Software selloff giving you deja vu? We’ve been here before, says Deutsche Bank, when the dotcom bubble burst

Equity markets are seeing a rotation away from IT and SaaS after Anthropic launched Claude Cowork plug‑ins and following Amazon’s earnings reaction tied to a reported $200 billion AI spend; the software component of the S&P is down nearly 30% from its October peak while the S&P 500 remains only about 2.6% below last month’s record. Analysts highlight that flows have moved into energy, materials and consumer staples, and argue the current AI-driven repricing differs from the dot‑com era because leading firms generate significant free cash flow and engage in buybacks and dividends.

Analysis

Market structure: The near-term rotation is reallocating index leadership from high-multiple SaaS/software (software component ~30% off its Oct peak) into cyclicals — energy, materials and consumer staples — which have taken up the liquidity baton and kept the S&P only ~2.6% below its record. Winners are AI infrastructure owners (cloud, chips, data centers) and commodity-linked sectors that benefit from re-leveraging; losers are high-growth, low-margin SaaS firms whose recurring revenue is now re-priced for AI obsolescence risk.

Risk assessment: Tail risks include an AI-capex pullback (a shock similar to Amazon’s $200bn AI spend reversing), regulatory constraints on model deployment, or a major model failure causing mass contract repricing; any of these could trigger >15% downside in concentrated tech names over months. Immediate (days) volatility will be earnings-driven; short-term (weeks–months) sees continued rotation and sector dispersion; long-term (quarters–years) outcome depends on realized ROI from AI capex and consolidation into winners.

Trade implications: Tactical plays should favor infra and commodity exposure while using options to limit drawdown: favor 3–9 month exposure to cloud/chip leaders and cyclicals, hedge with downside protection on software ETFs/large-cap tech. Relative-value: long AI-infra (NVDA/AMZN/MSFT) vs short high-multiple SaaS (IGV or SNOW) to capture reallocation; use 3–6 month option structures to manage timing.

Contrarian angles: Consensus underestimates survivorship among profitable, sticky enterprise software — firms with >40% gross margins and multi-year contracts will retain pricing power and consolidate market share. The sell-off may be overdone for cash-flow-positive SaaS trading <25x EBITDA; conversely, capex-heavy players who meaningfully miss ROI (threshold: >20% QoQ margin erosion) will re-price dramatically, creating high-conviction short opportunities.

Key Decisions for Investors

• Establish a 2–3% portfolio long position split 60/40 between XLE (energy ETF) and XLP (consumer staples ETF) for 3–6 months to capture rotation; set a tactical profit target of +10–15% and a stop-loss at -7%.
• Reduce direct SaaS exposure by 40–60% or implement a 1–2% portfolio short via IGV 3–6 month put spreads (buy 1x 15% OTM put, sell 1x 30% OTM) to limit capital while capturing continued de-rating risk.
• Enter a 1% portfolio tactical 3-month put spread on AMZN (buy 1x 5–10% OTM put, sell 1x 20–25% OTM) to hedge near-term downside from AI capex disappointment; if AMZN falls >20% from current levels, begin incremental long accumulation up to 2% for multiyear AI exposure.
• Establish a pair trade: 1% long NVDA (or 6–12 month 10–20% OTM call spread) vs 1% short IGV (cash or options) to express AI infra upside and SaaS downside; rebalance monthly and close if spread narrows by 50% in 60 days.
• Set hard event monitors: flag quarterly capex guidance (AMZN/MSFT/NVDA) and Anthropic/OpenAI enterprise plug-in adoption metrics — if incremental AI capex guidance rises by >$50bn annualized or adoption signals 20%+ ARR uplift for infra vendors, increase infra/commodity allocations by +2–3% within 30–90 days.