Premarket movers: Dell, GitLab surge and MongoDB tumbles

Oil prices edged lower amid renewed US–Iran tensions, while US index futures fell (S&P 500 -0.2%, Nasdaq 100 -0.4%) as investors weighed a global bond selloff and expectations of a Fed rate increase later this month. On earnings, Dell raised annual revenue/profit forecasts with AI-optimized server demand and HPE climbed 4.7% ahead of results, while MongoDB dropped 13.6% despite Q2 revenue up 30% to $771.8M and adjusted EPS $1.90 vs $1.61 consensus. GitLab surged 22% after beating and raising full-year guidance, while Uber gained 2.4% on a plan to cut ~3,300 jobs (~10%).

Analysis

The market is rewarding AI infrastructure names only where the story can convert into margin and guide visibility, not just order momentum. That argues for a widening dispersion trade inside the server complex: DELL and HPE can keep rerating if they prove they are monetizing demand without blowing up working capital, while SMCI remains more exposed to any normalization in pricing or customer concentration. The second-order winners are in the picks-and-shovels chain — networking, power, cooling, and memory — but the near-term bottleneck is still who can ship at acceptable gross margin, not who can claim the most AI exposure.

MongoDB’s reaction matters more than the beat itself: it signals that investors are now discounting growth quality and forward consumption trends rather than trailing revenue surprises. That keeps pressure on high-multiple software until management teams can show a path to sustained free-cash-flow conversion and not just one-quarter upside. GTLB is the cleaner relative-long because the guide raise gives it a near-term catalyst and a better operating-leverage narrative; MDB risks becoming a multiple compression proxy if any enterprise spend scrutiny spills over into the rest of infra software.

UBER’s restructuring is a margin defense move, but it is also an admission that scale alone no longer supports premium valuation without visible operating leverage. SIRI’s upgrade looks tactical rather than structural; higher rates and a weak secular growth profile limit how far a rerating can run. Over 1-3 months, the key falsifier for the bearish duration view is falling bond yields or broad software guidance upside; over 6-18 months, the real winners should be balance-sheet-strong AI enablers, not levered or low-growth names.

Key Decisions for Investors

• Long DELL / short SMCI for 1-3 months: express quality dispersion inside AI servers. DELL has the better risk/reward if the next print confirms margin discipline; invalidate if SMCI stabilizes gross margin and re-accelerates bookings.
• Long GTLB / short MDB into the next 1-2 earnings cycles: pair the name with a raised guide and operating leverage against a name the market is already punishing for any hint of deceleration. Stop if MDB guides consumption growth back above expectations.
• Buy HPE only on post-earnings weakness, not pre-earnings momentum: this is a cleaner way to own enterprise AI spend without paying peak enthusiasm. Fails if commentary points to margin dilution from mix or pricing pressure.
• Avoid chasing SIRI on the upgrade; if positioning is needed, use it as a tactical short on strength with tight risk controls. The thesis breaks if rates back down sharply or if management shows an unexpected path to sustainable growth.