DA Davidson reiterates SoundHound AI stock buy rating at $14 target

Q4 2025 EPS of -$0.02 beat the -$0.10 estimate and revenue was $55.1M vs a $54.29M forecast, with LTM revenue growth of 99%. D.A. Davidson reiterated a Buy and $14 price target (≈90% above the $7.39 share price) while H.C. Wainwright cut its PT to $20 from $26; InvestingPro flags the stock as overvalued and the company remains unprofitable. CFO Nitesh Sharan will depart on April 3, 2026, with CPO James Hom as interim—company retains a healthy balance sheet (current ratio 4.59, more cash than debt) and the firm does not expect a material strategy change.

Analysis

On-device, multimodal inference for vehicles materially shifts wallet share away from cloud-first AI providers toward silicon-and-stack incumbents; expect edge-AI wins to cascade into procurement preferences by Tier-1 suppliers who prize validated software/SDK relationships over standalone models. That elevates capture rates for companies providing the full hardware+software integration, and increases lumpiness of revenue for smaller specialist vendors that rely on licensing or OEM pilots — one multi-quarter OEM qualification miss can erase two years of modeled growth for a small-cap AI vendor.

A governance/financial-execution vector is underappreciated: management instability at nimble AI firms historically compresses forward revenue multiples by ~20–35% for 3–9 months as buyers mark down execution probability and factor in refinancing/M&A risk. Near-term catalysts that will re-rate or de-rate these names are product demos at major industry conferences (days–weeks), auto OEM qualification milestones (3–12 months), and the completion of C-suite hires (1–4 months). Tail risks include qualification setbacks, sovereign auto-supplier consolidation, or a broad software valuation derating — any of which flips conviction quickly.

This creates a tactical spread opportunity: overweight the hardware/stack beneficiaries to play durable edge demand while expressing a controlled short stance on smaller, execution-sensitive AI platform vendors. Position sizing should reflect asymmetric payoff — hardware beneficiaries tend to comp backlogs into FCF sooner, while small vendors carry binary risk around contracts and capital raises. Monitor OEM win announcements and CFO/finance hires as stop-loss and re-assessment triggers.

Key Decisions for Investors

• Long NVDA (3–9 month horizon): Buy a 3–6 month call spread (e.g., buy 1 ATM call, sell a 15–20% OTM call) sized to 2–3% portfolio. R/R: capped upside ~2.5–4x premium if edge inference narrative accelerates post-conference; downside limited to premium (loss = 100%).
• Long SMCI (6–12 month horizon): Buy stock or 6–12 month LEAPs sized 1–2% portfolio to play server/order-book pull-through from edge-AI validation. R/R: target +30–50% on successful channel ramps; set tactical stop at -15% to limit execution/valuation risk.
• Hedge/short SOUN (3–6 month horizon): Establish a modest short position via a put spread (buy 6-month 30% OTM put, sell 15% OTM put) sized 0.5–1% portfolio to express execution/valuation downside while capping margin. R/R: ~3–5x payoff if multiple compresses/announcements slip; max loss limited to premium paid.
• Pair trade (3–9 months): Long SMCI (or NVDA call spread) / Short SOUN put spread sized dollar-neutral to isolate hardware vs execution risk. R/R: historically this pairing captures 200–400bps alpha in the first 3–6 months during edge-adoption cycles; reduces net market exposure.
• Event hedge: If SOUN announces a material OEM qualification or major CFO hire within 4 months, trim shorts and reallocate to hardware names. Conversely, if demos disappoint at major conferences, take profits on hardware longs and increase short exposure to small-cap AI platforms.