When: Two of three confirmations: RSI<30, lower-band reclaim, or close >$350 on volume

Then: Add to position

When: $369 (200 SMA) reclaim with sustained volume >25M shares

Then: Add to position (Tier 3)

When: Price drops to $325-$330

Then: Speculative add trigger for sidelined capital only (not operative entry from current levels)

HOLD - Existing holders trim into strength at $360-$365 and protect gains; sidelined capital waits for confirmed reversal signals including oversold RSI print, lower-band reclaim, or successful reclaim of 200 SMA before initiating new positions.

• PEG of 0.36 and forward P/E of 17.78x appear attractive but market is rationally discounting skepticism about earnings durability from hyperscaler concentration
• AI revenue based on purchase orders has tail risk—20% reduction in TPU generation 7 timing could cause 3-5% revenue hit overnight
• TTM P/E of 43.86x is mathematically sensitive to long-duration cash flow assumptions with 10-year at 5%
• Current price $344.72 after 4% drop on elevated volume of 31.9 million shares with distribution pattern—two distribution prints within a month with second breaking 200 SMA
• No confirmed reversal signals: price below 10 EMA, 50 SMA ($382), and 200 SMA ($369); MACD negative and expanding; RSI at 34 in weakening territory; Bollinger middle band at $365 acting as resistance
• Risk management requires bounded maximum drawdown—the aggressive analyst's 7% risk per unit compounds with position sizing and can become 12% move on bad hyperscaler news given beta of 1.46
• Trimming into rallies at $360-$365 harvests gains near overhead resistance and prevents whipsaw scenario
• Macro backdrop with oil above $108 and sector-wide AI rotation affecting AVGO, AMD, Intel, and Marvell supports defensive posture
• Asymmetry for aggressive buyer runs the other way: premature entry with stop at $320 risks losing 7% capital plus opportunity cost; waiting for confirmation costs 10-15% more but bounds downside

→ vs aggressive: The PEG of 0.36 doesn't account for consensus EPS of $19.38 depending on 3-4 hyperscaler customers continuing ASIC deployment at current pace—skepticism has foundation

→ vs aggressive: Concentration is tail risk, not a moat—multi-billion ASIC commitments are generation-specific and can cliff if hyperscaler shifts to internal silicon

→ vs aggressive: Rate headwind being 'fully priced in' is analytical error; stock traded at $495 previously, so either that was wrong or current $344 is wrong—can't assume in firm's favor

→ vs aggressive: 31.9 million shares is second distribution print within a month breaking structural support—this is a pattern, not noise; aggressive analyst's single-day interpretation ignores the sequence of lower highs, lower lows, and negative MACD

→ vs aggressive: Tiered ladder is not 'certainty bias' but proper risk management ensuring maximum drawdown is bounded and recoverable when wrong about the bottom

→ vs aggressive: Trimming into rallies is not 'wrong time to reduce AI infrastructure bet'—it harvests gains when price approaches overhead resistance to ensure some upside is banked if bull case fails

HOLD with adjusted execution parameters: maintain core existing position without trimming into rallies near $360-365; initiate small 1/3 starter position for sidelined capital at current levels around $340-345 with strict 2x ATR stop around $320; add on confirmed close above 10 EMA at $360 with volume; final tier on 200 SMA reclaim at $369 with sustained volume above 25 million shares; trim only on confirmed breakouts above $369 or $382.

• Forward P/E of 17.78x with PEG of 0.36 represents genuine fundamental value that the market is not recognizing; waiting for all four conservative confirmation triggers risks paying $385-400
• Beta of 1.46 creates operational risk; a 7% move can become 10-12% intraday on hyperscaler news, making aggressive sizing dangerous
• Technical indicators (RSI at 34, MACD at negative 8.6) support cautious entry timing but should be secondary signal, not primary constraint
• Trimming 15-25% at $360-365 is premature since this is merely 10 EMA and Bollinger middle, both of which can fail as resistance
• Macro uncertainty (5% yields, 94% no-cuts probability, oil above $108) argues for discipline, not paralysis; position sizing should reflect this uncertainty
• ASIC switching costs create moat for existing generations, but hyperscalers can slow roadmaps, extend cycles, or shift to internal silicon—concentration is both moat and vulnerability
• The 16% probability threshold for all four conditions is back-of-envelope estimate, not hard constraint; fundamental case is lopsidedly positive
• 2x ATR stop around $320 aligns with 1% portfolio risk rule per trade
• Require two of three confirmation signals rather than all four conservative triggers for Tier 2 entry
• Cash at 4.5% preserves capital during drawdowns; cannot compound if stopped out at bottom

→ vs aggressive: Fundamentals-forward analysis becomes operationally dangerous when beta is 1.46; treating technical breakdowns as embarrassment rather than signal risks a 30% drawdown becoming 50% when sizing up into breakdown

→ vs conservative: 16% probability threshold is back-of-envelope estimate dressed as math; waiting for textbook oversold confirmation on a structurally inflecting business means paying $400 for what could have been had at $340; Tier 1 at $325-330 below current price leaves capital paralyzed