InvestAnswers: Bitcoin's CAGR has crumbled to single digits over five years

Bitcoin's compound returns have quietly collapsed

Responding to a Patreon question about Bitcoin's long-term compound annual growth rate (CAGR), James lays out Bitcoin's CAGR across different entry points: 60.1% over ten years, 36.8% over eight years, and just 4.8% over five years. Buying three years ago yielded 30.7%, while buying one year ago produced a loss of 29.4%. He stresses that a 4.8% CAGR "is not going to get you to your 14% hurdle," which he considers the minimum needed to avoid "drowning."

He pushes the comparison further by dividing Bitcoin's price by US M2 money supply growth, finding only a 15% gain over five years, equivalent to a 2.8% CAGR. He also runs through street forecasts: ARK's bear case implies a 38.4% CAGR to a $300,000 target by 2030, its base case 70.2% CAGR for $710,000, and its bull case 103.7% CAGR for $1.5 million, which he says "is not going to happen." He cites Bitwise's $1.3 million 2035 target (36% CAGR) and Michael Saylor's 2024 plan targeting $13.4 million by 2045 at a 30.9% CAGR, but cautions that distant forecasts should be taken "with a grain of salt, the size of the moon."

James also cites a JPMorgan report suggesting Bitcoin remains the top-performing asset class over the next decade but still only at a 10% CAGR, which he says "will not save you." He contrasts this with AI-linked stocks: Micron up nearly 800%, Palantir 343%, and Nvidia 54% over two years, versus Bitcoin's 10.7%, Ethereum down 10.3%, Solana down 44%, and MicroStrategy down 46% over the same period.

Quantum computing risk remains distant but not dismissed

A viewer asks whether AI-accelerated quantum computing poses a real threat capping Bitcoin's price. James agrees the risk "isn't a zero probability risk" but disagrees with "Shemath" (as transcribed) that Bitcoin goes to zero. He notes about 7.15 million Bitcoin, roughly 35% of supply, sits in addresses with exposed public keys, including roughly 1.71 million Satoshi-era coins worth "over 550 billion" at current prices.

"I do think that the price could stay capped until this issue is more convincingly addressed." — the question read by James

He cites Google's estimate that fewer than 500,000 physical qubits could break Bitcoin's encryption "in minutes," about 20x fewer than previously thought, while Google's Willow platform has roughly 105 qubits, meaning "the gap is still" very large. James says he expects things to be fine for a few years but could "get hairy" post-2028 unless fixes like BIP 360/361 move from drafts to activation. He adds that if ECDSA breaks, RSA 2048 "is right behind it," meaning broader internet and banking infrastructure would also be compromised, so "we've got bigger worries than just Bitcoin."

War spending and defense-adjacent AI positioning

Asked about positioning for a "war trade," James says he has held heavy defense exposure in his IA13 portfolio since June 2025, citing US defense spending near $1 trillion (with an attempt to reach $1.5 trillion), $74 billion earmarked for drones and the "Golden Dome" program, NATO's push toward 5% of GDP, and global military spending of $3 trillion.

Rather than buying legacy defense primes, James says he favors AI-linked proxies: Palantir for military intelligence and enterprise AI, SpaceX for its role in Golden Dome military communications (which he estimates at a small single-digit percentage of SpaceX revenue), and Anduril for autonomous drone systems, noting Anduril is "pre-IPO stock" not available on public markets.

Small-cap risk, exponential scale, and the Tesla-SpaceX merger math

On a question comparing a small-cap like Unusual Machines to mega-caps, James cites a Bessembinder-type study showing 57.4% of stocks lose to treasury bills and 4.3% of firms generate most market wealth, plus a JPMorgan stat that 44% of Russell 3000 names from 1980–2020 fell 70% and never recovered. He argues the top 0.3% of stocks drive the majority of returns, and bigger firms compound advantages through scale, citing AWS growing from $21 million in annual revenue in 2006 to roughly $21 million "every hour" today, a 9,000x increase he calculates as a 76% CAGR.

On a potential Tesla-SpaceX merger, James reiterates his RetireOn model targets of 655 SpaceX shares and 323 Tesla shares, and notes the current market ratio is 2.97 SpaceX shares per Tesla share, close to his long-stated "three SpaceX shares per Tesla" expectation, driven by Tesla rising 38% in 18 trading days while SpaceX weakened. He says depending on the final conversion ratio (3:1 vs 4:1), a holder with roughly 322 Tesla shares would end up with either 969 or 1,300 shares of the combined company, which he expects to be named "X," though he does not know when the merger will close, guessing "before the end of 2027."