Nvidia is now cheaper than Hershey after a $1 trillion wipeout — time to buy the AI dip

Verdict: the Nvidia dip idea is a watch-list setup, not a trade — and the cash-flow numbers say check the brakes

**The verdict: this is a watch-list idea, not a trade yet — and the fundamentals add a caution the headline ignores.** The strongest point for the thesis is that the business keeps compounding: revenue hit $96.2B in the quarter ended July 26, 2026, up 17.9% sequentially, with gross margin holding near 75% and net income of $59.7B, even as Bloomberg's July 8, 2026 reporting noted the stock shed roughly $1 trillion in value. The strongest point against is that the entry has never triggered in 1,236 daily bars over 60 months — NVDA sits at $218.81 with RSI at 44.6 versus the 35 threshold and 32.8% above its range low, so the market has not delivered the capitulation this strategy is designed to buy. There are also two uncomfortable facts the dip-buying framing glosses over: free cash flow fell 56% sequentially to $21.4B and operating cash flow fell 52.2% to $24.1B, and debt-to-equity jumped from roughly 0.04 to 0.14 — a 269.9% one-quarter swing that deserves explanation before treating the dip as cheap. Meanwhile, ownership filings for the period ended June 30, 2026 show 28 holders with net open-market insider selling of roughly $565.4 million — a June 30-period disclosure, not a read on current behavior. What would flip the verdict: an NVDA close at or below $210 with RSI dropping to 35 or lower, combined with 20-day momentum turning positive on both NVDA and SMH. Until then, set alerts and do nothing. **Conviction breakdown:** - **Thesis support — 60:** The contrarian premise (record-scale fundamentals at a de-rated valuation) is real, but the thesis itself names unresolved operational risks — geopolitical fears and next-gen rack-system delays — and the "cheaper than Hershey" framing rests on a single July 8 valuation datapoint. - **Trade readiness — 25:** The rules never triggered in any evaluated window, and the sensitivity evaluation exceeded its time budget and returned no robust setup, so the published thresholds are used as-is without a validated nearby-parameter alternative. - **Risk quality — 45:** The strategy's defined risk controls are sensible, but NVDA carries roughly 44% annualized volatility, an expected maximum drawdown near 60% for the two-name book, and NVDA anchors 17.8% of SMH, so the sector-confirmation leg offers little true diversification. - **Trigger proximity — 20:** Only one of the entry conditions is met (ADX well under its cap); RSI is nearly 10 points from threshold, price is 4.2% above the $210 support tag, and the NVDA 20-day rate-of-change needs to swing about 2.6 points. - **Fundamentals trend — 55:** Record revenue and stable gross margins, offset by the sequential cash-flow drop, net margin compression from roughly 71.5% to 62.0%, and the leverage jump.

Trade now: NVDA is a watch-list setup — two entry conditions are still far away

Do nothing yet. NVDA closed at $218.81, well above the first support level at $210 that the entry rule requires the price to touch, and the 14-day RSI is 44.6 — nearly 10 points above the 35 threshold. Of the entry conditions, only one is currently met: ADX (14) at 1.6, comfortably under the 40 cap. The 20-day rate-of-change is -2.1%, and the rule wants it above 0.5, so the momentum condition is also unmet. SMH's own condition — its 20-day change above 0.5 — is closer at 0.24, but the NVDA-side conditions are the gating factor. What "wait" means concretely: set alerts at an RSI of 35 and a low at $210 on NVDA, and at a 20-day rate-of-change crossing above 0.5 on both NVDA and SMH. The trade arms only when all five conditions line up — oversold RSI, subdued ADX, a tag of $210 support with the close holding above it, and positive 20-day momentum on both tickers. Today, none of the price or momentum legs is in range, and the price sits 7.2% below its recent range high. Risk framing once triggered: the strategy defines a hard stop at a 2.5% loss and a take profit at roughly 5.0%, an effective reward-to-risk of about 2 to 1, with a secondary invalidation if the close breaks the second support zone near $208.86 and a signal exit if RSI rises above 65. Position sizing is fixed-risk at 2.5% of the account per trade, capped at 25% of capital. One factual scope note that affects planning: the rules were evaluated on real daily bars but produced no entries in the evaluated windows, so this is a watch-list setup rather than an active signal. Separately, no robust parameter setup was established — the sensitivity evaluation exceeded its time budget — so the published thresholds are used as-is.

Why the bull case still has support

The core of this idea — that Nvidia's selloff has taken valuation to pre-AI-boom levels while the business keeps compounding — is at least partially supported by the numbers. Per Bloomberg's July 8, 2026 reporting, Nvidia shed roughly $1 trillion in market value, pushing its valuation below candy maker Hershey's even as it remains the world's most valuable company. That combination of peak-scale fundamentals and compressed multiple is exactly the setup a contrarian long thesis needs. The most recent quarterly snapshot (quarter ended January 25, 2026) shows $68.1B in revenue, up 73.2% year over year — a figure sitting in the 86th percentile among 788 Information Technology peers — with a 75.0% gross margin and a 65.0% operating margin in the 99th percentile of 854 peers. The latest fundamental changes reinforce momentum rather than undermine it. For the quarter ended July 26,…

Scores

• Conviction score breakdown: 41
• Thesis support: 60
• Trade readiness: 25
• Risk quality: 45
• Trigger proximity: 20
• Fundamentals trend: 55

Watch items

• NVDA — RSI (14)
• NVDA — Price vs support level 1
• NVDA — ROC (20)
• NVDA — ADX (14)
• SMH — ROC (20)
• NVDA — Insider net open-market activity
• NVDA — Quarterly free cash flow (next XBRL refresh after 2026-07-26)