This page tracks every stock BWB - Business With Brian has discussed on YouTube, extracted from AI transcription of their 4 most recent videos covering 19 tickers. They are currently bullish on GLD, SAND, MU, CEG, GOOGL, UBER and others. Bearish on GOOG, NVT. Each take below includes their stated reasoning and, where given, a price target.
Brian frames the current environment as a standoff between the Treasury (borrower, foot on gas) and the Fed (rate-setter, foot on brake), with long-term yields at 5.3%—the highest since 2007. Washington's traditional tools to lower rates are all shut: long-term borrowing is locked through 2026, the Fed is threatening hikes instead of cuts, QE has stopped, and Congress won't cut spending. The only historical pattern that breaks long-term yields is an economic crack (bad jobs report) while the Fed is still hiking, as seen in 2000, 2006, and 2018. Globally, rates are rising everywhere—UK 30-year at highest since 1998, Germany since 2011, Japan 10-year above 3% for the first time since 1996. Japan, America's largest foreign lender at $1.1T, has already shed ~$120B of Treasuries since February and may keep pulling back as domestic Japanese yields become more attractive. Brian views bond funds paying ~4.7% as ahead of inflation and preferable to cash at 3.5%, and sees new 30-year Treasuries paying 5%+ as historically likely to beat inflation. Overall sentiment is cautious/risk-off with elevated uncertainty.
| Ticker | Action | Why | Target | When |
|---|---|---|---|---|
SAND●●● | buy | SanDisk's gross margin surged 41.4 points while revenue nearly tripled, and it trades below the creator's bear case with a margin of safety comparable to Micron's. As a recently spun-out company with only 18 months of public history, it lacks the valuation percentile penalty that constrains other names, and the scarcity fingerprint matches the broader memory category. | — | 6d ago |
MU●●● | buy | Micron's gross margin jumped from 37.7% to 84.6% year-over-year, out-earning Microsoft on the margin line, with 16 long-term supply agreements running to 2030 and the CEO stating they cannot see when supply catches demand. At 6x forward earnings the market is pricing a cycle top, but the creator's system still allows starting a position, with the key risk being a Chinese state-backed memory competitor ramping capacity. | — | 6d ago |
GOOGL●●● | buy | The creator leans positive long-term, noting Alphabet looks cheap against earnings (higher margins than historically) but expensive against sales. He views it as a dollar-cost-averaging candidate rather than a high-conviction buy, starting with a 50% starter position. | — | 2w ago |
MU●●● | add | The creator's model puts fair value near $1,450, placing it in the deepest buy band, but the same system flagged it as richly valued against its own trading history. He takes a smaller bite due to this conflict and notes memory is cyclical, with the market assuming profits roughly triple from here. He continues buying on his normal schedule but won't add aggressively until the stock climbs back above its 50-day moving average at $961. | $1,450 | 2w ago |
UBER●●● | buy | The creator's independent fair value of $109 closely matches Parkev's $120, and Uber sits in the cheapest third of its own 10-year history while generating roughly $10 billion in annual free cash flow against a ~$150 billion market cap. He leans on cash flow rather than earnings, and the key risk (driverless cars disrupting the model) is clearly identified. This is his deepest buy band and the only name where his system didn't conflict with itself. | $109 | 2w ago |
ORCL●●● | add | Oracle has $638B in signed contracted revenue, largely from a $300B OpenAI deal, yet trades 57% below its high. Even stripping out OpenAI entirely, the stock costs less than before the deal was announced, making it cheap on already-signed sales. The main risk is negative free cash flow from heavy data center build-out and dependence on a single unprofitable customer. | — | 3w ago |
INOD●●● | buy | Innodata is a neutral AI data-labeling supplier whose independence became its biggest selling point after Meta bought half of rival Scale AI, driving competitors' customers away. Revenue grew 58% last quarter and customer concentration is improving, but the stock is volatile, trades at 46x earnings, and has 14% short interest. | — | 3w ago |
APP●●● | add | AppLovin runs an AI-powered ad engine with ~80% operating margins, higher than Visa, and trades at ~21x next year's earnings after falling over 50%. The SEC closed its investigation with no wrongdoing found, removing the biggest overhang, yet the stock dropped 20% on a slight guidance miss. Brian calls it the highest quality name on the list and is buying more once technicals stabilize. | — | 3w ago |
In the scenario where Japan raises rates and pulls capital home, U.S. rates rise while the dollar falls, and Brian says 'the people who bought gold, well, they collect.' Gold is positioned as a hedge against a Japan-exit scenario.
SanDisk's gross margin surged 41.4 points while revenue nearly tripled, and it trades below the creator's bear case with a margin of safety comparable to Micron's. As a recently spun-out company with only 18 months of public history, it lacks the valuation percentile penalty that constrains other names, and the scarcity fingerprint matches the broader memory category.
Micron's gross margin jumped from 37.7% to 84.6% year-over-year, out-earning Microsoft on the margin line, with 16 long-term supply agreements running to 2030 and the CEO stating they cannot see when supply catches demand. At 6x forward earnings the market is pricing a cycle top, but the creator's system still allows starting a position, with the key risk being a Chinese state-backed memory competitor ramping capacity.
Constellation Energy passed the margin test with a 24.5-point increase, making it the honest exception to the idea that electricity scarcity is impossible to own. It owns already-built, already-licensed, already-running nuclear reactors that no factory can replicate this decade, representing the scarce thing itself rather than commoditized equipment.
The creator leans positive long-term, noting Alphabet looks cheap against earnings (higher margins than historically) but expensive against sales. He views it as a dollar-cost-averaging candidate rather than a high-conviction buy, starting with a 50% starter position.
The creator's independent fair value of $109 closely matches Parkev's $120, and Uber sits in the cheapest third of its own 10-year history while generating roughly $10 billion in annual free cash flow against a ~$150 billion market cap. He leans on cash flow rather than earnings, and the key risk (driverless cars disrupting the model) is clearly identified. This is his deepest buy band and the only name where his system didn't conflict with itself.
Price target: $109
Oracle has $638B in signed contracted revenue, largely from a $300B OpenAI deal, yet trades 57% below its high. Even stripping out OpenAI entirely, the stock costs less than before the deal was announced, making it cheap on already-signed sales. The main risk is negative free cash flow from heavy data center build-out and dependence on a single unprofitable customer.
Innodata is a neutral AI data-labeling supplier whose independence became its biggest selling point after Meta bought half of rival Scale AI, driving competitors' customers away. Revenue grew 58% last quarter and customer concentration is improving, but the stock is volatile, trades at 46x earnings, and has 14% short interest.
Mayfair Gold is a pre-production gold developer in the Timmins Gold District with a pre-feasibility study showing C$652M after-tax value and 24% IRR. Insiders hold 35% and have invested C$20M over two years, with Howard Marks' Oaktree Capital and Muddy Waters among large holders. This was a sponsored segment.
Sterling does unglamorous but essential data center groundwork with a $5.5B backlog that grew 50% in a year, no net debt, and 40% ROE. The stock sold off because an acquisition temporarily lowered operating margins by 4 points, but that same acquisition lets Sterling win bigger, more complete jobs.
MasTec has a record $21B backlog tied to rebuilding the grid for AI-driven power demand, up 30% year over year. The stock fell ~40% because a small telecom division hit an air pocket, but the core power delivery business is booming and the selloff is a tail-wagging-the-dog situation. The company carries real debt and the chart is broken, so patience is warranted.
AppLovin runs an AI-powered ad engine with ~80% operating margins, higher than Visa, and trades at ~21x next year's earnings after falling over 50%. The SEC closed its investigation with no wrongdoing found, removing the biggest overhang, yet the stock dropped 20% on a slight guidance miss. Brian calls it the highest quality name on the list and is buying more once technicals stabilize.
Google is used as an example of a top-tier corporate borrower facing rapidly rising borrowing costs—its 40-year rate went from 5.75% to 6.5% in six months, partly due to Treasury yields rising and partly due to lenders charging more given its massive $200B/year data center capex. In a rising-rate scenario, Brian notes shareholders 'pay more' alongside the company.
nVent's gross margin declined 2.2 points while revenue grew 46%, indicating its bus bars and cooling loops are commoditized products that competent factories can build. The margin is flowing to the buyers rather than being captured by nVent, suggesting it lacks pricing power despite exploding demand.
Nvidia's gross margin declined to 75% and guided to 72-73%, but this is simply a return to pre-boom levels rather than a collapse. The company committed $279 billion in supply and capacity footnotes, primarily for memory and data center infrastructure, indicating Nvidia is paying through the nose for scarce memory supply.
Vertiv passes the margin test with a 4.3-point gross margin increase on 26% revenue growth, and deferred revenue doubled as customers prepay billions. However, the modest margin expansion despite massive prepayments suggests customers still have alternative suppliers, meaning Vertiv doesn't fully own the bottleneck.
GE Vernova marginally passes the test with a 2.2-point margin increase, but at a 20% gross margin it has a very long order book with little pricing power. It will do enormous business in the build-out but sells something a factory can make more of.
Apple has been testing Chinese memory for China-market iPhones and Macs, but faces bipartisan Senate opposition and Commerce Department disapproval. The creator's concern is that Apple's supplier approval would grant credibility to the Chinese memory competitor, which is more valuable than money since state funding already covers factory costs.
The creator declines to provide a valuation ladder because CoreWeave has only been public for 17 months, insufficient history to measure against. His model produces an optimistic number but he doesn't trust it given the limited data, and analysts disagree on whether demand is real or seller-financed. He stands aside rather than buying or holding.
Brian explains the mechanics behind the U.S. Treasury buying back its own 30-year bonds at 52 cents on the dollar, framing it as basic bond math rather than a crisis. He argues Washington has exhausted its tools to deliberately lower long-term rates, leaving the market to resolve it organically—likely through an economic crack. He lays out four scenarios (jobs cracking, bad auction, Fed hiking, Japan exiting) and urges viewers to identify whether they are 'paying' or 'collecting' in the current rate environment.
The creator investigates where the money from Nvidia's declining gross margins is flowing by running a gross-margin test across the AI infrastructure supply chain. He finds that memory companies—particularly Micron and SanDisk—show the strongest margin expansion, indicating they own the true bottleneck. Power/cooling companies like Vertiv and nVent Electric show weaker or negative margin trends, suggesting their products are more commoditized. Constellation Energy is the exception on the power side, as it owns already-licensed nuclear reactors that cannot be replicated. The creator flags a Chinese state-backed memory competitor as the key risk that could break the memory trade.
The creator reviewed 141 stock videos across 21 channels and found five companies most frequently discussed: Alphabet, Nvidia, Micron, CoreWeave, and Uber. He provides his own fair value estimates and entry ladders for each, while highlighting that four of the five are essentially the same AI trade (chips, clouds, models), with Uber being the only outlier. He emphasizes objective rules, pre-set exit points, and four key questions investors should ask before buying.
Brian presents five stocks he believes are on sale after a market correction, focusing on companies with large contracted revenue backlogs ('receipts') rather than speculative forecasts. He covers Oracle, Innodata, Mayfair Gold (sponsored), Sterling Infrastructure, MasTec, and AppLovin, arguing the market overreacted to short-term issues while underlying businesses remain strong.
Based on AI analysis of this creator's recent videos. Theses are paraphrases of the creator's own words, not verbatim quotes. Not financial advice.