Article
Updating the thesis after the pullback — conviction unchanged, position re-entered
54d · 3 min read
NVDA pulled back 18% from its high over six weeks. In that period:
The pullback was multiple compression in response to broader market risk-off, not fundamental deterioration. When the stock is down 18% and the thesis is intact, that is a buying opportunity. I re-entered my full position at $118.40.
This piece documents the updated thesis and why I have not lowered the target.
Original thesis (published at $134):
Post-pullback update: All four pillars are intact. If anything, the channel check data from the consolidation period is more constructive than at the time of the original thesis:
PillarOriginal signalCurrent signalDeltaBlackwell demandConstrainedStill constrainedUnchangedGM expansionH200 ASP +35%B200 data confirming +40–45%More positiveSovereign AIEmergingUAE/Saudi orders confirmed in earningsConfirmedTAM$150B data center 2025Raised to $170B+ by consensusUpgradedBear argument 1: "AMD MI300X is taking share" MI300X is gaining in inference workloads. This is true. But training — where NVDA has 85%+ share — is where the revenue is. Inference is growing but is a fraction of training revenue in dollar terms. The competitive threat is real but overstated relative to the revenue line.
This content is for informational purposes only and should not be considered financial advice.
Be one of the first credible voices on FinancialPress.
Discussion is gated until launch. Sign up now to contribute analysis, build your Reputation Score, and earn visibility from day one.
Bear argument 2: "China export controls are an existential risk" H20 restrictions reduced China TAM by ~8%. This has already happened. The question is whether further restrictions arrive. My base case: no material escalation in the next 12 months given the diplomatic environment. The risk is real; the probability of the extreme scenario is low.
Bear argument 3: "The valuation is too high at 35x" 35x forward earnings on a company growing revenue 200%+ is not obviously expensive. The relevant comparison is not absolute P/E but P/E relative to growth (PEG ratio). At 35x with 200% revenue growth, the PEG is 0.17 — one of the cheapest in the S&P 500 on that metric.
The risk/reward from re-entry is more attractive than the original entry. Same target, 17% lower cost basis.
None of these have occurred. Until they do, the thesis stands.
Author holds a full long position in NVDA, re-entered at $118.40. Also holds BTC and META at time of publication.
High-conviction positions carry concentrated risk. This analysis reflects the author's personal view and is not financial advice. The 18% pullback example illustrates that even thesis-intact positions can experience significant drawdowns.
Thesis review methodology: systematic comparison of original thesis pillars against updated observable data. Bear case rebuttal based on publicly available competitive data. Risk/reward recalculated from new entry point.
Sign in to join the conversation.
No comments yet - be the first to share your take.
