|
Issue No. 16 • Thursday June 11, 2026
THE TRADING ADDICTNEWSLETTERby Maria Helmick » Daily Trading Update · Day 163 · Tap for live dashboard
Here is how our AI trading robot Phil performed today. » Trades of the Week are at the end of the Newsletter « » Story No. 1 of 2 · Oracle Earnings
Tap the image to view full size Same Story, Different StockOracle beats expectations, but Wall Street focuses on the cost of the AI buildout. Oracle put up a strong quarter, but Wall Street still found something it didn’t like. The company reported record quarterly revenue of $19.18 billion, up 21% from last year and slightly ahead of estimates. Adjusted earnings came in at $2.11 per share, well above the $1.96 Wall Street expected.
Those are impressive numbers. The Problem: AI Growth Is ExpensiveOnce again, investors looked past the earnings beat and focused on how much money Oracle is spending to build out its AI business. Oracle spent $55.66 billion on capital expenditures during fiscal 2026. That heavy spending pushed free cash flow to a negative $23.69 billion for the year. The company now expects to raise roughly $40 billion in fiscal 2027 to help fund more data centers and AI infrastructure. That includes the previously announced $20 billion stock offering.
Wall Street Wants ProofThe concern is simple: Oracle’s cloud business is growing fast, but investors are not yet convinced it will be as profitable as the company’s traditional software business. Software revenue actually fell 2% during the quarter to $6.8 billion. Management says many of Oracle’s largest cloud contracts are prepaid or use equipment provided by customers, which should help protect margins. The company also believes its cloud business can eventually produce returns in the high-20% range. Oracle is still expecting strong growth. Revenue is projected to rise 27% to 29% in the first quarter.
Educational only. Investing involves risk.
» Story No. 2 of 2 · AI Memory · MU Explainer
Tap the image to view full size Nvidia Gets the Headlines. Memory May Get the Profits.
Everyone talks about Nvidia when they talk about AI. That makes sense. Nvidia’s GPUs are powering the AI data centers behind OpenAI, Anthropic, Google, Meta, xAI, and nearly every major AI platform. But the AI boom is not only about GPUs. Those chips need massive amounts of memory to work. That is where Micron (MU) comes in. Micron makes DRAM, which is the short-term memory used inside computers, servers, and AI systems. Think of DRAM as the workspace. The more complex the AI job, the more workspace the system needs. That is why memory has become one of the most important parts of the AI infrastructure trade. The Fox Business segment showed DRAM demand projected to grow from roughly 250 exabytes in 2024 to more than 1,200 exabytes by 2031. That is nearly a fivefold increase in seven years. The key driver is not phones or PCs. It is data centers. That matters because AI data centers use far more memory than traditional servers. They need advanced DRAM and high-bandwidth memory to keep the GPUs fed with data. If the memory is not there, the expensive AI chips cannot perform at full speed. For years, memory was treated like a boom-and-bust commodity business. Prices moved in cycles, supply gluts hurt margins, and investors often avoided the group. AI may be changing that setup. If demand keeps rising faster than supply, DRAM prices can stay firm or move higher. For Micron, that is important because stronger memory pricing can flow directly into better margins and stronger earnings.
The long-term DRAM pricing chart also matters. Historically, DRAM prices trended lower for decades. But recently, pricing has started to turn higher as AI demand absorbs more supply. That is why Wall Street is starting to look at Micron differently. This may no longer be just an old-school memory stock. Micron is becoming a direct way to play the AI buildout without buying Nvidia after a massive run. The big question is whether AI creates a longer and stronger memory cycle than the market is used to seeing. If it does, MU could stay important well beyond one earnings report. There are still risks. Memory stocks can move fast. DRAM pricing can turn if supply catches up. Expectations can get too high. Micron can also sell off after earnings even when the long-term story remains intact. But the big-picture story is clear: AI does not just need chips. AI needs memory. And Micron is one of the companies supplying it.
Maria’s TakeMicron is not moving because investors suddenly love old-school memory stocks. It is moving because AI is changing the memory cycle. If data-center demand keeps growing and DRAM pricing stays firm, MU has a real earnings tailwind. But this is still a cyclical stock, so chasing after a big run can be dangerous. Micron reports earnings on June 24, so traders looking for an earnings play may want to focus on the option chain instead of chasing shares. Premiums are elevated, and some of the deep out-of-the-money puts are paying well while still sitting far below the current stock price. Maria’s take: MU stays on the watchlist, but the trade needs discipline. The AI story is real — the risk is paying too much after everyone else finally notices. Educational only. Not financial advice.
Trade small, trade often. Math Makes Money. Get a fill, Phil. — Maria Math Makes Money · AI Trading Holdings LLC |