Deep Dive
Crypto Inefficiency and Exchange Inflows
InvestAnswers kicks off with frustration over crypto's illogical valuations. Ethereum processes $114,000 of market cap per daily transaction versus Solana at just $500—a 230x gap that makes no sense fundamentally. Recent data shows 8,500 Bitcoin hit exchanges over three weeks, breaking a longer outflow streak, but critically the holders aren't selling into rallies. ETF flows turned positive with $192 million into Bitcoin last week and $118 million total inflow, marking the fifth consecutive positive week. Ethereum saw $81.6 million exit—unusual given its size—with rumors of unstaking pressure. The broader message: if Bitcoin and crypto were purely rational pricing mechanisms, Solana should dominate Ethereum on transaction efficiency alone, yet market cap tells a different story.
The PMI Correlation and AI Capex Boom
This is where InvestAnswers plants his flag. Every other Bitcoin correlation has failed—the four-year cycle, historical blowoff tops, seasonal patterns. The lone survivor is the PMI business cycle framework. He maps three historical bull runs: 2011-2013 saw PMI recovery drive Bitcoin up, 2016-2018 had PMI peak at 60-plus with another bull, and 2020-2021 delivered a massive spike to 64-65 on COVID stimulus. Now comes the capex boom for AI, which InvestAnswers calls bigger than freeways, railways, and factories combined. ISM manufacturing PMI is currently in the 50s, signaling expansion. His bet: if Bitcoin doesn't pump on this capex cycle, he's done chasing correlations. The stakes feel real—he's explicitly betting his credibility on PMI holding as the last explanatory variable for crypto's next leg.
Gold-Bitcoin Flip and Liquidation Carnage
A chart tracking gold versus Bitcoin reveals a pattern worth watching. End of 2021 saw gold smash Bitcoin from roughly 36 ounces per BTC down to 9 by late 2024. Then Bitcoin soared relative to gold from late 2024 through August-September 2025. Then it tanked—gold rallied hard, Bitcoin underperformed. If these cycles last two years as historically typical, the pattern suggests a Bitcoin-bull/gold-bear trade could be forming. The math is staggering: if Bitcoin ever reaches gold's total market cap, each coin prices at $1.82 million. Meanwhile, Bitcoin bears have liquidated $8.7 billion since February with no major catalyst—they just keep shorting at $80,000 expecting $40,000 and keep getting whacked. The green liquidation spikes on his chart show the highest squeeze since early 2025. Bears are wrong 80% of the time historically; this liquidation carnage suggests they're wrong again.
AI Stocks and the Capex Explosion
Stock fear and greed flipped to 67—pure greed. The market is grabbing AI exposure with both hands. Micron jumped 900% in a year on AI chip demand, and Intel, dormant for 26 years, exploded 30% in a single week. Google added $1.4 trillion in market cap over 30 days—wealth created from thin air. AMD faces earnings today after rallying 74% in April, now outside what he calls the confidence cloud; historically, stocks that break above the cloud get pulled back in. The S&P 500 earnings growth outlook charts show extreme consensus optimism driven entirely by AI. InvestAnswers' thesis: ride these horses until they run out of steam, then rotate. Everyone's a hero in a secular bull market, but separating winners from losers requires timing the rotation. AI is driving everything right now—it's the money printer going burr.
Tesla Robo-Taxi Rollout and Inference Dominance
Tesla's Cybercab manufacturing line can produce one unit every 5 seconds. With current capacity making 2 million cars annually, hitting 10,000 robo-taxis by year-end is trivial once safety clears. New job postings for field response specialists in Florida and Nevada signal imminent rollout to those states—Tesla doesn't hire until they're ready to deploy. If 10,000 Cybercabs hit by year-end, one analysis suggests Tesla stock reaches $1,000 from current ~$400 levels. The broader AI story is inference over training: self-driving cars, humanoid robots, Cybercabs don't need learning capability—they need real-time inference in 20 milliseconds. Jensen Huang reportedly said inference will be 99% of future compute. This represents a structural shift in AI workloads and capital spending, not just another quarter of GPU sales.