Shares of Rackspace Technology jumped 9.1% to $5.16 after the company reported second-quarter results that cleared a low bar — raising a pointed question: does a slightly smaller loss justify optimism for a heavily indebted firm betting its future on artificial intelligence?
The Beat Was Real, but the Bar Was Underground
Rackspace posted revenue of $670 million, a slight increase of 0.6% year over year. The adjusted loss of $0.08 per share topped the $0.09 consensus estimate. That follows Q1, when the company beat revenue expectations by 2.6% and swung to $0.03 earnings per share versus an expected $0.22 loss. Two consecutive beats look like momentum — until you note that net losses actually widened to $68 million from $55 million a year ago, with loss per share deteriorating to $0.27 from $0.23.
Private Cloud Is Growing, but Public Cloud Is Shrinking
Private Cloud revenue hit $263 million, up 5.5% year over year, while Public Cloud revenue fell 2.3% to $407 million.
The public cloud decline signals potential customer attrition or pricing pressure — exactly the low-margin business Rackspace says it is deliberately exiting. Management is scaling back colocation and traditional hosting to free capacity for enterprise AI workloads. The risk: old revenue disappears before new revenue arrives.
The AI Bet Is Expensive and Unproven
Rackspace signed a definitive deal with AMD for 30 megawatts of GPU and CPU capacity across global data centers from 2026 through 2028.
Each megawatt is expected to generate $15–$20 million in annual revenue with EBITDA margins above 50% — far richer than the legacy business. But to fund this, the company announced a $250 million at-the-market equity offering , meaning new shares sold gradually that dilute existing investors. Meanwhile, interest expense surged 60.6% to $34.2 million, and total liabilities stand at $4.03 billion against a stockholders' deficit of $1.28 billion.
Guidance Holds, but the Margin for Error Is Razor-Thin
Full-year guidance calls for $2.45–$2.55 billion in revenue and $285–$295 million in adjusted EBITDA.
Management says benefits from new AI deployments won't materialize until 2027.
RBC raised its price target to $4 and UBS to $5.50 — meaning today's price already sits at the high end of analyst expectations. Investors buying here are paying for an AI future that remains, for now, entirely on paper.