Nvidia CEO Jensen Huang "Loves" This Artificial Intelligence (AI) Company. The Stock Could Soar 77% in 2026, According to 1 Wall Street Analyst

Serve Robotics, a maker of sidewalk food-delivery robots and a strategic partner/former investee of Nvidia, reported Q3 revenue of $687,000 (up 209% year-over-year) while posting a $33 million loss that widened nearly fourfold; management projects a tenfold revenue increase by 2026. The company claims the largest U.S. sidewalk delivery fleet (over 2,000 robots), rapid operational growth (delivery volume +66% quarter-over-quarter and +300% year-over-year), and multiyear commercial partnerships including DoorDash and Uber, but trades at over 400x sales and remains unprofitable despite unanimous buy ratings (average price target ~ $19, Street-high $26).

Analysis

Market structure: Nvidia (NVDA) and large platform partners (DASH, UBER) are the primary beneficiaries — NVDA sells chips and software, DoorDash/Uber capture margin on asset-light fulfillment while offloading capital risk. Serve (SERV) could win share in dense urban corridors if it reaches scale, but labor-cost parity (management's $1/delivery claim at 2.5 miles) is the choke point that determines pricing power and restaurant economics. Incumbent human-delivery models face margin pressure in high-density zones, while equipment manufacturers, insurers, and local telecom/edge-compute vendors stand to gain from physical-AI infrastructure demand.

Risk assessment: Key tail risks are regulatory bans/sidewalk restrictions, single-incident public safety/PR events, and capital exhaustion — at ~ $33M GAAP loss in Q3 and revenue of $0.687M, a continued burn at Q3 levels implies runway measured in quarters without financing. Immediate risks (days–weeks): headlines or municipal rulings; short-term (months): door-to-door rollouts and partner KPIs (fail rate, uptime) that must improve >50% QoQ to justify valuations; long-term (years): successful unit economics at scale (target >$1/delivery sustained + maintenance cost <20% of revenue).

Trade implications: Favor long exposure to NVDA (enabler) and platform partners (DASH, UBER) over pure-play robotics hardware. For high-risk allocation, a small tactical long in SERV (<2% portfolio) with strict downside controls is reasonable; alternatively, short-biased trades or put spreads on SERV capture the 400x sales valuation mismatch. Use pair trades (long UBER or DASH, short SERV) to isolate execution/regulatory risk while maintaining exposure to last-mile demand.

Contrarian view: Street optimism (consensus Buy, $19 PT) underweights execution and funding risk; 2,000+ robots vs plans for >1M is classic hardware scale friction — historical parallels include early autonomous vehicle vendors that required multiple financing cycles. If SERV fails to show step-function improvements in revenue per robot or secure profitable pricing within 6–12 months, downside could be 50–80%. Monitor non-obvious dependencies: municipal permitting cadence, insurance cost per incident, and NVDA’s ongoing support/license terms as optional de-risking events.

Key Decisions for Investors

• Establish a tactical, high-risk long in SERV (ticker: SERV) sized 1–2% of liquid portfolio now; set a hard stop-loss at -50% and add only if quarterly revenue exceeds $5M or delivery volume growth >3x QoQ (confidence trigger) within next 12 months.
• Implement a relative-value pair: short SERV equal notional vs. long UBER (ticker: UBER) or DASH (ticker: DASH) sized 2–3% net exposure to exploit SERV execution/regulatory risk while retaining exposure to last-mile demand; reassess after 2 quarters or material partner KPIs release.
• Buy a protective put spread on SERV (if liquid) or long-dated (9–12 month) put calendar: buy 2026 Jan puts ~10–20% OTM and sell nearer-term puts to finance cost, targeting max drawdown protection if SERV falls >40% on regulatory/event risk.
• Increase exposure to NVDA (ticker: NVDA) by 1–3% via LEAP call spreads (12–18 month) to capture upside from physical-AI acceleration while limiting capital — example: buy 2027 Jan 1 LEAP 1–2 call spread to the 1.5x strike to cap premium and time for adoption signals.
• Reduce/avoid direct exposure to small-cap robotics/hardware peers and reallocate 3–5% to platform-enabled logistics (UBER, DASH) and semiconductor enablers (NVDA) over the next 3–6 months pending municipal regulatory developments in Chicago, LA, Dallas, Miami (monitor permit denials or restrictive ordinances; if any city restricts robots, increase short SERV weight).