Executive Note: As I noted yesterday, insider buying had picked up, and we hit a bottom on the number of breakout stocks. I view both of these as contrarian opportunities… but I’m still cautious. I go back to 2022, a period of lower highs and lower lows, but with intensive, fast squeezes that gave way to more selling later. I highly encourage anyone investing and trading to look at the Relative Strength Index and Money Flow Index on the hourly and daily charts to assess momentum. Our signal is still red, but we’ve seen a very strong move off those lows. If momentum returns, consider all those names I discussed this week that are tied to liquidity and capital efficiency. Good Morning: NVIDIA reported its numbers, and the financial world reacted as it always does. Eyes widened. “Wow…” I said… watching the first green candle take the stock up over $190 in no time at all… Analysts sprinted to rewrite price targets. Commentators cheered the future of everything, everywhere, all at once. Anyone who has paid attention to how markets function today should already know the real story. NVIDIA is no longer just a semiconductor company. It has become the prime example of a modern asset that benefits from three powerful financial currents.
That trifecta is why the stock behaves like an unstoppable object. It is not a cult. It is a utility… a liquidity instrument dressed as a narrative about the future. So let’s pull away the confetti and look at what actually happened with its earnings. The Good of Earnings…NVIDIA posted record results. Revenue up 62% year over year. Data center revenue up 66%. Gross margins are hovering around 75%. Those are numbers you don’t find in many established mega-cap companies. NVIDIA has a monopoly on the world’s most important infrastructure. The current version of the gold mine is AI compute capacity. NVIDIA sells the picks and shovels, and the miners have no alternative. There is real demand… not a marketing stunt. Blackwell chips really are sold out. Every government, every cloud company, every sovereign wealth fund, every startup, and every data center developer is chasing GPU capacity the way railroads once chased steel. There is nothing imaginary about their product. It has become the power grid of machine intelligence. The BadThe industry around NVIDIA is still searching for a way to pay back the cost of these chips. That’s showing up in the balance sheet… Revenue is exploding, but earnings from the companies buying the hardware are not. Users of NVIDIA’s infrastructure have yet to prove they can monetize the applications built on it. They are racing to build capacity, but the payoff could be years away, and the economy is slowing. NVIDIA is earning money the way utilities earn money during industrial booms. Everyone else is still guessing how to turn electricity into profit. That disconnect has not mattered yet. Eventually, it does. With $19 billion in inventory and an expectation of $65 billion in fourth-quarter revenue, NVIDIA must keep selling at breakneck speed to maintain equilibrium. Yesterday, I had to get into a conversation about Cost of Goods Sold and inventory turnover… something that I haven’t done in nine years… Turnover - based on the back of the envelope was only around 3.3x to 3.6x (someone correct me)… and if that’s the case, I find that… low. Wouldn’t a hyper-growth hardware business with sold-out demand turn inventory far faster? Isn’t… AI is “going everywhere, doing everything, all at once” Well… Inventory has doubled since the end of last fiscal year. Working capital doubled… and the footnotes around China quietly remind us that policy can quickly take a brick to the narrative if both the U.S. and its rival can’t find common ground. A small slip in demand from China, government restrictions, a geopolitical hiccup, or delayed data center buildouts can quickly distort pricing and growth assumptions. When an asset trades at a valuation built on perfect visibility, even a fog bank can unsettle the road. The WTFThe CEO statement is just #@$%#$ wild… Jensen Huang did not just describe demand. He described inevitability. During the announcement, he said that compute demand keeps accelerating and compounding across training and inference, each growing exponentially. He also declared that the world has entered a virtuous cycle of AI. Not a cycle of investment. A virtuous cycle of AI. According to him, the AI ecosystem is scaling fast, with more new foundation model makers, more startups, across more industries, and in more countries. Not slowly. Not cautiously. Scaling fast. Then he went further. AI, he said, is going everywhere, doing everything, all at once. That was not a fever dream written by a Wall Street quant. I’ve never seen a CEO talk like this… That was the CEO of a trillion-dollar company. Saying… AI is going everywhere, doing everything, all at once. If a railroad baron had said railroads were going everywhere, carrying everything, all at once, it would have marked the precise moment historians would later cite as the peak of that cycle. The guidance embodies that tone… Next quarter revenue is expected to jump again. Margins are expected to expand again. Spending on data centers is assumed to rise again, regardless of interest rates, geopolitical friction, sovereign restrictions, cloud cost pressures, or the growing pile of unfinished monetization sitting on balance sheets around the world. None of that exists in this story. Was there any risk assessment? Not really… It all happened under the glow of inevitability. Infrastructure booms always start with real necessity… They become bubbles when the necessity is treated as infinite. Railroads before overbuilding. Fiber optic networks before the glut of dark fiber. Power generation before deregulation snapped demand assumptions in half. The technology was real in every case. The valuations were not. That’s the challenge here… I’m not saying we’re there yet… but I will bet that everyone, everywhere soon becomes an expert on infrastructure bubbles and starts comparing NVIDIA and the data center boom to the 1840s railroads… NVIDIA is selling the most valuable industrial commodity of the decade. The world is racing to buy it. But the market may start behaving as if demand cannot slow, cannot pause, cannot meet friction, cannot require a breather, and cannot fall short of the exponential curves used in presentations. That belief is never sustainable. The product will survive. The expectations rarely do. Jensen might be right about the future. He might even be early. But markets are pricing his confidence as if there is no timeline, no risk, and no chance of disappointment along the way. But that means… that the music is still playing… and everyone will keep dancing… Let’s get to the market…... Continue reading this post for free in the Substack app
|
Subscribe to:
Post Comments (Atom)
0 Response to "NVIDIA Earnings: The Good, the Bad, and the WTH?"
Post a Comment