Back to News
Market Impact: 0.35

2 Top Artificial Intelligence Stocks to Buy Right Now

SOUNVRTNVDA
Artificial IntelligenceTechnology & InnovationCorporate EarningsCorporate Guidance & OutlookCompany FundamentalsAnalyst InsightsAnalyst EstimatesInvestor Sentiment & Positioning
2 Top Artificial Intelligence Stocks to Buy Right Now

Despite a recent pullback in AI equities, two names stand out: SoundHound AI and Vertiv. SoundHound is down ~36% over the past month but reported Q3 revenue up 68% year-over-year, raised full-year guidance to $165–$180M (midpoint implying >100% growth), carries a $1.2B revenue backlog, trades at ~30x sales, and has a median analyst 12‑month target of $16 with 7 of 10 analysts recommending buy. Vertiv, a data-center infrastructure supplier, saw Q3 orders rise 60% YoY, revenue increase 29% to $2.7B, exited with a $9.5B backlog (+30% YoY) and a 1.4 book-to-bill, and trades with a PEG of ~0.83 after a ~13% one‑month pullback — positioning both as buy candidates to capture AI software and hardware adoption.

Analysis

Market structure: AI-driven winners are infrastructure and thermal/power specialists (Vertiv VRT, rack/PDU, liquid-cooling suppliers) and GPU/cloud suppliers (NVDA, AMZN, MSFT) as hyperscaler capex ramps — Vertiv’s $9.5bn backlog and 1.4 book-to-bill point to supply-constrained demand and pricing power in 2025–27. Software/edge players like SoundHound (SOUN) benefit from broad TAM but face competitive pressure from hyperscaler-embedded NLP and must justify a 30x sales multiple against execution risk. Cross-asset: accelerating data‑center capex supports copper, industrial cyclicals and energy demand (near‑term upward pressure on commodity prices), while heavier capex expectations can push rates up and flatten credit spreads, hurting long-duration growth names. Risk assessment: Tail risks include abrupt GPU supply normalization or a 100bp rise in rates that defers 2025–26 capex (could reduce VRT orders 15–30%), EU/US AI regulation curbing voice data monetization for SOUN, and customer concentration or cancellable backlog dynamics. Near-term (days–weeks) volatility will track NVDA/cloud earnings; medium term (3–12 months) backlog conversion and margin trends matter; long term (2–5 years) the 37% AI CAGR is plausible but contingent on sustained hyperscaler spend and competitive moat. Hidden deps: SOUN relies on large enterprise pilots converting to multi-year contracts; VRT relies on steady GPU supply and FX stability in EMEA. Trade implications: Favor durable hardware exposure: establish long VRT (2–3% portfolio) targeting 12–18 months and hedge with 12‑month 25% ITM calls (0.5% notional) for convexity. Keep SOUN as a small, event-driven position (0.5–1%): use 3–6 month call spreads sized small and only add on revenue/backlog conversion confirmation; consider a relative-value pair long VRT / short SOUN (1:1 dollar) to capture multiple compression risk. Reduce broad high-duration AI software exposure by 2–4% and redeploy into data‑center industrials if NVDA/cloud guides confirm >20% y/y capex growth. Contrarian angles: Consensus underestimates cancellability and lead‑time in VRT’s backlog — upside exists if backlog converts at >60% annualized, but downside if <40%. Conversely SOUN’s multiple may already price near-perfect execution; a miss in conversion or margin would produce >30% downside quickly. Historical parallel: 2017–19 hyperscaler capex benefited infrastructure vendors for multiple years, but 2022 rate shocks shortened cycles — monitor rate path and NVDA supply as the decisive second-order variables.