Mackenzie Sigalos 2.8 14 ideas

Google is poised to dramatically expand its defense business by deploying Gemini AI in classified Pentagon systems, capitalizing on Anthropic's failed talks, due to its long-standing government relationship, own chips, cloud business, and removal of ethical bans, making it a frontrunner for Pentagon AI contracts.

Meta is making a long-term bet on building custom AI chips through partnerships like with Broadcom, following Google's playbook, to reduce dependence on NVIDIA and build cheaper, more energy-efficient AI infrastructure, despite lacking a cloud business to directly monetize the infrastructure and relying on ads to pay for the massive capex.

The extended partnership with Meta for the design of custom AI accelerators through 2029 makes Broadcom the near-term winner, as investors are focusing on the stock due to the positive market reaction.

Speaker reported that Google's TPU chip business could capture 20% of the AI market (a $900B opportunity), underscoring Google's vertical integration edge in AI with chips, cloud, models, and products under one roof, and noted Google trades at a discount to Apple and Amazon with a new revenue driver not fully priced in by the market. TPUs are transitioning from internal tools to a core revenue pillar, with demand accelerating from customers like Anthropic, offering cheaper and more efficient alternatives to NVIDIA GPUs for specific AI workloads, supported by a five-year deal with Broadcom. The combination of a large market opportunity, competitive advantages in integration and cost, and current market discount suggests upside potential, warranting a LONG view. TPUs may fail to gain broad adoption due to flexibility limitations, NVIDIA's strong market position, or execution challenges in scaling the chip business.

Google is monetizing its in-house TPU chips by selling them to third parties like Anthropic, Meta, and Apple, while its ad revenue continues to beat estimates. Google's vertically integrated AI stack—including models, cloud platform, distribution, and chips—is becoming a significant revenue driver, alleviating past cannibalization fears. With revenue expanding, a competitive advantage in AI infrastructure, and positive analyst estimates, Google is well-positioned for growth, justifying a LONG view. Heightened competition in AI, slower-than-expected monetization, or economic downturns impacting ad revenue could pose challenges.

Broadcom shares are moving higher after agreeing to a five-year deal to develop and supply Google's in-house AI chips. This deal establishes Broadcom as a critical supplier in the AI custom chip era, likely boosting its revenue and market position. The positive stock reaction and strategic role in a high-growth market suggest upside potential, supporting a LONG view. Execution risks, increased competition, or shifts in Google's chip procurement strategy could undermine the thesis.

Apple lost the first round of AI due to privacy constraints, is now opening Siri to rival chatbots and using Google's Gemini models, and was the best performer in the MAG-7 last quarter. Apple's strategy of avoiding massive CAPEX—unlike punished rivals—protects its financials and stock performance, while partnerships aim to address AI lag without heavy spending. WATCH because Apple shows strategic patience and cost discipline that benefits current stock, but its ability to catch up in AI via Siri improvements is critical for future competitiveness. Failure to effectively integrate AI or sustained lag behind rivals in model development could undermine Apple's market position and stock gains.

The speaker explicitly states Google is "the biggest beneficiary" of the battle between Anthropic/OpenAI and the Pentagon. Google is actively courting the Pentagon, has updated its AI principles to allow defense work, and has already landed a new AI deployment deal. While competitors face public and legal battles (Anthropic's lawsuit, OpenAI's heat), Google is positioning itself as a reliable, lower-conflict provider of the same caliber AI tools to the Department of Defense. LONG because Google is capitalizing on a competitor's weakness to capture significant government contracts in a high-stakes, high-value market (national security AI), signaling a major business expansion. The thesis breaks if internal employee revolt (similar to 2018's Project Maven) successfully forces leadership to reverse course, damaging client trust and halting the growth initiative.

"Wells Fargo may soon offer new crypto services... moved to trademark the term WFUSD... JP Morgan told CNBC over the summer it was developing a stable coin like deposit token dubbed JPMD. Federal regulators issued new guidance... banks shouldn't have to hold extra capital against securities that are tokenized." Favorable regulatory guidance (capital parity for tokenized assets) removes the primary financial penalty for traditional banks engaging in blockchain technology. Large TradFi institutions will now aggressively launch proprietary stablecoins and tokenized services. This allows them to capture new high-margin revenue streams (staking, trading fees) and retain customer deposits that would otherwise leak to crypto-native platforms like Tether or Circle. LONG WFC / JPM as they leverage their massive existing deposit bases and new regulatory clarity to dominate the institutional stablecoin and tokenized real-world asset (RWA) markets. Crypto-native competitors maintain insurmountable network effects, slow consumer adoption of bank-issued tokens, or future SEC/Fed leadership reverses the favorable capital parity guidance.

Shares popped 15% after President Trump expressed support for "market structure legislation" and explicitly stated, "We are not going to allow [banks] to undermine our powerful crypto agenda." Regulatory uncertainty has been the primary overhang for Coinbase. Presidential backing for market structure bills—specifically regarding stablecoins and custody—removes existential risk and potentially unlocks new revenue streams (like stablecoin rewards) that banks were previously blocking. LONG. The political environment is shifting from headwinds to tailwinds for the dominant US exchange. Legislation fails to pass Congress or bipartisan support fractures.

Miners are the primary source of marginal selling pressure right now as they sell coins to keep lights on. With Bitcoin near their production cost ($60k), miners are in survival mode. They are immediately selling mined Bitcoin to service debt. Some are pivoting to "High Performance Computing" (AI) to survive. Production cost cited at $60k. If BTC drops below $60k, widespread bankruptcies and industry consolidation will occur.

The "flywheel" strategy—issuing stock to buy Bitcoin to boost share price—has broken. Previously, MSTR traded at a premium to its Bitcoin holdings. Now, the stock trades *below* the value of its Bitcoin. This prevents the company from efficiently raising capital to buy more BTC. Additionally, the company faces a massive unrealized loss ($17B+) and has long-dated debt maturing next year. MSTR posted a $17B+ unrealized loss; stock trading at a discount to NAV (Net Asset Value). If confidence erodes further, MSTR may be forced to sell Bitcoin to service debt or pay dividends, adding massive sell pressure to the market.

Bitcoin is trading around $68k-$69k, having erased post-election gains. The market is seeing heavy selling from ETFs and a lack of new money. The asset is currently trapped between selling pressure and a structural floor. * Bearish Pressure: ETFs have shifted from buying to selling ($5B outflows in 3 months). Holders are underwater (average cost basis ~$90k), creating "overhead supply" where investors sell rallies to break even. * The Floor ($60k): This is the breakeven price for miners. Below $60k, miners stop selling because they go out of business, or they hold to survive, reducing supply. $5B net outflows from ETFs; 10x Research notes ETF holders are down ~31%. A break below $60k could trigger a "death spiral" where miners shut down, reducing network security and forcing liquidations.