Bloomberg Tech
2026-06-12 · Hosted by Caroline Hyde, Ed Ludlow · Bloomberg / iHeartMedia
Executive Summary
On the eve of SpaceX’s record-breaking IPO, retail orders topped $70 billion — at least three times oversubscribed from retail alone — with the deal priced at a fixed $135/share for a roughly $1.7 trillion valuation and $75 billion raised. The unusual fixed-price, ~4% float structure dominated coverage, alongside Ark Invest’s Brett Winton arguing Starlink economics alone justify the valuation while the orbital-AI-data-center thesis (a claimed $26.5 trillion enterprise-AI opportunity) remains speculative. Oracle was the day’s worst large-cap, falling 12% ($61 billion in market cap erased) on higher-than-expected capex and plans to tap debt and equity markets, reigniting fears about AI infrastructure profitability. A separate thread tracked AI’s spillover into consumer inflation, with memory/chip prices pushing tech hardware components up a record ~15% year-over-year and potentially adding 0.4 percentage points to headline inflation.
Key Stories & Changes
1. SpaceX IPO Mechanics & Demand
SpaceX set to finalize IPO pricing; retail orders exceeded $70 billion, “at least three times oversubscribed from retail” alone
Priced as a private-style fixed deal: $135/share, fixed size, no traditional range walk-up — “this is the price, this is the size, are you in or out”
Order books closed; allocations expected this evening/early tomorrow; first trades anticipated tomorrow afternoon
~4% float — small relative to the ~$1.7T valuation; demand also strong from the Middle East, Japan (a couple billion+), and US investors
Pre-float investment-grade ratings from all three main agencies — read as a signal SpaceX will return to markets to raise more capital
BlackRock order of at least $5 billion (per WSJ); several institutions placed ~$10 billion orders (Bloomberg)
Index fast-track inclusion creates built-in passive demand; lockups expected to expire relatively quickly
2. The SpaceX Thesis — Ark Invest’s Brett Winton
SpaceX is the largest position in Ark’s venture interval fund (open to all investors)
Starlink alone could drive $300–400 billion in revenue by 2030; launching tens of gigawatts at ~$15 billion/gigawatt market rate
SpaceX has rented AI capacity to Google and Anthropic at just under $40 billion/gigawatt on a blended basis
Q1: SpaceX net lost $4.28 billion on revenue of $4.69 billion; Winton expects profitability “very quickly” as Starlink scales
Flagged a likely future Tesla merger to fund XAI/Grok catch-up; competing to the AI frontier by 2030 needs 30–50 gigawatt-years of compute
Musk target: 1 million tons to orbit in ~5 years — Ed Ludlow noted Starship’s 150 metric-ton capacity implies ~1,333 launches/year, “a long shot”
3. Oracle’s Costly AI Bet
Oracle down 12%, worst day in six months; $61 billion market cap wiped out
Quarterly capex above expectations; plans to tap debt and equity markets next fiscal year
Core question: how fast can it convert AI demand into revenue, and at what margin — gross margins falling as AI business ramps; fired thousands to control costs
“Empty calories” risk: growing massively but margin/profitability uncertain
4. AI’s Inflation Spillover
Memory chip prices skyrocketing as hyperscalers build data centers; tech hardware + electricity costs spiking
Tech/computer components in inflation data up ~15% year-over-year — a record increase
Bloomberg Economics: memory squeeze may add 0.4 percentage points to headline inflation before easing
Small/medium businesses facing much higher costs for RAM, SSDs, and new computers/laptops
5. GoPuff × SpaceX AI (Grok) Enterprise Case Study
GoPuff built its new AI shopping assistant “Go” on XAI/SpaceX models — voice shopping, auto-built carts, TikTok-style discovery
Chose SpaceX over OpenAI/Anthropic for voice + image quality, lowest-cost provider, and willingness to co-develop (10–11 month build)
Early metrics: basket sizes up, frequency up within first week
Context: SpaceX claims enterprise AI is a $26.5 trillion opportunity; WSJ reports OpenAI may cut prices aggressively
Trends Identified
1. Retail Investors as a Structural Market Force
Retail demand of $70B+ for a single IPO — enough to fund the deal on its own — marks a step-change in retail’s market power, echoing lessons from GameStop. Strategist Christina Hooper noted retail “understands the story better” and may hold longer, supporting valuation, but warned it’s a double-edged sword: if cracks form in core businesses, retail can react sharply and amplify volatility.
2. The AI Capital-Consumption Paradox
Oracle’s selloff crystallized a market-wide anxiety: AI infrastructure players cannot build compute fast enough, yet the capital required (debt + equity raises) is eroding the thesis that hyperscalers can self-fund buildouts from free cash flow. The same compute-constraint urgency that excites SpaceX bulls is exactly what’s punishing Oracle’s margins.
3. AI Costs Bleeding Into the Real Economy
The data-center boom is no longer an abstract capex story — it’s lifting memory/chip prices, raising costs for ordinary consumers and small businesses, and meaningfully nudging inflation. This connects the equity euphoria directly to policymaker concerns and the rate outlook.
4. Imagination-Driven Valuation
From orbital data centers to a $26.5T enterprise-AI TAM and 10-million-people-on-Mars ambitions, SpaceX’s valuation rests on vision rather than current fundamentals (losing money on ~$4.7B quarterly revenue). Guests repeatedly framed this as “science fiction becoming science fact,” while urging tempered, diligence-based expectations. —-
Sentiment Analysis
Overall Market Sentiment: Cautiously Euphoric
The dominant mood was excitement around SpaceX strong enough to let markets “look through” geopolitical risk, tempered by clear unease about AI infrastructure economics and elevated valuations.
Risk Factors Highlighted
AI infrastructure profitability: Oracle shows margins compressing and capital needs ballooning as AI ramps — is it a good business to be in?
Mounting debt/equity issuance: Oracle, Alphabet ($80B+) tapping markets repeatedly undercuts the self-funding hyperscaler thesis.
SpaceX execution risk: Starship reusability unproven; 1M-tons-to-orbit target implies an implausible ~1,333 launches/year.
Valuation re-rating: High valuations leave long-duration tech vulnerable if rates/yields continue rising.
Retail volatility: Heavy retail ownership could amplify swings if confidence cracks.
AI-driven inflation: Memory/chip price surge adding ~0.4pp to headline inflation, complicating policy.
Geopolitical risk: Middle East conflict cited as a catalyst keeping investors on edge.
Grok competitiveness: Whether XAI can compete with Anthropic/OpenAI at the frontier, or merely rent capacity.
Capital concentration: 65% of VC money has gone to just 0.05% of companies — a crowded, consensus AI trade.
This episode was covered in today’s The Market Signal — 2026-06-12, a cross-source synthesis of multiple podcast reports.