NVIDIA's decline reveals a surprising weakness in passive investing: a diversified index can still expose investors to concentrated risks.
KEY TAKEAWAYS
NVIDIA declined 4.27% despite the broader technology rally. The company's board announced an independent investigation following reports that CEO Jensen Huang had been targeted by a foreign intelligence effort, while stating that it had no reason to believe company secrets or national security were compromised.
Systematic and idiosyncratic risks affect portfolios differently. Systematic risk reflects market-wide developments, while idiosyncratic risk arises from individual companies and can generally be reduced through diversification.
The S&P 500 remains heavily concentrated in its largest constituents. According to the ETF holdings cited in the note, Apple, NVIDIA, and Microsoft represented approximately 21% of the fund, while the ten largest holdings accounted for roughly 37.6%.
NVIDIA's size gives its individual stock movements considerable influence over index performance. At an approximately 7% index weighting, Monday's 4.27% decline represented roughly a 0.30-percentage-point drag on the S&P 500, ahead of NVIDIA's November 17 earnings announcement.
MY HOT TAKES
Diversification is not something investors should simply assume because they own an index fund. When nearly 38% of an index sits in ten companies, investors are making a considerably more concentrated bet than the name S&P 500 might suggest.
The founder premium is a two-way street, and Monday provided a painful reminder. Investors celebrate visionary leadership when everything is going right, but the same dependence can create extraordinary downside when uncertainty surrounds that individual.
The greatest risk to passive investors may be the concentration they don't realize they own. Holding several different ETFs does not necessarily provide meaningful diversification when those funds repeatedly own the same handful of mega-cap companies.
NVIDIA's selloff was a company-specific event, but its consequences extended well beyond NVIDIA shareholders. When a single stock represents approximately 7% of a benchmark, idiosyncratic risk becomes increasingly difficult for index investors to ignore.
Concentration is neither inherently good nor inherently bad, but misunderstanding it is dangerous. Market-cap weighting has rewarded investors handsomely, and replacing it with equal weighting involves trade-offs, but every investor should understand precisely which risks are embedded in their portfolio.
You can quote me: "Stocks price uncertainty the minute it shows up, and on Monday they put a price tag on uncertainty around ONE person."
Take the A train. You know my morning routine–first espresso, local NYC news on the TV, and the subway traffic report. Whether or not I'm riding that day, I listen anyway. It's a ritual (um…er…I like routine, and I've never apologized for it). This morning, somewhere between the weather and the traffic, came a line I have heard a THOUSAND times: "Delays on the 4 and 5 trains due to a sick passenger at 59th Street." 🤮
For those of you who don't live here, the 4 and the 5 are the Lexington Avenue Express–the green trains. The 4 starts up in Woodlawn in the Bronx, the 5 up in Eastchester, and both barrel down under Lexington Avenue along the East Side of Manhattan–125th Street, the Upper East Side, 59th Street, Grand Central (which you have to see if you haven’t)–straight down to Wall Street 👋 and then under the river into Brooklyn. The "Lex" has long been called the busiest subway line in the country, and anyone who has tried to squeeze onto a 4 train at Grand Central at 8:30 in the morning doesn't need the statistic. By the way, we call it the train, not the subway. The Lex is my train.
Now, here is the thing about that announcement. One sick passenger is ONE problem, at ONE station, on ONE line. The A train doesn't care. The 7 doesn't care. That is very different from a blizzard that shuts down the entire system–that's everybody's problem. But when one line carries that much of the city on its back, one sick passenger at one station becomes a LOT of people's problem. Hold that thought.
Monday, the stock market had its own sick passenger at 59th Street.
It was a GREAT day for stocks. The S&P 500 jumped 2.35%, the Nasdaq Composite climbed 2.74%, and the Dow Jones rallied 1,104 points–a snapback from last week's ugly tech selloff. Chip stocks led the charge. AMD rose more than 13%. Broadcom gained nearly 9%. Micron added more than 6%. Everybody was on the express.
Everybody except NVIDIA–NVIDIA!--the company at the very center of the AI trade. It fell 4.27%, and it did so on roughly 463 million shares, about four times its normal volume. That is not a stock drifting lower. That is a stock being sold–hard. The reason had nothing to do with chips, demand, interest rates or the Fed. On Sunday, NVIDIA's board announced an independent investigation "in light of recent reporting that founder and CEO Jensen Huang has been the target of a foreign intelligence effort." A committee of independent directors, assisted by outside counsel, will oversee it. The board said it has "no reason to believe that either national security or NVIDIA secrets were compromised," and that Huang fully supports the review. NVIDIA has also disputed the underlying allegations involving its customer Megaspeed. Let me be clear–these are reports, not findings. Nobody has proven anything. But markets don't wait for verdicts. They price uncertainty the minute it shows up, and on Monday they put a price tag on uncertainty around ONE person.
There's an old rule on the Street that founder-led companies get a premium on the way up. Investors pay extra for the vision, the drive, the face on the stage in the leather jacket. What people forget is that the premium runs in both directions. When the founder IS the story, any cloud over the founder becomes a cloud over the stock–and Monday, that cloud cost shareholders dearly.
That brings me to a lesson every finance student learns in the first few weeks of class–and one I think every investor should revisit right now. There are two kinds of risk in the stock market. The first is systematic risk (sometimes called market risk). That's the blizzard. Rising interest rates, a recession, an oil shock, a war–these hit nearly every stock at once. You can't diversify your way out of systematic risk. If the whole system shuts down, it doesn't matter which train you're on.
The second is idiosyncratic risk (fancy word, simple idea–risk that belongs to one company). That's the sick passenger. A failed product, a lawsuit, an accounting problem, a recall–or a cloud over a founder and CEO. Idiosyncratic risk is exactly what diversification is designed to kill. Own enough different companies and one company's bad day gets drowned out by everybody else's normal day. That's the whole promise of the index fund–you're not betting on one horse, you're buying the racetrack.
Monday was a textbook case of idiosyncratic risk. The AI trade was alive and well. Just look at AMD and Broadcom. The sector was bid. ONE company was sold, because of ONE story about ONE man. So here's my question. If idiosyncratic risk is the risk you diversify away, why should someone who owns an S&P 500 index fund, and not a single share of NVIDIA directly, care? Because of how the index is built.
The S&P 500 is market-cap weighted (the bigger a company's total stock market value, the bigger its slice of the index). Most of the time that sounds perfectly reasonable. But as of last Friday's holdings in the largest S&P 500 ETF, Apple was 7.85% of the fund, NVIDIA was 7.10%, and Microsoft was 6.19%. Add up the top ten holdings and you get roughly 37.6%. Think about that–roughly 37 cents of every dollar you put into "the market" goes into ten companies. The technology sector alone is about 36% of the fund. That is not five hundred equal passengers spread across the whole system. That's a third of the city riding the Lex.
Here is the simple math. When a stock that makes up roughly 7% of the index drops 4.27%, that one move by itself subtracts about three-tenths of a percentage point from the index. On Monday nobody noticed, because the other 499 stocks were busy rallying. On a flat day, you would have noticed. On a down day, it piles on. By my rough math, NVIDIA shed something like $190 billion of market value on a day when the rest of the market was adding a fortune.
And this didn't start Monday. NVIDIA closed at a record $239.24 on October 6. Monday's close was about 25% below that. Back in early October, NVIDIA was north of 8% of the index—today Apple has taken back the top spot. The biggest passenger on the train changes seats faster than you'd think, and every time it does, index investors ride along, whether they meant to or not.
So yes–you can diversify away idiosyncratic risk. In theory. But when the index is this top-heavy, a single company's problem stops being a rounding error. It walks right back into the "diversified" portfolio through the front door. Idiosyncratic risk doesn't disappear in a concentrated index. It just gets a bigger seat at the front of the cabin!
Now, in fairness concentration has been a very good friend to index investors. The same weighting that dinged the S&P on Monday has been the engine behind a lot of the gains over the past few years. When the biggest companies win, a cap-weighted index wins BIG. Equal-weight approaches spread the risk out more evenly, but they come with trade-offs of their own (I wrote about that a few weeks ago–same destination, different route). There's no free-lunch version of this. You pick which risk you want to carry.
What I want is for you to KNOW which risk you are carrying. A few questions worth asking yourself this week. How much of my "diversified" portfolio is really the same ten stocks? Do I own NVIDIA three times over–once in my S&P 500 fund, again in a technology or growth fund, and maybe again in a handful of individual shares–without realizing it? If one of the top three companies in the index had a really bad quarter, what would that do to my retirement account? And am I comfortable with the answer?
Mark your calendar, too. NVIDIA reports earnings on November 17. That's a company event, but with a stock this big, it's also an index event. Earnings day for NVIDIA is earnings day for your 401(k).
None of this means panic. The board says it has no reason to believe anything was compromised, the investigation is just getting started, and NVIDIA's business didn't change between Friday and Monday. Key-person scares sometimes fade as fast as they arrive. But this episode is a perfect, real-time reminder that "diversified" is a word you have to check–not a word you get to assume.
Which brings me back to my espresso and the morning news. When I hear "sick passenger at 59th Street," I don't panic. I check which line I'm taking. If it's the 4 or the 5, I make other plans, or at least I leave early. If it's the A train, I finish my coffee.
Your index fund rides the Lex every single day. It's crowded, it's fast, and most mornings it gets you there quicker than anything else. Just don't be shocked when one sick passenger makes you late for work. 😉
YESTERDAY’S MARKETS
Stocks rallied Monday as chipmakers led a rebound from last week’s technology selloff, with the S&P 500 rising 2.35% and the Nasdaq Composite gaining 2.74%. The Dow Jones Industrial Average climbed 1,104.02 points to 52,389.75. Industrials led the sector gains, up 4.59%, while NVIDIA fell 4.27% after its board opened an independent investigation involving CEO Jensen Huang. The 10-year Treasury yield edged up to 4.88%, and gold rose 1.74% to $4,138.25. Oil slipped.
NEXT UP
NFIB Small Business Optimism (October) slipped to 97.8 from 97.9 missing estimates.
CPI / Consumer Price Index MoM (October) is expected to come in at 0.3%, slower than September’s 0.6%. Core is expected to come in at 0.2% after printing 0.3% in the prior period.
Fed speakers today: Waller, Logan, Schmid, Bowman, and Paulson.
Important earnings today: Venture Global, TransDigm, D-Wave, Firefly Aerospace, Rigetti, Coreweave, and Inovio Pharma.