## Session overview Salesforce exploded roughly 22.6% higher after a strong second-quarter earnings beat and an expanded AI partnership with Anthropic, leading a software rally that lifted ADBE, AVGO, and INTC. Nvidia jumped about 8.7% on a blowout quarte

Verdict: a disciplined CRM watch-list, not a trade yet

The idea is a disciplined post-surge pullback entry on Salesforce, and the fundamental backdrop behind the surge is real: CRM's latest filed quarter (period ended 2026-07-31) showed revenue up 1.9% sequentially to $11.3B and net income up 67.4% to $3.5B, while fiscal-year revenue grew 36.9% to $41.5B with $14.4B of free cash flow. The strongest argument against is that the trade-ready evidence is thin — the entry rules never triggered across 1,237 daily bars over 60 months, and the newest quarter also shows sequential margin compression (net margin down about 10% to 25.9%) plus roughly $1.17M of net insider open-market selling in the June 2026 filings. The verdict flips if CRM actually delivers the setup: a daily close at or below the ~$255.13 EMA with RSI below 45, ideally confirmed by the parameter-sensitivity work establishing a validated rule neighborhood. Until both conditions are live, treat this as a watch-list item with mechanical risk once entered — a roughly 2.4% stop against a 4.8% target.

Trade now: CRM is pulling back toward the buy zone — but the trigger has not fired

This is a watch-list setup on CRM, not an active trade. The strategy wants to buy a post-surge pullback: a daily close at or below the 20-period EMA **with** RSI (14) below 45. Neither condition is live right now. CRM closed at **$257.22**, which is **$2.09 above** the EMA at **$255.13** — the price condition is close but not met. RSI (14) sits at **54.0**, needing to fall below **45** — a real gap, not a technicality. The reversal conditions (RSI turning back up through price, and price reclaiming the EMA) can only be evaluated after the pullback conditions are in place. What "wait" means concretely: do nothing until a daily close puts CRM at or below roughly **$255** *and* RSI (14) drops under **45**. Only after both pullback conditions hold does the crossover side of the entry come into play. Given the 22.6% post-earnings surge the idea describes, a normal digestion phase could deliver these readings within days to a couple of weeks — but chasing now would mean buying above the strategy's zone with no edge defined. Risk is fully mechanical once triggered. A position entered under these rules carries a hard stop at **−2.4%** from entry and a take-profit at **+4.8%**, an effective reward:risk of about **2:1**, sized at fixed risk so a stopped trade loses roughly 2.4% of the account. An RSI (14) reading above **70** or a 60-bar holding period would close the trade from the signal side. Do not pre-position: entering above the EMA means your stop would sit inside the zone the strategy intends to buy, which defeats the whole setup.

A Demand Shock That Makes the Pullback-Worth-Buying Thesis Credible

The session's thesis is that AI demand is repricing enterprise software and semis, and the reported numbers back that up rather than just the headlines. Salesforce's latest fiscal year (ended January 31, 2026) shows revenue of $41.5B, up 36.9% year over year, with free cash flow of $14.4B — near the top of its IT-sector peer group at the 99th percentile — and net income of $7.5B on a 18.0% net margin. The just-reported quarter extends that momentum: revenue rose 1.9% sequentially to $11.3B with net income jumping to $3.5B, and per the Reuters report Salesforce raised annual revenue forecasts while expanding its AI partnership with Anthropic. The idea is not resting on one print. The cluster story holds up in the filings, too. Nvidia's latest quarter (ended April 26, 2026) posted revenue of $81.6B, up 19.8% sequentially, with net income of $58.3B on a 71.5% net margin and free cash flow of $48.6B — and management publicly projected the AI-fueled sales surge extends into 2028, per the Yahoo Finance coverage. Broadcom grew revenue 14.9% sequentially to $22.2B with net income of $9.3B and free cash flow of $10.3B. Two of the largest AI beneficiaries confirming accelerating demand in the same week is the kind of fundamental corroboration a demand-driven thesis needs. Even the laggards carry optionality for a rotation thesis: Intel swung to positive operating cash flow of $7.0B in the June 2026 quarter from $1.1B the prior quarter, with revenue up 18.8% to $16.1B, so the software rally lifting ADBE and INTC has some fundamental footing beyond sentiment. On the strategy side, the setup is a disciplined post-earnings entry: wait for CRM to pull back to its 20-day moving average with RSI below 45, then enter long on a momentum reclaim. That structure is sensible given the magnitude of the move — a 22.6% single-day jump (per CNBC, on track for its second-best day ever) is exactly the kind of event that historically produces mean-reversion pullbacks worth buying. Importantly, the entry conditions have not occurred in the evaluated window — across 1,237 daily bars over 60 months the rules never triggered — so this is a watch-list setup rather than an active signal; the takeaway is that the setup is rare and specific, not that the rules are untrustworthy. The research author requested bounded expanded optimization of the trigger conditions while preserving the thesis, direction, exits, and risk rules, with an untouched final holdout. Finally, the mechanical exits keep the risk bounded: a stop around 2.4% of position value, a take-profit near 4.8%, an RSI-overbought exit above 70, and a maximum 25% position sizing under a fixed-risk method. If the AI demand story stalls, the plan exits quickly rather than riding a broken narrative.

placeholder removed

Start with the ownership tape. In the Form 13F batch for the period ending June 30, 2026, insider filings across this cluster show net open-market selling — at Salesforce, nine holders reported net open-market sales of roughly $1.2M; at Nvidia, 23 holders sold a net $375.3M; at Broadcom, $283.3M net sales. That is not a fatal signal, but the people closest to these businesses were reducing exposure into the very quarter the AI rally accelerated. Notably, Boston Scientific is the one cluster name with net insider buying ($10.0M) — and it is the one that fell 3.1% on the session. The newest Salesforce quarter also shows margin pressure the headlines gloss over. Net margin fell to 25.9% from 29.5% the prior quarter (down about 10%), operating margin slipped to 33.8% from 37.8%, free cash flow dropped 27.9% sequentially to $2.1B, and debt-to-equity fell 11.4% to 0.42. Net income fell 9.4% quarter over quarter even as revenue grew 3.4%. A 22.6% one-day repricing on a quarter with contracting margins and softer cash flow is a demanding valuation setup, and sales-stock rallies of this size routinely give back a large chunk in the weeks that follow — expect double-digit percentage swings in a bad month for a stock that just moved 22.6% in a day. The strategy's own evidence profile is thin in a specific way: the entry rules have not triggered at all in the evaluated window — zero trades across 60 months (1,237 daily bars), 24 months, and 12 months of daily CRM bars. The setup requires four simultaneous conditions (close at or below the 20-day EMA, RSI below 45, an RSI cross above price, and a close crossing above the EMA), and that combination has simply never lined up historically. Additionally, the parameter-sensitivity evaluation exceeded its time budget, so no robust nearby-parameter setup was established — you would be deploying rules with no validated neighborhood. Note also the loss legs of the cluster: iQIYI posted a net loss of $29.5M for 2025 with free cash flow collapsing 99.5% to $1.4M, and Xpeng still ran a $1.1B annual loss despite 87.7% revenue growth — if AI enthusiasm cracks, high-growth losers like these show how fast sentiment reverses. Management and governance questions add friction. Nvidia's insider selling of roughly $375M is large in absolute terms; Intel's CEO Lip-Bu Tan and CFO David Zinsner each received roughly $3.8M and $4.2M in fiscal 2025 compensation while Intel posted a $11.0B quarterly net loss and a $11.0B net-margin burn of -68.4% in the June 2026 quarter — a governance and credibility overhang for the cluster's weakest fundamental story. None of…

Scores

• Conviction score breakdown: 59
• Thesis support: 72
• Trade readiness: 48
• Risk quality: 65
• Trigger proximity: 40
• Fundamentals trend: 70

Watch items

• CRM — Daily close vs EMA (20)
• CRM — RSI (14), daily
• CRM — RSI (14) crossover of price
• CRM — Price reclaim of EMA (20)
• CRM — Position-level stop and target (post-entry)
• CRM — Insider open-market activity (June 2026 filing)