Meta Platforms (META)

The panel's take

Verdict: BUY · Conviction: MEDIUM · Last price: USD 560.66 (as of 2026-06-29)

This is the panel's most one-sided call to the upside — five of six frameworks are bullish, including the resident contrarian. Meta's core advertising engine is accelerating, not breaking: Q1 revenue grew 33% to $56.3B (the fastest since 2021), ad pricing rose 12%, operating margin held near 41%, ROE is ~33%, and net income jumped ~61%. Yet the stock has fallen ~32% from its ~$820 peak — almost entirely on AI-capex fear ($125–145B guided for 2026) and a persistent ~$4B/quarter Reality Labs loss — compressing the multiple to ~17x forward earnings, below Meta's own ~26x average and below the broad market. Independent fair-value estimates from the value, growth, valuation, and even forensic lenses land between roughly $680 and $720, with the analyst consensus above $800 — 20–48% above today's price. The case rests on the de-rating having overshot a still-compounding cash machine. The one genuinely non-confirming signal is the tape: the weekly chart is in a downtrend, which is why conviction is MEDIUM rather than higher.

Key levels

Key resistance starts at ~$590 (near-term swing), then ~$640 (the 50-week trend / a ~38% retrace of the decline), up to the ~$700 level where the long-term trend signal flipped negative. Key support sits at ~$545 (the recent swing low), then a ~$500 psychological zone near the 200-week trend, with a deeper level near $460. The analyst consensus target spans roughly USD 622 to 1,015, averaging above USD 800 — well above the current price, an unusually wide gap that reflects how completely the Street reads the selloff as capex-fear rather than a broken thesis.

What legendary investors think

We ran Meta Platforms past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Is Meta a wide-moat cash machine handed back at 17x forward on an overblown capex scare — or a business whose $125–145B AI bet and bottomless Reality Labs burn will keep free cash flow suppressed long enough to justify the market's skepticism?

The numbers

Metric	Value
Price / Market cap	USD 560.66 / ~1.4T
P/E (TTM / fwd)	~20x / ~17x (vs ~26x 12-mo avg)
ROE	~33%
Operating margin	~41%
Dividend yield	~0.4%
Debt / equity	~0.25x (net cash ~$22.5B)
Free cash flow	~USD 12.4B (Q1 2026; compressed by AI capex approaching operating cash flow)

Figures as of Q1 2026 / June 2026; sourced from Meta IR/SEC filings, StockAnalysis, GuruFocus, MacroTrends, MarketBeat. GAAP Q1 EPS was lifted by a one-time $8B tax benefit; adjusted EPS ($7.31) is the cleaner figure. Free cash flow is positive but compressed as 2026 capex (~$125–145B, up sharply YoY) approaches operating cash flow — the swing variable behind both the bull and bear cases.

The bottom line

Meta is the panel's clearest value-versus-tape standoff, and the value side carries the weight. The bull case is grounded and, this quarter, accelerating: a wide-moat advertising machine growing 33% with ~41% operating margins and ~33% ROE, AI-driven targeting visibly lifting ad pricing, a net-cash balance sheet, and a multiple compressed to ~17x forward — below the company's own history and the market — because investors are in "prove-it" mode on AI spending rather than because the business is faltering. Five independent frameworks, including the forensic skeptic who read the footnotes, place fair value between roughly $680 and $720, with the Street above $800 — 20–48% above the price — and they read the 32% drawdown as a capex scare that overshot. The caution is real but narrower than usual: the $125–145B capex and the ~$83.5B cumulative Reality Labs loss are compressing free cash flow with the depreciation drag still building, and the weekly tape is in a confirmed downtrend that the macro and technical lenses refuse to fight. What would convert the remaining doubters is evidence the capex converts — ad-revenue acceleration validating the AI spend, Reality Labs losses flattening, and a reclaim of the long-term trend. What would validate the skeptics is capex creeping higher with no margin payoff into 2027. On the weight of an accelerating business, a below-average multiple, and a fair-value gap five frameworks agree on, the panel leans BUY — with the open capex-ROI question and an unconfirmed tape the reasons conviction stays measured rather than high.

The panel's take

Verdict: BALANCED · Conviction: MEDIUM · Last price: USD 668.00 (as of 2026-07-11)

Meta lands on BALANCED after a violent week: the stock just ripped ~15% — its best week since early 2024 — as the July 1 launch of "Meta Compute" reframed the company's enormous AI spend from pure cost into a potential cloud-revenue line, and the panel is genuinely split on whether the pop is the start of something or a reversion into resistance. The bull case is strong: Q1 2026 revenue grew 33% to $56.3B, operating margin held near 41%, ad impressions rose 19% with price-per-ad up 12%, the balance sheet is a fortress ($81B cash, debt/equity ~0.36), and the Street is near-unanimous — no sell ratings and an average target around $838, well above the price. But the panel's valuation lenses push back hard on chasing the move: a story-to-numbers DCF lands fair value around $560-620 (i.e. slightly rich after the run), reported EPS was flattered by an ~$8B one-time tax benefit, and the real number to watch is capex — raised to $125-145B for 2026, roughly double 2025's ~$72B, with Q1 free cash flow already compressed to $12.4B. So three lenses are bullish on the AI-monetization turn, two are neutral, and the forensic lens is outright bearish on the cash-flow math. With the stock having run straight into technical resistance ($696-725) just ahead of the July 29 Q2 print, the synthesis is BALANCED.

Key levels

Key resistance begins at ~$696 (the falling weekly Supertrend flip level), then a ~$725 prior-high shelf, up to the ~$765-770 swing high before the final push to the all-time-high zone. Key support sits at ~$600 (round-number congestion), then ~$577 (this week's low and nearest historical accumulation zone), with a deeper structural level near ~$520 (the 2025 base before the rally). The analyst consensus target spans roughly USD 622 to 1015, averaging near USD 838 — a wide band that captures the debate: the Street sees meaningful upside, while the panel's own valuation work sits below the current price after the pop.

What legendary investors think

We ran Meta Platforms past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Is Meta's ~15% week the market correctly pricing an AI-monetization inflection in one of the best businesses on earth — or a momentum pop into resistance that has pushed the stock past its own intrinsic value while $125-145B of capex and a compressing free cash flow remain unproven?

The numbers

Metric	Value
Price / Market cap	USD 668.00 / ~1.7T
P/E (TTM / fwd)	~22-24x / ~20x
Q1 2026 revenue / EPS	USD 56.3B (+33% YoY) / USD 10.44 (incl. ~$8B one-time tax benefit)
ROE / operating margin	~33% / ~41%
Dividend yield	~0.35%
Debt / equity	~0.36 (cash ~$81B)
Free cash flow	~$46-48B TTM (Q1 ~$12.4B as capex ramps)

Figures as of Q1 2026 (reported Apr 29, 2026) / July 11, 2026; sourced from Meta IR, CNBC, Yahoo Finance, TIKR, Investing.com, MacroTrends, StockAnalysis, GuruFocus. Q1 EPS ($10.44) was inflated by an ~$8.03B one-time tax benefit — revenue growth (+33%) and operating margin (~41%) are the reliable signals. 2026 capex was raised to $125-145B (from ~$72B in FY2025); Reality Labs Q1 loss $4.03B (cumulative >$90B). ROE varies by source (~33% financecharts / ~46% GuruFocus). The "Meta Compute" AI-cloud launch (July 1) drove the 15% weekly rally. Analyst consensus ($838 avg, range ~$622-1015) sits above the price. Next print: July 29, 2026 (Q2 2026).

The bottom line

Meta is the panel's most timing-sensitive BALANCED verdict — a genuinely elite franchise whose ~15% week has split the room on whether the AI-monetization turn is real or already priced. The bull case is strong and current: Q1 revenue up 33% with margins near 41%, ad impressions and pricing both accelerating, a fortress balance sheet, and a Street that is near-unanimous with an average target around $838. The launch of "Meta Compute" is the reason the stock re-rated, reframing the year's most-questioned line item — $125-145B of AI capex — as a potential revenue asset rather than pure cost. But the panel's valuation and forensic lenses are the reason this is not a chase: a disciplined DCF lands fair value around $560-620, roughly 10-16% below where the stock now trades; reported EPS was flattered by an ~$8B one-time tax benefit; free cash flow compressed to ~$12.4B in Q1 as capex nearly doubled; and Reality Labs' $90B+ cumulative burn still has no path to breakeven. What would tip the call decisively bullish is the capex starting to convert — a July 29 print showing Meta Compute revenue traction and margins holding near 41% despite the depreciation wave, with the stock clearing ~$725 on volume. What keeps it BALANCED is that the stock just ran into resistance and past its own intrinsic value on a monetization promise the cash flow hasn't yet confirmed — the same elite business is easier to own once the capex proves it pays.

The panel's take

Verdict: WAIT · Conviction: MEDIUM · Last price: USD 646.01 (as of 2026-07-18) · Changed from BALANCED

What changed since 2026-07-11:

• Price: USD 668.00 → 646.01 (-3.3%) — a second straight rejection at the $696-700 resistance zone pulled the stock back toward support.
• Panel: 3 of 6 → 1 of 6 now Bullish (The GARP Operator and The Macro Opportunist flipped from Bullish to Neutral) — both now want the July 29 Q2 print to show the capex bet converting into cash flow before turning bullish again.
• Trigger: The July 29, 2026 Q2 earnings print — whether Meta Compute revenue shows up and free cash flow recovers from the $12.4B Q1 low.
• Verdict: BALANCED → WAIT

The operating business is not the argument here — 33% revenue growth to $56.3B and a 41% operating margin are numbers nearly every lens on the panel treats as elite. The four neutral lenses (value, GARP, valuation, macro) share one objection: they want the July 29 print to show the $125-145B 2026 capex guide actually converting into revenue before calling it cheap, given Q1 free cash flow compressed to just $12.4B and $8B of the headline $10.44 EPS came from a one-time tax benefit. The Disruptive-Innovation Seeker stays bullish on the multi-year "Meta Compute" platform thesis regardless of this quarter's noise; the Forensic Skeptic stays bearish, reading the same tax benefit and capex-driven cash burn as deteriorating earnings quality rather than a temporary lull. The dominant objection is "prove it, don't price it yet," not a broken thesis.

Key levels

Key resistance begins at ~$696-700 (the falling weekly Supertrend line, rejected for a second straight test this week), then the ~$725 level a volume-backed close above would need to clear to flip the technical picture, up to the ~$768 prior swing high. Key support sits in the ~$600-630 band — a historical accumulation zone tested repeatedly over recent weeks — with a deeper level near ~$577. The analyst consensus target spans USD 622 to 1015, averaging near USD 838, a band that sits well above both the current price and the panel's own fair-value estimates.

What legendary investors think

We ran Meta Platforms past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Does the July 29 Q2 print show Meta Compute revenue starting to convert the $125-145B capex bet into durable cash flow — or does free cash flow stay compressed while Reality Labs' burn continues, leaving the panel's own $560-620 valuation work as the more reliable signal than the Street's $838 average target?

The numbers

Metric	Value
Price / Market cap	USD 646.01 / ~1.65T
P/E (TTM / fwd)	~21-23x / ~20x
Q1 2026 revenue / EPS	USD 56.3B (+33% YoY) / USD 10.44 (incl. ~$8B one-time tax benefit)
ROE / operating margin	~33% / ~41%
Dividend yield	~0.35%
Debt / equity	~0.36 (cash ~$81B)
Free cash flow	~$46-48B TTM (Q1 ~$12.4B as capex ramps)

Figures as of Q1 2026 (reported Apr 29, 2026) / July 18, 2026; sourced from Meta IR, CNBC, Yahoo Finance, TIKR, Investing.com, MacroTrends, StockAnalysis, GuruFocus. Q1 EPS ($10.44) was inflated by an ~$8B one-time tax benefit — revenue growth (+33%) and operating margin (~41%) are the reliable signals. 2026 capex remains guided at $125-145B (from $72B in FY2025); Reality Labs' cumulative loss exceeds $90B. Analyst consensus ($838 avg, range ~$622-1015) sits above both the current price and the panel's own valuation work. Next print: July 29, 2026 (Q2 2026).

The bottom line

What would tip this decisively bullish is the July 29 print showing Meta Compute revenue traction alongside a volume-backed close above $725; what would tip it bearish is free cash flow staying compressed with no Meta Compute revenue disclosed and Reality Labs' burn continuing unresolved. A break below the $600-630 support band without either signal in hand would put the deeper $577 level in play next. Until the print, the panel's own $560-620 valuation range — sitting well below the Street's $838 average target — is the more conservative reference point to watch against.

The panel's take

Verdict: WAIT · Conviction: MEDIUM · Last price: USD 595.19 (as of 2026-07-25)

What changed since 2026-07-18:

• Price: USD 646.01 → 595.19 (-7.87%) — a seven-session losing streak that broke clean through the $600-630 zone flagged last week as a "historical accumulation zone," now trading roughly 5.5% below the ~$630 40-week average proxy for the first time this cycle.
• Panel: 1 Bullish / 4 Neutral / 1 Bearish → 1 Bullish / 3 Neutral / 2 Bearish. The Macro Opportunist is the one genuine flip, moving Neutral → Bearish on the confirmed support break; The Disruptive-Innovation Seeker stays the lone Bullish holdout as the AI-compute catalyst gets more concrete, not less.
• Catalyst: Same dated event — Q2 2026 earnings, Wednesday July 29 after the close — now four trading days out. Sell-side trimmed targets into the print (JPMorgan 875→800, BofA 900→810, Citizens 825→800, Piper Sandler 880→840) while holding Buy/Overweight ratings, a coordinated "trim into earnings" rather than capitulation.
• New offset: The reported ~$10B, two-year AI-compute leasing arrangement now has a named counterparty, sharpening The Disruptive-Innovation Seeker's thesis even as a new risk data point (capex tracking as high as ~$200B per one sell-side estimate, well above the $125-145B guide) raises the stakes on the same print.
• Verdict: WAIT (unchanged)

Nothing has actually resolved — that's the throughline across every lens this week. The operating business is untouched: ~33% ROE, ~41% operating margin, and revenue growth still running in the mid-20s% are numbers no lens disputes. What moved is the setup around that business — price fell into, and now through, the level the panel spent last week calling a floor, which is real technical deterioration even if the fundamentals haven't budged. Four lenses (moat/quality, GARP, DCF, macro/timing) treat this as the market re-pricing risk correctly into a binary event rather than handing out a discount; the Forensic Skeptic reads the same capex escalation as tightening the vise on free cash flow with no resolution yet; the Disruptive-Innovation Seeker reads the same spend, plus a firming AI-compute deal, as the clearest evidence yet that the platform thesis is real. The dominant read is still "prove it on July 29," now with a slightly worse chart and a slightly more concrete bull catalyst pulling in opposite directions.

Key levels

The $600-630 band that anchored last week's support case is now the first overhead resistance on any bounce attempt — it was breached intraweek and closed below, flipping its role. Above that, the ~$696-725 zone (the falling weekly Supertrend line plus the prior rejection band) stays the next real ceiling, with the ~$768 prior swing high further out. On the downside, the ~$577-581 zone (the prior article's S3, untested until now) becomes the front-line support to watch, with the ~$540-557 June 2026 swing-low band next, and a deeper ~$520-526 multi-quarter low in view if both give way. The analyst consensus target — trimmed this week but still averaging near USD 826 — sits well above both the current price and the panel's own $560-620 fair-value work, a gap that hasn't narrowed.

What legendary investors think

We ran Meta Platforms past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Does the July 29 print show free cash flow stabilizing alongside visible AI-compute revenue and capex discipline — confirming the panel's own $560-620 valuation work as the more reliable signal than the Street's ~$826 average target — or does free cash flow stay compressed while capex risk runs toward the higher end flagged this week, confirming this week's support break as the start of a deeper move toward the $540-557 and $520-526 bands?

The numbers

Metric	Value
Price / Market cap	USD 595.19 / ~1.52T
P/E (TTM / fwd)	~21.65x / ~18.27x
Q2 2026 estimate (reports Jul 29)	Revenue ~USD 60.26B (guide USD 58-61B) / EPS est. USD 7.23
ROE / operating margin	~33% / ~41%
Dividend yield	~0.38%
Debt / equity	~0.36 (cash ~$81B)
Free cash flow	USD 48.25B TTM; Q1 quarterly figure disputed across sources ($12.4B vs. ~$16.2B) — unresolved pending the Jul 29 10-Q

Figures as of July 24-25, 2026; sourced from StockAnalysis.com, GuruFocus, ts2.tech, Motley Fool, Alphastreet, TipRanks/TheFly, 247wallst.com. TTM revenue of USD 214.96B (+26.2% YoY) and ~41% operating margin are unchanged from the prior update and remain the reliable signals. Forward P/E compressed to ~18.27x from ~20-23x as price fell faster than estimates moved. The free-cash-flow figure is deliberately flagged as conflicting rather than resolved — reconcile against the actual 10-Q at the July 29 print. Analyst consensus (~USD 826 average) reflects this week's target cuts (JPMorgan, BofA, Citizens, Piper Sandler all lowered, ratings held) and sits above both the current price and the panel's own USD 560-620 valuation work. Next print: July 29, 2026 (Q2 2026), after market close.

The bottom line

What would tip this decisively bullish is a July 29 print showing free cash flow stabilizing or improving alongside disclosed AI-compute revenue and capex-discipline language — that would validate the Disruptive-Innovation Seeker's read and give the Moat Compounder and Intrinsic-Value Modeler the margin of safety they've been waiting for inside the $560-620 band. What would tip it bearish is free cash flow staying compressed while capex guidance holds or climbs toward the higher estimates now circulating, with Reality Labs' burn still unresolved — that would confirm the Forensic Skeptic and Macro Opportunist's read and put the untested $577-581 support in play, with $540-557 and then $520-526 next if it gives way. Until the print, the now-broken $600-630 zone is the level to watch as resistance, not the floor it was treated as last week.

The panel's take

Verdict: WAIT · Conviction: HIGH · Last price: USD 556.71 (as of 2026-08-02)

What changed since 2026-07-25:

• Price: USD 595.19 → 556.71 (-6.47%)
• Panel: 1 of 6 → 2 of 6 now Bullish — composition also changed: added The Quality Rationalist, The Structural Tailwind Rider; removed The Disruptive-Innovation Seeker, The Moat Compounder
• Trigger: A bullish signal would come from a weekly close above $579.32 (78.6% Fibonacci retracement) or a monthly test of $525.73 with a bullish reversal candle; a bearish signal (less likely) would be a monthly close below $490.93 (weekly EMA200), confirming a Stage 3→4 transition and opening a measured move toward $464-442.
• Verdict: WAIT (unchanged)

Price fell 6.5% from $595 to $557, bringing META closer to the cautious voices' accumulation zones ($525, $490) without reaching them, while the two new panelists — The Quality Rationalist and The Structural Tailwind Rider — split their calls, reinforcing rather than resolving the BUY-versus-WAIT structure. The net effect: conviction rose from MEDIUM to HIGH that waiting is correct, because the price move validated the cautious timing thesis without triggering any entry signal for either camp.

The split is BUY vs WAIT — not BUY vs SELL. The fundamental scout, The GARP Operator, and The Structural Tailwind Rider are all Bullish at current levels. The remaining cautious voices (technical, The Quality Rationalist, The Macro Opportunist, The Forensic Skeptic, The Intrinsic-Value Modeler) all acknowledge the business is exceptional; their objection is timing — they want the support test at $525-490 or a momentum reclaim above $579 before shifting stance. Only The Forensic Skeptic shades into genuine business concern (negative FCF as structural impairment), but even he would turn bullish, not more bearish, at $464-480. This is a classic BUY vs WAIT structure: great business, wrong price/timing. Synthesized to WAIT.

Key levels

Key resistance sits at USD 579.32 (R1), USD 604.67 (R2), and USD 623.73 (R3) above current price, while key support sits at USD 525.73 (S1), USD 520.26 (S2), and USD 490.93 (S3) below it, with the analyst consensus target averaging USD 769, well above both.

What legendary investors think

We ran Meta Platforms past a panel of 6 legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Is Meta's $506 billion AI infrastructure buildout a temporary investment cycle that will generate incremental revenue and expand the moat — creating an attractive setup at 17x forward earnings — or the beginning of a structural margin compression that will keep the multiple depressed for years, vindicating those who wait for $490–525 before shifting stance?

The numbers

Metric	Value
Price / Market cap	USD 556.71 / n/a
P/E (TTM / fwd)	20.98 / 17.28
ROE	29.85%
Operating margin	n/a
Dividend yield	~0.38%
Debt / equity	0.43
Free cash flow	USD 40.98B (TTM)

Figures as of 2026-08-01; sourced from stockanalysis.com (https://stockanalysis.com/stocks/meta/statistics/).

The bottom line

A weekly close above $579.32 — the 78.6% Fibonacci retracement — would signal that momentum has shifted and align with the case the four cautious voices are waiting on. A monthly test of $525.73 with a bullish reversal candle is the alternative setup that aligns fundamental value with technical support. The Q3 2026 earnings report is the next dated catalyst: any upward revision to capex guidance beyond $185B without a corresponding revenue acceleration would harden the bear case, while management articulating a credible ROI timeline or committing to maintain buybacks through the build-out would move several WAITs toward BUY. Watch the Advantage+ run rate: crossing $100B would provide the first hard evidence that AI monetization is outpacing the spend, directly challenging the Forensic Skeptic's core thesis. A monthly close below $490.93 — the weekly EMA200 — confirms a Stage 3-to-4 transition and opens the measured move to $464–442, where even the bears would turn bullish.

The panel's take

Verdict: WAIT · Conviction: MED · Last price: 592.1

What changed since 2026-08-02:

• Price: 556.71 → 592.1 (+6.36%)
• Panel: 2 of 6 → 0 of 6 now Bullish — composition also changed: added The Moat Compounder, The Scuttlebutt Growth Investor; removed The Macro Opportunist, The Structural Tailwind Rider
• Trigger: BUY trigger: weekly close above $606-$629 (weekly EMA20 / monthly Fib 23.6%) with volume confirming the weekly MACD bullish crossover, AND/OR Q3 FCF stabilizes above $5B with Meta Business Agent monetization showing traction. SELL trigger: weekly close below $520.26 (February 2026 swing low), confirming Stage 3 transition toward $442.
• Verdict: WAIT (unchanged)

Since the prior HIGH-conviction WAIT at $556.71 in early August, META rose 6.4% to $592.10, but the Q2 print revealed FCF collapsing to near-zero ($784M vs $8.6B a year ago), intensifying the capex debate. The conviction dropped from HIGH to MEDIUM not because the outlook improved, but because nascent bullish technical signals (weekly MACD crossover) have made the debate more evenly balanced, even as the FCF urgency has grown. The business quality remains unquestioned; the question is purely one of price and timing.

This is a BUY vs WAIT split, not BUY vs SELL. The fundamental-scout says BUY now. The other five voices (technical-scout, The Intrinsic-Value Modeler, The Forensic Skeptic, The Moat Compounder, The Quality Rationalist) either explicitly recommend WAIT or express Neutral-with-bearish-tilt driven overwhelmingly by price/timing concerns. The business quality is near-universally praised (30% ROE, 22% revenue growth, fortress balance sheet, wide moat). The objections are: (1) price at $592 offers no margin of safety above the $520-525 support zone, (2) FCF has collapsed and needs to stabilize, and (3) the technical structure requires a reclaim of $606+ before the trend is investable. The GARP Operator and The Scuttlebutt Growth Investor are Neutral/Neutral-bullish but also cite price/timing hesitations. The Intrinsic-Value Modeler is the lone BUSINESS-level objector (intrinsic value below current price), but even he acknowledges the business quality. The structure is textbook WAIT: great business, wrong entry.

Key levels

What legendary investors think

We ran Meta Platforms past a panel of 6 legendary investors' frameworks.

Where they agree — and where they clash

The question it comes down to: Can Meta's AI capex generate enough monetization to restore free cash flow before the market's patience (already worn thin by a near-zero FCF quarter) exhausts completely, or will Q3 2026 provide the definitive answer one way or the other?

The bottom line

• Trigger to confirm the breakout: A weekly close above $606-629 (weekly EMA20 / monthly Fib 23.6%) with volume confirming the weekly MACD bullish crossover would signal the correction is over and the uptrend has resumed.
• Next catalyst: Q3 2026 earnings: FCF stabilizing above $5B and Meta Business Agent paid monetization showing enterprise traction would directly address the market's core concern about the AI capex return.
• Risk to monitor: A weekly close below $520.26 (the February 2026 swing low) would confirm a Stage 3 transition and open downside toward $442, the structural failure level every voice on the panel is watching.
• Further out: The September Connect Conference, where management is expected to unveil the 'Watermelon' frontier model: benchmark wins or concrete performance claims could shift the AI monetization narrative before Q3 earnings land.