Morgan Stanley reiterates Overweight on WeRide stock, $14.70 target

Morgan Stanley reiterated an Overweight rating with a $14.70 price target on WeRide while the stock trades at $6.76 (down ~60% Y/Y and ~60% below its 52-week high of $16.86). WeRide plans to grow its robotaxi fleet from 1,125 to 2,600 units by year-end (≈+131%), with ~30% deployed internationally, and reports operational progress including 15 avg daily orders (26 peak), 5 km average trips at 2 yuan/km (50% discount to ride-hailing) and a target of 25 trips/car/day at 3 yuan/km. Financially, revenue rose 40% LTM, the company remains unprofitable with a $2.17B market cap but has more cash than debt; it achieved a 38% total cost of ownership reduction in 2025 via remote assistance ratio improvements. Separately, NVIDIA’s latest 13F shows it exited stakes in Applied Digital, Arm and WeRide, signaling notable institutional selling pressure to monitor.

Analysis

The structural winner from accelerated robotaxi deployment is the platform/marketplace layer that captures demand and pricing power while offloading asset-heavy risks. Incumbent aggregators that can integrate autonomous supply into an existing app and pricing engine will see unit economics improve faster than vertically integrated robotaxi operators, because they avoid vehicle capex, logistics and cross-border shipping exposure. Suppliers of AV compute and sensing stand to gain if volume ramps, but their order book and pricing discipline will depend on a concentrated set of large OEM/robotaxi customers — a single major contract shift could swing supplier margins materially.

Key risks are operational (remote assistance and disengagement metrics), regulatory (city-by-city approvals and insurance regimes), and funding/market-structure driven (large tech investor rotation away from early-stage AV names which reduces secondary liquidity). Near-term catalysts that will re-rate valuations are repeatable per-car utilization improvements, transparency on autonomous kit cost per unit, and multi-city commercial launches that prove cross-jurisdictional scalability. Conversely, shipment delays or a single high-profile safety incident would compress multiples quickly and could force capital raises at distressed prices.

Tactically, prefer exposure to durable demand-capture platforms and short/neutral sizing on pure-play capital-heavy operators until consistent unit economics are visible across multiple geographies. Use calendar and cross-asset hedges to protect against rapid de-risking by large institutional holders. The consensus is focused on headline expansion; what’s underappreciated is the path-dependence of per-car utilization and the nonlinear impact of fleet downtime and shipping friction on cash burn — that asymmetry favors option structures and pair trades that long platform upside while hedging execution risk.

Key Decisions for Investors

• Pair trade (6-12 months): Long UBER equity (or 9-12 month calls) / Short WRD equity — size to net 1-2% portfolio risk. Rationale: play platform monetization and demand capture vs execution/asset risk. Target relative outperformance of 30-60%; stop if the pair moves 20% against position.
• Long WRD optionality (18-36 months): Buy LEAPS (~24–36 months) rather than large outright equity exposure, funded partly by selling near-term calls or buying protective 9–12 month puts (ratio ~1:0.3). R/R: asymmetric payoff if unit economics improve (3:1 upside potential vs limited funded downside).
• Short APLD (3-9 months): Buy puts or small outright short given potential de-risking by large holders and limited liquidity. Rationale: negative investor concentration and fallback financing risk. Target 40–100% P/L on realized weakness; cap loss via buy-stop or bought call for defined risk.
• Hedge/watch (0-6 months): Buy a small NVDA 1–3 month put spread as insurance across AV/AI equity positions to protect against accelerated tech-investor de-risking events that propagate through small-cap AV names. Keep this hedge sized to cover correlated drawdown of core AV exposure (~25–50% of notional).