Back to News
Market Impact: 0.28

Buy These 5 Small and Mid-Sized AI Stocks for Stellar Returns in 2026

PATHFIVNQLYSTDCBBMSFTNVDASNOWSAPCRMNOWVRNTGOOGLACNIBMCTSHINFYHUMNDAQ
Artificial IntelligenceTechnology & InnovationCybersecurity & Data PrivacyCorporate EarningsAnalyst EstimatesProduct LaunchesAutomotive & EVCompany Fundamentals
Buy These 5 Small and Mid-Sized AI Stocks for Stellar Returns in 2026

Zacks highlights five small-to-mid cap AI-related names—UiPath (PATH, Zacks #1), Five9 (FIVN, #2), Qualys (QLYS, #1), Teradata (TDC, #1) and BlackBerry (BB, #2)—as potential 2026 outperformers on continued AI infrastructure momentum and recent product rollouts and partnerships. The firms show projected revenue/earnings growth: PATH +9.3% revenue / +13.7% earnings for year ending Jan 2027 (consensus EPS +5.6% last 30 days); FIVN +9.5% revenue / +8.3% EPS (current year, EPS +1.3% last 60 days); QLYS +7.7% revenue / +6.5% EPS (current year, EPS +10.4% last 60 days); TDC −0.6% revenue / +3.6% EPS (current year, EPS +8.3% last 60 days); BB +7.4% revenue / +13.3% EPS for year ending Feb 2027 (EPS +6.3% last 30 days). Key drivers cited include new generative AI features and LLMs at UiPath, Five9’s Genius AI on Google Cloud, Qualys’ AI/security acquisitions, Teradata’s vector/store and Agentic AI capabilities, and BlackBerry’s QNX wins across ~275M vehicles.

Analysis

Market structure: Small/mid AI infrastructure and security vendors (PATH, FIVN, QLYS, TDC, BB) are primary beneficiaries as enterprises shift capex to software-defined automation, contact-centers-as-code, vector stores and AI-native security. Legacy on-prem vendors and non-AI SaaS products face pricing pressure and churn; hyperscalers (MSFT, GOOGL) and GPU suppliers (NVDA) capture disproportionate margin upstream, concentrating bargaining power. Strong enterprise demand implies continued ARR growth of ~7–12% for winners but potential ASP compression over 12–24 months as features commoditize. Cross-asset: expect elevated equity IVs in AI names, modest tightening in IG spreads if growth sustains, higher electricity demand and copper intensity for data centers, and marginal USD strength on tech-led risk-on flows. Risks: Tail risks include rapid regulatory constraints on LLMs/data (policy shock within 6–18 months), a sudden GPU oversupply that collapses pricing, or major cybersecurity incidents undermining trust in AI ops. Short-term (days–weeks) volatility will hinge on earnings/partnership announcements; medium-term (3–12 months) on product integrations with MSFT/NVDA/GOOGL; long-term (1–3 years) on ARR monetization and free-cash-flow conversion. Hidden dependency: heavy reliance on hyperscaler integrations and Nvidia GPUs; loss of one partner could cut TAM access by >20% for certain vendors. Key catalysts: large enterprise design wins, hyperscaler co-sell ramp or meaningful beat-and-raise on ARR guidance. Trade implications: Direct longs — establish concentrated exposure to PATH and QLYS where fundamentals + partnerships imply 20–30% upside if ARR execution holds; size 1–3% portfolio positions each. Pair trade — long FIVN (2%) / short VRNT (1.5%) over 3–9 months to capture migration to cloud-native contact-center AI. Options — use 3–6 month call spreads on PATH/QLYS to cap premium; buy a 3-month NASDAQ-100 put spread (tail hedge) sized to 2–3% of portfolio for macro risk. Rotate 4–6% from non-AI legacy software into cybersecurity and automation over next 3 months. Contrarian angles: Consensus underestimates integration risk and customer churn from poor AI ROI — contracts may extend sales cycles by 20–30% for some vendors. Some small/mid names are under-owned; PATH and QLYS could re-rate materially on 2–3 sequential quarters of ARR beat, but valuation dispersion is wide so idiosyncratic risk dominates. Historical parallel: 2013–15 SaaS consolidation where feature-led winners gained enterprise penetration — winners then captured 30–50% higher multiples. Unintended consequence: rapid agentic-AI adoption increases hyperscaler and chip leverage, turning software vendors into service resellers if they fail to capture model/IP margins.