The AI safety panic is a broad sentiment shock, but the damage is not evenly spread — it is hitting companies whose products AI could replace (Salesforce, ServiceNow fell 4-5% on the GPT-6 launch) far harder than the company building the infrastructure. N

Verdict: the pair is right, the entry is not yet — wait

The verdict is wait — this is a genuinely interesting thesis whose trade has not yet arrived. The strongest point for it is the fundamental asymmetry the idea leans on: Nvidia grew revenue 65.5% year over year to $215.9 billion with a 55.6% net margin, while ServiceNow's latest quarter (ended June 30, 2026) saw operating margin collapse 9.3 points to 4.1%, exactly the divergence the pair needs. The strongest point against is insider posture: filings for the June 30, 2026 reporting period show net open-market selling at all three companies, including roughly $565.4 million at Nvidia — insiders selling while the thesis asks you to buy the long leg. On the trigger side, NVDA closed at $218.29, just $2.28 below the 20-day moving average at $220.57, and the 14-day RSI at 43.87 sits only about one point under the required 45.0 — both arming conditions are close but unmet, and the entry never fired once across the full tested history of 1,235 daily bars. If a confirmed Nvidia-Anthropic IPO investment (per the idea's citation of Reuters) lands or NVDA closes above $220.57 with RSI at or above 45.0, the setup arms; a daily close below $210 support would push it further away. Position sizing matters most here: the equal-risk basket runs about 34.4% annualized volatility with an estimated maximum drawdown near 68.7%.

Trade now — NVDA long leg is close but not armed

Nothing is triggered yet. NVDA closed at $218.29, which is $2.28 below the 20-day moving average at $220.57 — the price leg of the entry is 'near' status, not met. The 14-day RSI sits at 43.87, just over one point below the required 45.0, and it must also stay below 50.0, which it does. So both arming conditions are within reach but unconfirmed. Because this is a watch-list setup (the rules were evaluated on real daily bars but opened no entry across the tested windows), the correct action today is to wait, not deploy. Waiting means two things concretely: alert at a daily close above $220.57, and alert when the 14-day RSI reaches 45.0 or above while remaining under 50.0. Only when those line up does the long leg arm. Risk is defined in advance. The position rules call for a hard stop at a 2.5% loss from entry, a take-profit at roughly 5.0% — an effective reward-to-risk of about 2 to 1 — plus a time-based exit after 45 bars and a structural stop if price breaks back under the second-ranked support. Price breaking below the nearest support at $210 would push the setup further away rather than invalidate an active trade, since none is open. On the parameter question: the research author requested bounded optimization to relax the compiled entry conditions, but the sensitivity evaluation ran out of its time budget, so no robust alternative setup was established — the live thresholds above are the ones to watch.

The Infrastructure Is Earning Like It, and the Panic Is Not Spread Evenly

The core of this idea — that AI disruption falls hardest on application software rather than infrastructure — is visible in the numbers. Per the Yahoo Finance piece, the GPT-6 launch knocked 4%–5% off Salesforce and ServiceNow in a single session, while the fundamental gap between the three names is stark: Nvidia grew revenue 65.5% year over year to $215.9B with a 55.6% net margin and a 76.3% return on equity, versus Salesforce at 9.6% growth and ServiceNow at 20.9% growth. Those are not companies facing the same demand curve, and the thesis argues exactly that asymmetry. The bull case also rests on business economics that hold up in a bear tape. Nvidia converted $96.7B of free cash flow in its latest fiscal year — the 99.6th percentile of its sector — with a 3.9 current ratio and $10.6B of cash, so the company on the long side of the pair does not need market sentiment to fund itself. ServiceNow still posted 98.9th-percentile free cash flow of $4.6B and an 81st-percentile operating margin for its most recent full year, which matters for the short leg only insofar as the thesis expects multiple compression rather than fundamental collapse. The catalyst side of the argument is the reported Nvidia–Anthropic angle. Per Reuters, Nvidia is in talks to invest in Anthropic's mega IPO — the idea reads that as keeping the infrastructure leader on the inside of every major AI player, so competition among AI labs still flows through Nvidia's silicon. In a panic where the long AI trade is being repriced, holding the company that profits from that competition accelerating is a defensible hedge structure. One scope note the reader should carry: the rule set was evaluated on real daily bars but did not open a single entry over the last 60 months (1,235 bars evaluated, and also across 24- and 12-month windows). This is a watch-list setup, not an active signal — the…

Scores

• Conviction score breakdown: 55
• Thesis support: 70
• Trade readiness: 35
• Risk quality: 45
• Trigger proximity: 55
• Fundamentals trend: 68

Watch items

• NVDA — Price vs 20-day EMA
• NVDA — RSI (14)
• NVDA — Nearest support level
• NVDA — Insider net open-market activity (period ended 2026-06-30)
• NVDA — Anthropic IPO investment confirmation
• NOW — Operating margin (Q2 2026, quarter over quarter)