the Magnificent Seven fell roughly 40% while the S&P 500 dropped 18%. Our data confirms the pain: SPY shed 20% that calendar year, while Meta Platforms (NASDAQ:META | META Price Prediction) collapsed 64%, NVIDIA (NASDAQ:NVDA) fell 51%, Microsoft (NASDAQ:MSFT) lost 28%, and Apple (NASDAQ:AAPL) gave back 28%.

As the host said, “We talk so much about the return of these names, but we don’t talk about the risk.”

The Return Side Is Seductive For A Reason

NVIDIA just reported FY2026 revenue of $216 billion, up 65% year over year, with Jensen Huang calling this “the agentic AI inflection point.” Meta posted Q1 2026 revenue of $56.31B with operating margins of 41.4%. Microsoft’s AI business runs at a $37 billion annual rate, growing 123% year over year per CEO Satya Nadella. Apple just printed its best March quarter ever at $111.18B.

These are toll bridges of the AI economy. I’ve owned Meta since $121 in December 2022 and Apple since 2012, so I’m living the upside.

The Risk Side Is Where Portfolios Break

NVIDIA carries a beta of 2.2, a P/E around 45, and a price-to-sales ratio of 25. Meta is plowing $125 to $145 billion into capex this year, and Microsoft already spent $30.88 billion in a single quarter. If AI monetization slips, these multiples compress fast. The VIX sits at 18 today, but it spiked to 31 as recently as March 2026.

The Ballast Side: Where Diversification Earns Its Keep

Johnson & Johnson (NYSE:JNJ) rose 6% in 2022 while tech imploded. Its beta is just 0.3, and it has raised the dividend for 64 consecutive years. Berkshire Hathaway climbed 2.7% in 2022 with a beta of 0.6 and a trailing P/E near 14. Read the JNJ Q1 filing to see what a 27% operating margin looks like outside of software.

How Much Tech Can You Actually Carry?

The On Investing host’s framework is the cleanest I’ve heard: “The good news is you can own those in a diversified portfolio.” You can overweight the Mag 7 if you believe AI capex pays off, but only if “it fits your risk profile.”

You should consider tilting heavier into NVIDIA, Meta, Microsoft, and Apple if you believe agentic AI demand compounds for another decade and you have 15-plus years before needing the money. You should consider trimming and adding JNJ or Berkshire ballast if a repeat of that 2022 drawdown would force you to sell at the bottom or delay retirement. The host’s closing instruction stuck with me: “Maybe you own some of those concentrated positions, but make sure that there’s some diversification around those positions.” The Mag 7 doubled the index’s losses once. Plan for the next time before it happens, not during.