NVIDIA is a $5 trillion company and the stock trades at 18 times earnings. Revenue is supposed to grow 97% this quarter after six straight quarters of reacceleration. Even Jensen is out calling it the seventh wonder of the world while asking why the stock is not moving the way it should.

Shay Boloor and Brad Freeman spend the whole episode on that dislocation. The easy story is over. For two years the pitch was that GPUs were sold out and NVIDIA could not make enough of them. Now the market is pricing this as the peak capex year while the physical supply chain says the boom runs at least three more years, and the real question is whether the customers sitting above NVIDIA can monetize AI fast enough to justify what is being built underneath them.

They get into what NVIDIA has actually become, a networking leader with a $20 billion CPU business created overnight, the $500 billion financing pool with Apollo, BlackRock, and Blackstone, whether custom silicon from Google, Amazon, and Meta is really zero sum, the CPU renaissance that agents are driving, and what to watch when earnings land this week.

The verdicts split. Shay put 80% of his son's brokerage into NVIDIA and treats it like a T bill, the foundation of a new AI economy at the start of the S curve. Brad owns the cycle through an ETF instead, giving up upside so a bad ending costs him 30% and not 80%. Both think it is hard to imagine much more multiple contraction while the profit growth keeps coming.

Watch the full episode with chapters on YouTube: https://youtu.be/qa-gJCJbM_s

New episodes every Monday, so you know what to watch before the week trades.

Nothing in this episode is investment advice. Shay and Brad discuss their own portfolios; Shay's family accounts own NVIDIA, and Brad owns AI infrastructure through an ETF rather than the stock.

For the last few years, AppLovin was a beat and raise machine. Beat the top line by 4%, guide 3% ahead, see you next quarter. It was boring in the best way. This quarter the streak broke. Revenue still grew 53%, but the company came in 1% light against its own guidance, and the stock now trades at 15 times earnings, the same multiple as The Trade Desk.

That comparison is why Shay Boloor and Brad Freeman spend the whole episode here. Trade Desk's slide started with very reasonable sounding excuses too, and they both watched it happen up close. So they want to know whether AppLovin is just a victim of the trust Trade Desk burned, or whether this really is the first crack.

They dig into the late model update behind the miss, whether Unity's gaming data is real competition, what it costs to keep making Axon better every three months, and the push into e-commerce against Meta, Google, and Amazon.

Shay comes out bullish. He thinks the new risks are already priced in at 15 times earnings, and further weakness would make him more aggressive, not less. Brad stays neutral. If he wants advertising exposure, he'd rather own the multi-trillion dollar companies that are growing like weeds. Both agree Q3 decides it.

Watch the full episode with chapters on YouTube: https://youtu.be/Rv8LDYcBbRI

New episodes every Monday, so you know what to watch before the week trades.

Nothing in this episode is investment advice. Shay and Brad discuss their own portfolios; neither owns AppLovin, and they hold positions in some other names mentioned, including Meta and MercadoLibre.

CrowdStrike (CRWD) trades at 110 times free cash flow, and the two investors who know it best split on what to do about it. Shay Boloor and Brad Freeman of Stock Market Nerd open their new weekly show with a full CrowdStrike deep dive, then preview the seven names reporting in a loaded AI earnings week.

The CrowdStrike conversation starts with the July 2024 outage: the company shut the world off for a few hours, handed its entire competitive landscape a free shot, and lost no meaningful market share. Two years later the SIEM, cloud, and identity businesses run at a $2 billion rate, Falcon Flex is changing how customers buy security by letting them commit to a pool of spend instead of signing product-by-product contracts, and the agentic SOC is the next act. The problem is the price. Brad sold every share on valuation at 35 times sales and has no complaint about the business. Shay trimmed 15% in July to fund AI power names and still cannot bring himself to exit a stock he says may neither reach $300 nor break below $100 for a long time. Both of them dismantle the idea that frontier AI models replace security vendors: a model that hallucinates 15 to 20% of the time does not get the keys to your digital estate.

Then the week ahead. Nebius (NBIS) reports Wednesday pre-market with a capacity story: estimates already came down for Q2 while 2027 held, because data centers need power connections, transformers, and cooling before a GPU earns a dollar. Hims (HIMS) is mid-transition with gross margin down 8 points and GAAP net income margin swinging from 8% to negative 15%. Sea Limited (SE) is taking Brazil while the watch item is credit risk in Indonesia. CAVA at 35 times EBITDA works out to roughly 102 times earnings once store depreciation is counted. Rocket Lab (RKLB) rides entirely on Neutron staying on track. AST SpaceMobile (ASTS) is the most speculative name in the group and does not own its own launches. Cerebras (CBRS) is expected to grow from $865 million this year toward $8 billion by 2029 and already costs 45 times sales. The show closes with Shay's tin hat theory on a data center pullback into the midterms and Brad's case that mega-cap token pricing and open source models are squeezing OpenAI and Anthropic from both ends.

Watch the full episode with chapters on YouTube: https://youtu.be/jpvWWyEOCNk

New episodes every Monday, so you know what to watch before the week trades.

Nothing in this episode is investment advice. Shay and Brad discuss their own portfolios and hold positions in some of the names covered, including CrowdStrike and Nebius.