Nvidia has sold off going into Wednesday's earnings, and much of the chip sector has followed it lower — but analysts are framing the weakness as pre-report jitters rather than evidence the AI boom is over. When a market leader drops on anticipation inste

Verdict: the story is intact — but the trade isn't armed yet

The strongest case for this trade is that Nvidia's fundamentals flatly contradict the 'AI unwind' narrative — $215.9B in revenue growing 65.5%, a 74.9% gross margin in the latest quarter, $96.7B of free cash flow (99.5th percentile in its sector), and near-zero leverage — so a pre-earnings dip on anticipation rather than deterioration is exactly the setup where confirmation sparks a relief rally. The strongest case against it is that Wednesday is a binary event the rules can't shield you from: a 2.4% stop is thin armor for a stock whose post-earnings gaps routinely exceed that distance, exits were modeled on daily bars (so fills may be worse than shown), and the 18-trade backtest sample is modest. The verdict is **wait** — one entry condition met, two near-misses (NVDA is $1.51 below $210 and $5.54 below the $214 mid-band), and the OBV cross unconfirmed. The 60-month backtest (61.1% win rate, 118% return, 13.1% max drawdown) earns respect but not pre-emption of the confirmation discipline. Note the cross-stock warning: the three-leg basket is economically one semiconductor bet, so size it as one position. **Conviction breakdown:** Thesis support 78 — fundamentals and news framing (per the IBD piece) back the 'jitters, not unwind' read. Trade readiness 35 — entry conditions not armed across all three names. Risk quality 45 — tight 2:1 frame but daily-bar fill fidelity and gap risk cut against it. Backtest evidence 62 — solid but small sample; the 24-month regime improved to 64.7% wins with an 11.7% drawdown. Fundamentals trend 90 — margins, cash flow, and balance sheet are top-percentile across the board.

Trade now

**Do not buy today.** NVDA last closed at $208.5, and the strategy's primary entry is a four-condition long trigger that is not yet armed. Here is the live scorecard: the 21-day EMA ($214.8) is above the 50-day SMA ($207.8) — met. But price must be above the 20-day Bollinger middle band ($214) — NVDA is $5.54 short, or about 2.7%. On-balance volume must cross above price — no live reading, so treat it as unconfirmed until the cross prints. Price must also break above nearest resistance at $210 — NVDA is $1.51 below. In short, one condition met, two near-misses, one unknown. The pre-earnings dip thesis argues you wait for confirmation, and the tape agrees: NVDA's RSI (14) sits at 33.6, SMH at 32.4, SOXX at 33.9 — oversold but nowhere near the momentum-confirmation the entry rules demand. **If filled, the risk frame is tight.** The hard stop is a 2.4% loss (about $5.10 from a $208.5 equivalent entry), and the paired take-profit is 4.9% (about $10.20) — a roughly 2:1 reward-to-risk that matches the fixed 2.45% risk sizing per trade. The strategy also exits if price closes below the 21-day EMA, which is already a live condition today ($208.5 vs. $214.8), or after 45 trading bars. Over the 60-month backtest this rule set produced 18 trades with a 61% win rate and a 118% return, but with a 13.1% peak drawdown — so size accordingly, no more than the 25% maximum position per name. **What "wait" means concretely:** set alerts at NVDA $210 (resistance break), $214 (Bollinger mid-band reclaim), and an OBV-over-price crossover on the daily. A strong earnings print Wednesday could clear all three within a session or two — that is exactly the confirmation the thesis wants. A miss instead pushes NVDA toward support near $200, and the ETFs' weaker trend (SMH and SOXX both have their 21-day EMA *below* the 50-day SMA, 18.6 and 22.9 points short respectively) means their entries are even further away — closer to three-plus percent of repair work. Parameter-sensitivity screening ran out of budget and produced no recommended alternative setup, so the published rules stand as-is.

The dip-buy thesis has a deep fundamental cushion

The core of this idea is that Nvidia's pre-earnings selloff is anticipation, not deterioration — and the fundamentals back that framing. The latest fiscal year (ended January 2026) shows revenue of $215.9B, up 65.5% year over year, with the most recent quarter (ended April 2026) still printing 74.9% gross margin. A company whose margins are expanding, not compressing, is an odd candidate for the 'AI trade is unwinding' narrative. The cash machine behind the thesis is extraordinary even by mega-cap standards. Free cash flow was $96.7B for the fiscal year — in the 99.5th percentile of the Information Technology sector — and net margin sits at 55.6%. Return on equity of 76.3% (92nd percentile among peers) and operating margin of 60.4% (99th percentile) confirm pricing power that has so far survived every round of 'competition is coming' skepticism. The balance sheet adds safety: a current ratio of 3.9 and debt-to-equity of just 0.047, down from 0.44 in fiscal 2023, mean leverage…

Scores

• Conviction score breakdown: 62
• Thesis support: 78
• Trade readiness: 35
• Risk quality: 45
• Backtest evidence: 62
• Fundamentals trend: 90

Watch items

• NVDA — Close vs nearest resistance
• NVDA — Close vs Bollinger (20) middle band
• NVDA — RSI (14)
• NVDA — Close vs second support shelf
• SMH — Close vs Bollinger (20) middle band
• SMH — Close vs nearest support
• SOXX — Close vs Bollinger (20) middle band
• SOXX — Close vs nearest support
• NVDA — Price above Bollinger (20)
• NVDA — EMA (21) above SMA (50)