3 Beaten-Down Software Stocks: 2 to Avoid and 1 to Buy

ServiceNow reported Q4 2025 subscription revenue up 21% to $3.47B and cRPO up 25% to $12.85B but trades at ~63x P/E (forward ~26x), pricing near-perfection. Salesforce grew FY2026 revenue 10% to $41.5B, with $2.9B in Q4 recurring revenue from AI products, trading at ~24x P/E. Adobe posted Q1 FY2026 revenue +12% to $6.4B, trailing-12-month free cash flow of $10.3B versus a ~$98B market cap, trades at ~14x P/E and is down ~31% YTD; author favors Adobe given valuation but flags generative-AI disruption risk.

Analysis

The market is increasingly pricing a binary outcome for incumbents: either they monetize AI as an augmentation layer that deepens enterprise lock-in, or they face a multi-year re-pricing as commoditized models erode pricing power. That creates asymmetric outcomes where misses on execution or timing produce sharp multiple compression, while successful integration of proprietary data + models can sustainably widen gross margins.

Second-order winners include infrastructure and tooling that underpin enterprise AI deployments — GPU vendors, cloud IaaS, MLOps providers, and data-integration platforms — because most large enterprises will prefer managed, audited stacks over one-off point solutions. Conversely, standalone point-product disruptors that rely on model output without substantial data or workflow embedding face user churn risk and rapid margin decline, pressuring their channel partners and nascent marketplace ecosystems.

Key catalysts to watch on multiple horizons are product rollouts and pricing experiments (3–12 months), enterprise pilot-to-deploy conversions and contract re-pricing (6–18 months), and regulatory/standards developments around IP and model provenance (12–36 months). A reversal in the current sentiment will likely be triggered more by evidence of durable revenue per customer (not raw ARR growth) and successful new pricing mechanics than by headline AI announcements alone.

For portfolio construction this implies favoring optionality and defined-risk structures: overweight assets that control unique customer data or content libraries, hedge with short exposure to names whose valuations require near-perfect execution, and keep a small tactical sleeve to capture a renewed AI-capex rotation into infrastructure names.

Key Decisions for Investors

• Long ADBE defined-risk call spread (12-month): buy ADBE 12-month ATM call and sell a higher strike to fund ~60% of premium. Position size 3–5% of equity; target 2x premium if shares re-rate, max loss = premium paid. Rationale: buy asymmetric upside to re-rating while limiting downside to AI-disruption noise.
• Pair trade — long ADBE / short NOW (6–12 months): weight 1.0 ADBE : 0.6 NOW to be roughly beta-balanced. Objective: capture relative re-rating; take profits if ADBE outperforms NOW by 20% or cut if the pair moves against by 15% (reassess on major product/capex headlines).
• Tactical NVDA call spread (3–6 months) as an AI tail-hedge: buy near-ATM call, sell nearer-term higher strike to reduce cost. Allocate small notional (1–2%) to capture upside should risk-on flows rotate back into AI infrastructure; defined downside = premium paid.
• Contrarian speculative: buy INTC long-dated call (18–24 months) sized 0.5–1% as optionality on a potential Intel share gain in AI servers. This is a low-probability, high-upside ticket — treat as lottery allocation and trim if clear share wins/losses are announced.