Arbe Robotics Ltd.

Executive Summary

SELL, Medium conviction. ARBE rates a SELL at Medium conviction. The stock has already collapsed 78.4% from its 52-week high, but that decline reflects a real reset in automotive production timing rather than an established earnings floor. Defense wins and the Gryphon launch create optionality, yet limited backlog, continuing losses and recent equity financing leave the path to self-funded scale too uncertain at $0.46.

Investment Thesis

The central error in owning ARBE at this price is treating technical validation as commercial validation. The company has real assets: a radar platform that can serve automated-driving and unmanned-ground-vehicle applications, a defense integrator framework, and improving gross profit behavior. The October Gryphon launch is strategically sensible because defense procurement can be faster than passenger-vehicle design cycles. But strategic relevance is not the same as a revenue base capable of supporting the present valuation.

The case depends first on revenue conversion. Management has identified chips, systems and engineering work as the source of recent growth, but those activities are still too small to establish recurring sensor economics. Second, the company must prove that the cost reductions are structural. A lower quarterly burn helps extend runway, but it does not eliminate dilution if commercial revenue remains below fixed R&D and selling costs. Third, HiRain and robotaxi activity must become serial production shipments rather than another round of technically successful trials with deferred purchase decisions.

The market has already punished the shares severely, which creates a temptation to buy a cheap-looking option on autonomy. It is not cheap on operating evidence. The stock is 47.5% below its 200-day average and 97.3% below its five-year high, but neither statistic establishes intrinsic value. A 4D radar winner could be worth materially more if autonomy adoption accelerates and customers standardize on Arbe's architecture. Conversely, AI may consolidate value in perception software, compute platforms and integrated sensor suites, leaving a small standalone radar vendor with weaker bargaining power.

Medium conviction reflects that the downside to the base case is not enormous from $0.46, but the evidence required to underwrite a durable upside case is missing. Avoiding the shares is preferable until booked orders, production revenue and cash burn show that the product roadmap has crossed from promise to repeatable commercial demand.

Key Metrics

Financial Strength

Liquidity is adequate for the immediate operating plan, supported by $41.9M of Q2 2026 cash, the subsequent $15M offering and a 2.21 current ratio as of December 31, 2025. Solvency is not the near-term problem: debt/equity was 0.68 as of that date. The problem is earnings absorption. LTM to June 30, 2026 operating margin was -2,374.3%, an improvement from FY2025 but still evidence that modest revenue gains do not yet cover the cost structure. Margin direction is improving at the gross-profit level, while cash generation remains unproven.

Competitive Position

Arbe's technical edge is its 4D imaging radar platform, combining radar hardware, processing and perception capability for difficult weather, terrain and autonomous-driving use cases. Customer validation through robotaxi trials, HiRain and defense projects is more valuable than the current revenue base, but the moat is not yet durable because large OEM production awards and scale manufacturing are still absent. Larger automotive sensor suppliers can bundle radar into broader platforms, making execution and time-to-production more important than patents alone.

Management & Guidance

Management reiterated its FY26 outlook rather than raising it after Q2 growth, which is appropriately cautious but not a sign of accelerating order visibility. The explicit targets are $4M-$6M of revenue, adjusted EBITDA loss of $28M-$31M and cash burn below $7M per quarter from Q3 onward. The cost target has some support because expenses had already declined and the 15% reduction was reportedly implemented versus Q1. Revenue guidance deserves less confidence because backlog was limited and the company has previously pushed automotive production expectations out by roughly a year.

Segment Analysis

• Chips, radar systems and engineering work (Not disclosed): These revenue streams drove Q2 2026 revenue of $0.7M; the company does not disclose separate revenue shares.
• Automotive ADAS and Level 4 autonomy (Not disclosed): HiRain's China Level 4 program and Phoenix robotaxi trials are prospective production channels, not disclosed revenue segments.
• Defense and homeland security (Not disclosed): The global integrator framework covers 3 projects, and Gryphon targets unmanned ground vehicles; management identifies this as a shorter-cycle opportunity.

Capital Allocation

Capital allocation is currently survival-oriented rather than shareholder-return-oriented. LTM to December 31, 2025 FCF was $0.0B, with $0.0B of buybacks, dividends and capex reported, while the September 2026 $15M direct offering supplied external capital. The useful discipline is the targeted lower cash burn and 15% cost-base reduction; the unfavorable tradeoff is that future investment in product and customer programs is likely to remain equity-funded until revenue becomes material.

Management Execution & Track Record

The record is mixed and argues for limited trust in timing claims. In Q3 2024, management guided FY2024 adjusted EBITDA loss of $30M-$36M, expected production-chip shipments in early 2025, a revenue ramp in H2 2025 and HiRain mass production in Q4 2025. By Q2 2026, the business had generated $0.7M quarterly revenue and still guided only $4M-$6M for the full year, indicating that the anticipated automotive scale-up slipped materially. Cost execution has improved: Q2 adjusted EBITDA loss narrowed to $8.7M from $8.9M year over year, and management delivered a 15% cost-base reduction. Earnings relative to estimates are also improving, with three consecutive 2026 quarters meeting or beating EPS expectations after three misses across 2025. Still, repeated production timing slippage and the CFO transition, with Assaf Pereg starting August 30, 2026 after Karine Pinto-Flomenboim's departure, leave management credible on cost control but not yet credible on commercialization timing.

Positioning & Flows

Ownership is relatively lightly institutionalized: the top three institutions hold 14.2%, while insiders own 6.6%, limiting the presence of a deep long-only holder base. Short interest is 10,888,955 shares, or 10.93% of shares, with 4.1 days-to-cover and a 27.1% change, leaving room for an earnings-driven squeeze but also signaling active skepticism. Options gamma is positive within 45 days to expiry, though zero-gamma and call-wall levels are unavailable, so positioning does not supply a reliable upside price level.

Bull Case

• A global defense integrator selected Arbe technology for 3 projects under a framework agreement, opening shorter-cycle system sales beyond slow automotive qualification cycles.
• Q2 2026 gross loss improved to roughly breakeven from a $0.2M gross loss a year earlier, showing that small revenue growth can materially improve unit economics.
• Management reduced the cost base about 15% versus Q1 2026 and targets quarterly cash burn below $7M beginning in Q3.
• Phoenix-equipped robotaxi vehicles have entered trials, while HiRain's Chinese Level 4 program is moving toward planned production in coming months.

Bear Case

• The 32.43x TTM P/S ratio is 13.0x the 2.5x peer average despite FY2025 revenue declining 51.1% year over year.
• FY2025 operating margin was -7,489.9%, worse than -6,465.2% in the prior fiscal year, underscoring how far revenue must rise before fixed costs are covered.
• Q2 2026 net loss was $9.1M, while the company targets $28M-$31M of FY26 adjusted EBITDA loss, making future capital raises plausible.
• Prior management expected a revenue ramp in H2 2025 and HiRain mass production in Q4 2025; the present outlook instead points to meaningful growth in H2 2026 into 2027.

Valuation & Price Target

Engine price target $0.39 (-15.4% vs current), Low confidence.

• P/S Multiple: $0.02 (6% weight)
• Quality-adjusted: $0.51 (20% weight)

The quant engine's HOLD reflects a low-confidence $0.39 target, only 15.4% below the current price, and a quality score of 62/100. The SELL diverges because the quality-adjusted method yields $0.51, yet that outcome assumes value for a future platform before the company has proven its production ramp; the P/S method produces only $0.02 and highlights the valuation dependence on a distant revenue outcome. ARBE's 48/100 Moderate risk score understates the practical financing and timing issue created by a $1M backlog against a large FY26 loss. There is no sell-side coverage available to cross-check, which is context rather than support for either view.

Risk Assessment

Risk score 48/100 (Moderate). The principal risk to shareholders is that the $1M backlog fails to convert into a durable production ramp before the guided $28M-$31M FY26 adjusted EBITDA loss consumes the benefit of recent financing.

The Bottom Line

ARBE is a SELL for the next 6-12 months. The company has enough liquidity and technical credibility to survive, but not enough contracted demand to justify paying a premium sales multiple for a loss-making platform. The next results need to show revenue conversion and lower cash burn, not simply more pilots or product announcements. The rating would change with a material backlog increase, evidence of repeat defense-system orders, or a production ramp that puts the business on a credible path beyond the current FY26 loss guidance.