Thursday, June 25, 2026 · Morning

Methodology note: Thursday June 25, 2026 morning briefing generated pre-open (~11:00 UTC / 7:00 AM EDT). Equity snapshot carries forward June 24 close (SPX 7,358.22, Nasdaq 25,476.44, VIX 19.49 from TheStreet/Yahoo Finance). BTC live at $61,285.75 (Fortune/CoinDesk, 10:54 UTC). Futures data from Yahoo Finance at 6:25 AM EDT (ES +0.7%, NQ +2.1%). 10Y yield 4.45% and DXY 101.27 from Trading Economics / Investing.com premarket. VIX premarket ~18.32 from StreetStats.finance. Micron Q3/Q4 data from Motley Fool / StockTitan. Qualcomm Dragonfly announcement from CNBC / Bloomberg. Asia session closes from Yahoo Finance / Bloomberg. Oil prices from Trading Economics. PCE consensus from Morningstar. Q1 GDP second estimate from BEA; third estimate due 8:30 AM EDT today alongside PCE. Russell reconstitution from LSEG / CME Group.

Morning briefing — pre-market open prep.

Verdict — HOLD — Two AI Demand Confirmations in One Night; PCE Is the Final Gate

The overnight tape delivered the AI demand confirmation the market demanded — twice. Micron Q3 FY2026 reported $41.5B in revenue against a $34.66B consensus (+20% beat), with Q4 guidance of $50B ± $1B (vs. ~$42.9B expected, +17% guide beat). GAAP gross margin reached 84.6%, HBM4 has been in production for NVIDIA's Vera Rubin platform since March 2026, and 2026 HBM supply is fully booked. Simultaneously, Qualcomm debuted its Dragonfly data center portfolio — C1000 CPU, AI300 inference accelerator — with Meta as its first major customer and a $15B data center revenue target by 2029. QCOM +12% premarket. Asia reversed the 'Black Tuesday' KOSPI panic in a single session: KOSPI +5.42% to 8,930.30, Nikkei 225 +4.61% to 72,366.34. ES futures +0.7%, NQ futures +2.1% as of 6:25 AM EDT.

Verdict: HOLD at medium conviction. Not BUY — because three regime indicators remain breached pre-open and the PCE binary prints at 8:30 AM ET.

What keeps the verdict at HOLD and not BUY:

1. Regime bars are still active. VIX premarket ~18.32 (above 17 gate; June 24 close 19.49). DXY 101.27 (above 100 gate; June 24 close 101.59). ES futures imply SPX open of ~7,409 — still 51 points below the 7,460 floor. Three of six regime indicators remain breached; two consecutive closes clearing all three are required for a full BUY upgrade.

2. MU told the story at $1,083 premarket. Micron's AH surge reached ~$1,199 (+14.6% from ~$1,082 regular close). By 7:00 AM EDT, MU had retraced to ~$1,083 — giving back all AH gains. The MU $1,200-$1,250 add zone was not available in premarket. The market is explicitly waiting on the PCE print before repricing AI memory multiples.

3. NVDA did not gap higher. NVDA premarket ~$199-201 (roughly flat from $200.81 close). Despite MU's $50B Q4 guide directly confirming Vera Rubin HBM4 demand and Jensen Huang's June 24 Annual Meeting confirmation that the AI ROI question "has been answered," NVDA has not gapped meaningfully higher. This is the market's message: AI demand is confirmed; rate regime must clear before paying up.

Two conditional outcomes define the rest of the session:

• Cold PCE (core below 3.2%): The disinflation path is confirmed. DXY should ease below 100, 30Y pulls back from 4.93%, VIX falls toward 17. SPX gaps above 7,460. HOLD upgrades to BUY same session. Buy MU at $1,200-$1,250, NVDA above $205, add QCOM on any pullback from premarket spike. Oil's WTI decline below $70 is the primary disinflation wildcard that could pull the headline below 4.0%.

• Hot PCE (core above 3.5%): BofA's 3-hike framework is confirmed. 30Y tests 5.00% same session, the June 24 regime upgrade evaporates, and the NQ gap-up reverses hard. Trim MU and NVDA adds within 45 minutes of the print. HOLD reverts to STEP ASIDE.

June 24 Close + June 25 Premarket

Today's Catalyst Sequence — Two Simultaneous Releases at 8:30 AM ET

Everything hinges on the 8:30 AM ET data drop.

Event 1 — May PCE (8:30 AM ET, BEA) Consensus: headline 4.1% YoY (highest since April 2023), core 3.3% YoY. The Cleveland Fed nowcast implied a hot monthly run rate (0.37% MoM core). Oil's 10%+ two-week decline is the dominant downside wildcard — if energy pulls the headline, core may remain sticky at 3.3% even as the headline surprises cold. Three scenarios:

• Headline below 4.0% AND core below 3.2% regime normalizes; HOLD BUY. Buy MU $1,200+, NVDA $205+.
• Headline 4.0-4.2%, core 3.0-3.4% broadly in-line; HOLD maintained. Continue selective adds only on confirmed SPX reclaim of 7,460.
• Core above 3.5% BofA 3-hike thesis confirmed; 30Y tests 5.00% same session; HOLD STEP ASIDE; trim within 45 minutes.

Event 2 — Q1 GDP Third Estimate (8:30 AM ET, BEA) Second estimate: 1.6% annualized (downward revision from 2.0% advance). Third estimate typically small revision. A further downward revision to 1.3-1.5% adds a growth-softening signal that paradoxically helps the disinflation narrative (weakening demand = lower inflation) but adds recession fear if paired with hot PCE (stagflation corner).

The Week Ahead — Sequence Matters

The Russell Reconstitution on June 26 is a structural vol event. FTSE Russell's 2026 June reconstitution is the first semi-annual rebalance, with ~$11T benchmarked to Russell indices and more than $200B typically traded at the close. The total market cap of the Russell 3000 has grown 29% to $75.6T since the prior reconstitution. Paired trades around additions/deletions (particularly Russell 1000 2000 migration candidates) remain the most documented June-end edge. Any names added to the small-cap index face massive passive inflows at the June 26 close; names dropped from the large-cap index face symmetrical selling.

Major Stocks — What to Watch Pre-Open

Crypto — Exit Gate Still Active

BTC: $61,285.75 (Fortune/CoinDesk, 10:54 UTC June 25). Down -2.30% from June 24's ~$62,729. The Day 2 full exit gate executed per the June 24 morning brief protocol remains in force. The $61,285 level is $2,214 below the $63,500 exit threshold. No re-entry criteria have been met: SPX is still below 7,460 and VIX is still above 17.

The continued BTC weakness despite MU's blowout is telling: speculative assets are not re-engaging until the macro regime clears. BofA's 3-hike framework is a structural headwind for high-beta assets. If PCE prints cold today and SPX closes above 7,460, BTC re-entry zone ($60,000-$62,000) becomes active for a retest-and-hold; but that gate does not open until after the print.

ETH: ~$1,620-1,650 (estimated). Tracking BTC weakness. No new entries.

AI / Semis — The Demand Cycle Is Definitively Intact

Three signals in 48 hours have re-established the AI demand thesis:

1. Micron Q3 FY2026 — the most decisive AI memory data point of this cycle. $41.5B revenue (346% YoY growth), 84.6% GAAP gross margin, Q4 guidance $50B ± $1B. HBM4 began shipping for NVIDIA Vera Rubin in March 2026 with a reportedly faster ramp than prior generations. 2026 HBM supply is fully booked with multi-year agreements securing demand. The June 23 -11.4% flush was KOSPI contagion mis-pricing, not company-specific deterioration.

2. Qualcomm Dragonfly — AI inference competition broadens. Qualcomm's new Dragonfly portfolio targeting $40B in non-handset revenue by 2029 (up from $22B prior est.) with Meta's multi-generation agreement is the first serious inference-at-scale challenger to NVDA since ARM entered the conversation. For the AI trade broadly, QCOM's move validates the inference layer as a second major spending category — complementary to NVDA's training dominance.

3. Jensen Huang's Annual Meeting confirmation. Jensen stated the AI ROI question "has been answered" with the national security and data center framework intact. Combined with MU's Q4 $50B guide directly confirming Vera Rubin HBM4 demand, the fundamentals of the AI capex super-cycle are unambiguously intact.

The KOSPI recovery seals the narrative: KOSPI +5.42% to 8,930 confirms the June 23 -9.99% 'Black Tuesday' was a single-market MSCI technical, not a read on AI demand fundamentals.

Fed & Macro — PCE Is Today's Everything

May PCE at 8:30 AM ET is the only data point that matters today. Consensus: 4.1% YoY headline, 3.3% core. The Fed's preferred inflation gauge will either confirm BofA's 3-hike framework or begin to invalidate it.

The oil disinflation wildcard is real. WTI briefly broke below $70 on June 24, settling at $70.34 with a session low of $69.63; today at $69.42 it is extending below that level. Brent closed at $73.05 (-5.2%) as Hormuz tanker traffic accelerates under the 60-day US-Iran license framework. This is May data, so the June oil crash may not fully appear in today's PCE — but the direction is clear and provides the market with a credible narrative for a softer-than-feared headline.

BofA's 3-hike framework (Sep/Oct/Dec 2026) remains the dominant rate map. The BofA June 22 note calling inflation "unambiguously worse" and Deutsche Bank's independent September hike addition predated both MU's earnings and QCOM's data center expansion. Neither changes the rate trajectory on its own. PCE is the first real-time data point that can.

Q1 GDP Third Estimate (8:30 AM ET): The second estimate came in at 1.6% annualized (downward-revised from 2.0% advance). A sub-1.5% third estimate would add a growth-softening signal; combined with oil deflation, it builds the case for a PCE undershoot. But hot PCE + weak GDP = stagflationary echo — the scenario the Fed most wants to avoid and markets fear most.

Don't Buy Right Now

• QQQ / SPY at the gap-up open — NQ futures +2.1% creates a gap-up open that prices in a favorable PCE result before the data arrives. Buying the gap means paying for an outcome that prints at 8:30 AM ET — 90 minutes into the session — and risks a -1.5% to -2.5% same-day reversal if core reads above 3.5%. ES futures imply ~7,409 open: still 51 points below the 7,460 regime floor. This framework specifically bars entering longs in a breached regime without a catalyst. The PCE print is the catalyst — it prints at 8:30, not at the open.

• MU at premarket levels (~$1,083) — Micron's fundamentals are unambiguously the strongest in its history, but the premarket fade from $1,199 AH to $1,083 by 7 AM EDT (-9.7%) is the market telling you to wait. Adding at the regular-hours close level ($1,082) provides no margin of safety against a hot PCE print and a reversal toward $1,000. The night brief's $1,200-$1,250 add zone was set for a post-PCE confirmation entry — honor it.

• GOOGL — Bounced +2.1% to $348.79 on June 24 with no new positive catalyst — reflexive relief after a cumulative -14.5% decline. QCOM's Meta data center win is an incremental negative for GOOGL Cloud market share. No structural floor until mid-July Q2 earnings. Short/puts at $355-370 remain the active trade on any further reflexive bounce.

Trade Setups

1. MU — conditional long post-PCE (medium conviction · 2-4 weeks)

• Thesis: Q3 $41.5B revenue (+20% beat), Q4 $50B ± $1B guidance (+17% beat), GAAP GM 84.6%, HBM4 in production for NVIDIA Vera Rubin since March, 2026 supply fully booked. The most decisive AI memory demand confirmation in this cycle. The premarket fade from $1,199 to $1,083 is PCE uncertainty, not fundamental deterioration.
• Entry: Buy $1,200-$1,250 ONLY after PCE core prints below 3.2% and MU reclaims AH levels. If PCE hot: no adds; watch $1,050 as secondary entry on a two-day washout. Stop $1,050 on a two-close basis.
• Invalidation: PCE core above 3.5% — exit within 45 minutes of print. Samsung HBM4E qualification with NVDA confirmed. Hyperscaler guidance cutting AI capex.

2. GOOGL — short / July puts (medium conviction · 2-4 weeks)

• Thesis: Cumulative -14.5% on AI talent exodus (Shazeer OpenAI, Jumper Anthropic) with no structural floor. QCOM+Meta data center partnership is an incremental cloud-spend diversification negative for GOOGL. At ~$348, still ~20x 2026 earnings with FCF -47% YoY.
• Entry: Short $355-370 on any post-PCE reflexive pop. July $330 puts as defined-risk alternative. Size down if PCE prints cold (improving regime reduces the tailwind).
• Invalidation: GOOGL reclaims $380 daily close. Google announces a named AI executive retention program with specific commitments.

3. TBT long / XLE short — hold into PCE (medium conviction · 1 day 1-2 weeks XLE)

• Thesis: 30Y at ~4.93%, 7 bps from the 5.00% trigger; TBT captures hot PCE regime confirmation. WTI $69.42, Brent $73.05 — Iran Hormuz normalization advancing; XLE H2 sell-side Brent models at $87-90 represent a $14+ estimate gap not yet revised. Dual hedge against both inflation surprise and commodity deflation.
• Entry: Hold TBT long (add 25% on dips toward $44-45); hold XLE short ($52-55 zone; trail stop to $56). Reassess both within 90 minutes of PCE.
• Invalidation: PCE core below 3.0% exit TBT immediately (30Y falls below 4.75%); Iran deal collapses Brent reclaims $85 exit XLE short.