The Rhyme: Nvidia Enters Its $313 Billion Week & Its 2000 Echo

"History doesn't repeat… but it rhymes." — Mark Twain

Nvidia enters this week on a six-day losing streak, its longest since 2022, with options pricing a single-day swing of roughly $313 billion in market value when the company reports earnings Wednesday after the close. That number is not a typo. It is larger than the entire market capitalization of 471 companies in the S&P 500. The world's most valuable company at $5.2 trillion now faces a week where its earnings, a new Fed chair's first Jackson Hole speech, and the July PCE inflation report all land within 48 hours of each other.

NVDA $214.75 (6-day slide, −4.7%)  |  Options imply ±6% = ~$313B  |  S&P 500 7,674 (−1.4% on week)  |  30-yr yield 5.28% (highest since '07)  |  SMH −4.7% on week  |  Fed funds 3.50–3.75%

The last time a single chip company carried this much weight into a week like this, central bankers were gathering in the same lodge in Jackson Hole, the S&P was flirting with its all-time high, and the company's name was Intel.

On Thursday morning, August 24, 2000, a convoy of black sedans wound up Teton Village Road toward the Jackson Lake Lodge. Alan Greenspan, seventy-four years old and thirteen years into his chairmanship, stepped out into air that smelled like pine and wildfire smoke. The symposium topic that year was "Global Economic Integration: Opportunities and Challenges." Greenspan would give opening remarks the next morning about how technology was reshaping the world economy. Nobody in the room thought the reshaping was about to go the other way.

Three thousand miles east, in Santa Clara, Intel's headquarters hummed along on autopilot. The company had just posted its second quarter a few weeks earlier, and the numbers were spectacular. Revenue was running at an annualized rate above $33 billion, up 23 percent from the prior year, the fourteenth consecutive year of growth. Earnings per share had climbed 37 percent in the first quarter alone. Every server rack being bolted into a dot-com data center needed Intel silicon, and the orders kept pouring in. The stock sat near $73, within striking distance of what would become its all-time closing high of $74.88 on August 31. Intel's market capitalization had reached $503 billion, making it the most valuable semiconductor company on earth and one of the five largest corporations by any measure.

The analyst consensus was almost comically unanimous. Out of the major Wall Street firms covering Intel, not one had a sell rating. The research notes from that August read like love letters. Growth was accelerating. The internet was still in its infancy. The Pentium 4 was coming. Buy, buy, buy.

Greenspan spoke Friday morning, August 25. He talked about global trade, about how technology was driving productivity gains that traditional models couldn't explain, about how the new economy had "altered the structure of the way the American economy works." The audience nodded. The S&P 500 was trading at 1,506, within two percent of its all-time closing record of 1,527. The NASDAQ had crashed in April from its March peak of 5,048, but it had clawed back to around 4,100, and the mood was that the worst was over. The Fed had raised rates six times since June 1999, pushing the funds rate to 6.50 percent, and had been holding there since May. The economy looked like it was settling into a comfortable cruise.

Twenty-seven days later, on September 21, at 4:16 in the afternoon, Intel released a press statement that read like a short obituary. Third-quarter revenue would come in only three to five percent above the second quarter's $8.3 billion, well below what every analyst on the street had modeled. The reason was four words long: weaker demand in Europe. The stock dropped 30 percent the next morning. A hundred and twenty billion dollars in shareholder wealth vanished between the closing bell and the opening bell. Analysts who had been recommending purchase at $75 scrambled to explain why the stock was now worth buying at $40. It wasn't.

Both moments feature the world's dominant chip company trading near peak confidence during Jackson Hole week, with the S&P within two percent of its record, the Fed on hold after a hiking cycle, and not a single analyst willing to say sell.

Here is what happened after Greenspan left Jackson Hole in August 2000. For about a week, nothing. Intel drifted higher, touching $74.88 on August 31. The S&P briefly traded above 1,530 on the first trading day of September, flirting with its all-time closing high. Everything looked fine.

Then the music stopped. Intel's September 21 warning did not just take down Intel. It took down the idea that semiconductor demand was bulletproof. Cisco followed with its own warning in February 2001. By March 2001, the NASDAQ was below 2,000. By October 2002, Intel had fallen 83 percent from its August high to $12.95. The S&P had been cut in half. The stock did not recover to its August 2000 closing price for twenty-six years.

The people who made money were the ones who noticed something small before the warning hit. In August 2000, the purchasing managers' index was dipping. European industrial production had started to stall. The data was there, hiding in plain sight under the noise of analyst upgrades and Jackson Hole speeches about the bright technological future. Nobody wanted to look because looking meant questioning the best trade on the planet.

Today that same question hangs over Nvidia. Not whether Wednesday's quarter will be strong — it almost certainly will be. The question is whether anything in the guidance, the gross margins, or the customer commentary hints at the kind of demand plateau that nobody wants to talk about while the stock is still the most important holding in every index fund on earth. Intel's August 2000 quarter was fantastic. The warning came three weeks later.

The edge is not in the earnings number. It is in the guidance language, the gross margin trajectory, and any change in customer concentration. Intel taught the market that the last great quarter before a warning looks exactly like every other great quarter. The only difference is what the company says about the next one.

Consumer Confidence hits Tuesday morning, and July new home sales land at 10 a.m. If confidence breaks below 50, it will be the kind of demand signal that rhymes with the September 2000 PMI dip that preceded Intel's revenue warning by three weeks — the number everyone ignored because the bellwether's earnings hadn't missed yet.