# SpaceX Has $100 Billion in Cash. So Why Is It Borrowing $20 Billion More? > Source: > Published: 2026-06-23 19:43:25+00:00 Fresh off an IPO that valued the company at nearly $1.8 **trillion** and raised more than **$85 billion** in fresh capital, [according to SpaceX’s bond](https://www.wsj.com/livecoverage/stock-market-today-dow-sp-500-nasdaq-06-23-2026/card/-analysts-on-spacex-s-borrowing-plans-al69FcdHYdg7NnygOBsD) filings, SpaceX is already back at the capital markets window. This time, it’s selling bonds—[at least $20 billion worth](https://www.axios.com/2026/06/23/spacex-musk-bonds-stock), with maturities stretching out to 30 years. The company reported **$100.8 billion** in cash as of June 19. That’s more than [Microsoft](https://www.gadgetreview.com/microsoft-is-force-installing-its-copilot-app-on-windows-11-again) held at the end of its last fiscal year. And yet here SpaceX is, borrowing again. The explanation is less dramatic than it sounds. The implications are not. ## The Actual Reason Behind the Bonds *This is financial housekeeping layered on top of a cash-burn rate that would make most CFOs sweat.* Before the IPO, [SpaceX](https://www.gadgetreview.com/spacex-stock-is-down-16-4-and-the-real-selling-hasnt-even-started) took out a [$20 billion bridge loan](https://www.reuters.com/business/media-telecom/spacex-bankers-prepare-bond-sale-least-20-billion-bloomberg-news-reports-2026-06-18/)—arranged by Bank of America, Citigroup, Goldman Sachs, JPMorgan, and Morgan Stanley, according to Reuters. That short-term facility funded the acquisition of Musk’s AI company [xAI](https://www.gadgetreview.com/residents-living-near-xais-data-centers-are-furious-about-the-spacex-ipo) and bridged the gap to going public. The bond sale replaces that expensive, short-dated debt with cheaper, longer-term investment-grade paper maturing between 5 and 30 years. Standard corporate treasury work. The spending firepower behind it is anything but standard. **Negative free cash flow** hit $14.1 billion last fiscal year—more than double the prior year’s $5.4 billion, per SpaceX’s bond filings. KeyBanc analyst Michael Leshock estimates roughly $28 billion in negative FCF this year alone. Meanwhile: - AI compute contracts with Google and Anthropic reportedly total nearly **$75 billion** over time - A deal with startup Reflection could generate **$6.3 billion** in sales - Data-center-related bond issuance in 2026 has already reached **$165 billion** year-to-date, according to JPMorgan—surpassing all of 2025 before the midpoint of the year, a trend exemplified by initiatives like the[Stargate Project](https://www.gadgetreview.com/openai-and-partners-launch-500-billion-stargate-project) [S&P Global warned](https://www.spglobal.com/ratings/en/regulatory/article/-/view/type/HTML/id/3584843) that “investments, especially in the AI business, should result in billions of dollars in free cash flow deficits for the next few years.” Moody’s flagged “elevated execution and financial risks.” Both agencies issued investment-grade ratings anyway. That tension—blessed by credit analysts, haunted by the cash-flow math—is worth examining closely. ## When the Market Doesn’t Buy the Story *A 20% stock drop in three trading days signals what investors actually believe about the bill coming due.* The market’s verdict landed fast and hard. SpaceX stock dropped more than **16%** in a single session after the bond announcement, erasing roughly **$400 billion** in market cap. Over three trading days, shares fell more than 20% from post-IPO highs—even after the $6.3 billion Reflection compute deal was announced that same week. Revenue headlines didn’t move the needle. The leverage did. Investors aren’t afraid of SpaceX’s ambition. They’re afraid of the tab—and whether they’re [paying too much](https://www.gadgetreview.com/things-youre-paying-too-much-for-without-realizing) for a runway that stretches years into the future. For anyone holding shares or watching the AI infrastructure financing wave, the risks are concrete: - Dilution from future equity raises - Leverage from the bond stack - An FCF-positive timeline that analysts don’t see arriving until around **2030** Bulls argue the Starlink-plus-AI-compute moat justifies every dollar of capex. Skeptics hear echoes of the 1990s fiber-optic boom—transformative infrastructure, spectacular overcapacity, and a very long wait for returns that tested even the most patient investors. SpaceX isn’t running out of money. It’s running a high-wire act where the burn rate is massive, the runway is long, and the payoff depends entirely on AI economics that nobody has fully priced in yet—including whether Starlink and xAI can deliver at the scale and speed this financing strategy demands.