Bulls & Bears
Monday opened on fumes. Volume was 28% below the 20-session average after the long Fourth of July weekend. The Dow punched above 53,000 for the first time, closing at a record 53,057. Chip stocks bounced 2.2% on the SOX after shedding 12% across the prior two weeks. Broadcom extended its Apple partnership to 2031. SK Hynix kicked off marketing for its US listing. But participation was thin, and the move told you more about positioning than conviction. As Mark put it on the morning call: "not a night to necessarily take a huge indicator from."
Tuesday was the day everything turned. Samsung reported a 19-fold increase in operating profit — a record — and the stock dropped 7%. SK Hynix fell 6%. The KOSPI slumped 5%. The Philadelphia Semiconductor Index cratered 4.7%. The read was straightforward: all the good news was already in the price. Away from chips, the real trouble arrived from the Strait of Hormuz. Three tankers were attacked in 24 hours. The US Treasury revoked the waiver allowing sales of Iranian oil. WTI jumped 5% to $72. Brent spiked 5.2% to nearly $76. SpaceX joined the Nasdaq 100 and promptly fell 7% on its first day in the index.
Wednesday was chaos. Trump told reporters at the NATO summit in Ankara that the Iran ceasefire was "over." He said the US would "probably hit them hard again tonight." Three hours later, it did. The US struck approximately 170 targets across Iran over Tuesday and Wednesday combined, including air defences, IRGC small boats, and coastal radar. Iran retaliated by hitting US bases in Kuwait and Bahrain. Brent briefly topped $80 for the first time since June 22. The S&P fell 0.3%. The Dow dropped 1.1%. Europe's STOXX 600 had its steepest one-day decline since March. But — and this is the part that mattered — the Nasdaq finished up 0.2%. Chip stocks bounced 2%. The AI trade was already finding buyers again while bombs were falling.
Thursday felt like a different planet. Trump said Iran had called wanting to make a deal. Oil dropped 2%. Chip stocks rallied 3.1%. Meta announced plans to put a custom AI chip into production in September and double its computing power to 14 gigawatts next year. Micron bumped its planned US investment to $250 billion through 2035. SK Hynix priced its IPO at $149 per ADR — seven times oversubscribed — raising $26.5 billion in the largest US listing by a foreign company ever. The S&P closed at 7,541. The Nasdaq hit 26,207. The market's attention span for geopolitics is about 18 hours. For AI, it's infinite.
- Dow record >53,000 Monday; S&P 7,541 Thursday; Nasdaq 26,207
- SOX −14% over 4 sessions through Tuesday, then +5% bounce Wed-Thu
- Samsung 19-fold profit increase — stock fell 7%. KOSPI −5%
- US struck ~170 Iranian targets Tue-Wed; Iran hit US bases in Kuwait & Bahrain
- Trump declared Iran ceasefire "over" at NATO summit; Brent briefly >$80
- Trump then said Iran called wanting a deal; oil fell 2% Thursday
- SK Hynix IPO: $26.5B at $149/ADR, 7x oversubscribed — biggest foreign US listing ever
- Meta: custom AI chip production in September; doubling compute to 14 GW in 2027
- Micron: $250B planned US investment through 2035
- SpaceX joined Nasdaq 100 — fell 7% on Day 1; analyst initiations mostly bullish
- Fed minutes (first under Warsh): committee evenly split on hikes; a few wanted to move immediately
- 3 tankers attacked in Strait of Hormuz; US revoked Iranian oil sales waiver
- Global memory monthly sales hit record $74.6B (UBS)
- Volume 28% below 20-session average Monday — thin participation
- 30-year Treasury auction drew highest yield in nearly two decades
- IMF cut 2026 global growth to 3.0%; expects 2027 rebound
| Index | Level | Mon | Tue | Wed | Thu |
|---|---|---|---|---|---|
| S&P 500 | 7,541 | +0.7% | −0.5% | −0.3% | +0.8% |
| Nasdaq Composite | 26,207 | +1.1% | −1.2% | +0.2% | +1.3% |
| Dow Jones | 52,487 | +0.3% (Rec) | −0.3% | −1.1% | +0.3% |
| SOX (Semis) | — | +2.2% | −4.7% | +2.0% | +3.1% |
| Brent Crude | ~$76 | ~flat | +5.2% | +5.4% | −2.2% |
| STOXX 600 | — | Pulled back | Tech selloff | Steepest drop since Mar | +first gain in 4 |
The story everyone thought was finished just got ripped back open. Brent crude went from pre-war levels near $68 to briefly topping $80 in a single week. The trigger was clear enough: three tankers attacked in the Strait of Hormuz, the US revoking Iranian oil sales waivers, and 170 strikes across Iran. Trump calling the ceasefire "over" at NATO in Ankara didn't help.
Bond yields followed oil higher. The 10-year UST jumped to 4.57% at Wednesday's peak. Germany's 10-year hit 3.09%. The UK pushed towards 5%. Euro zone CPI expectations leapt 14 basis points in a day. Traders priced in 36 basis points of Fed tightening by year-end, up from 25 bps just 24 hours earlier. The Bank of England saw 32 bps priced. This is a meaningful repricing.
The Fed minutes — the first under Warsh — added to the tension. A few members pushed for immediate hikes. The committee was evenly divided. "Most" saw scenarios where inflation cools on its own. "Most" also saw scenarios where it stays elevated due to AI-related demand, the Middle East, or tariffs. NY Fed's Williams said he'd grown less worried about inflation because of energy price declines. That comment aged badly by Wednesday evening.
Strait of Hormuz traffic is running at 40-50% of pre-war levels. The "peace dividend" on oil that drove the post-ceasefire rally has been partially erased. Whether it holds depends on a single variable: does Trump's "they want to make a deal" comment from Thursday translate into anything real? The oil market doesn't know. Neither does anyone else.
✓ Trump said Iran called wanting a deal Thursday; oil retreated 2%
Oil was the main event. Brent swung from $71 on Monday to above $80 on Wednesday and back to $76 by Thursday's close. The violence was the catalyst: three tanker attacks, 170 US strikes across Iran, Trump's "ceasefire is over" rhetoric, and Iran threatening to shut Hormuz entirely. Then Trump said Iran called wanting a deal, and crude dropped 3% in an afternoon. If you're trading oil on headlines alone right now, you're going to get whipped around. The underlying picture hasn't changed much: Hormuz is half-open at best, and the supply recovery that drove prices to pre-war levels in June was always fragile.
Gold drifted lower most of the week, which seems counterintuitive given the geopolitical chaos. The problem is yields. When 10-year Treasuries jump to 4.57% and the dollar firms, gold struggles to compete as a non-yielding asset. It finished around $4,081, still historically elevated but well off the January highs near $5,589. Goldman maintains a $4,900 year-end target. The next catalyst is either a genuine de-escalation in the Middle East (bearish for gold) or a funding crisis (bullish).
The story nobody is talking about: global memory chip monthly sales hit a record $74.6 billion, per UBS. Sam Altman said rising memory and compute costs were "definitely a headwind" for OpenAI's expansion. The scarcity trade in AI hardware is real. The question is whether the buyers can keep writing the checks.
Santos (STO) +6.2% Wed
Energy stocks were the obvious winners from the Hormuz chaos. Santos led the gains as Brent spiked above $80. If the US-Iran situation deteriorates further, the Australian energy sector is the direct beneficiary. Santos has Hormuz-independent LNG exposure, which makes it cleaner than pure crude plays.
Yancoal (YAL) +3.8% Wed
Coal miners rode the energy spike alongside oil names. With European gas prices sensitive to any Middle East disruption, thermal coal stays in demand as a fallback fuel source. Yancoal's balance sheet is in good shape and the dividend yield remains attractive if prices hold.
Healthcare — RMD / CSL / COH Value territory
Mark flagged all three as buying opportunities this week. ResMed is market-leading with Philips still banned in the US. CSL hit its cheapest valuation since listing. Cochlear likewise. The GLP-1 scare on ResMed hasn't shown up in the data — most patients use both therapies in conjunction. These are not broken businesses; they're broken prices.
BHP (BHP) −2.9% Wed
Port Hedland unions launched industrial action over pay, dragging BHP while the broader materials sector sold off on base metals weakness. Iran tensions hit copper and iron ore prices as the market priced in demand risk from a potential global slowdown. A strike at Pilbara is never welcome.
WiseTech Global Leading tech losses
The ASX tech sector followed its US counterpart lower on the chip selloff days. WiseTech continues to carry the governance overhang from last year. Until that cloud clears, any broad tech weakness hits it harder than peers.
Adairs (ADH) −2.2%
Warned of an expected $43 million loss for FY26. The consumer discretionary story in Australia remains ugly for anyone not named JB Hi-Fi. Housing-adjacent retail is getting squeezed by rates, cost of living, and a customer who has stopped spending on anything that isn't essential.
SK Hynix IPO $26.5B raised
The biggest foreign IPO in US history. Priced at $149/ADR, seven times oversubscribed. Baillie Gifford, Coatue, and Situational Awareness Partners indicated interest for up to $7 billion. The memory trade is alive and institutional appetite for AI hardware exposure is nowhere near exhausted. ADRs begin trading Friday on Nasdaq.
Meta Platforms (META) Neocloud pivot
Debuted its first AI image model under the rebuilt AI lab. Then the big move: planning to put a custom AI chip into production in September and double total compute power to 14 GW next year. Morgan Stanley sees every 250 MW leased at $40/Watt adding ~$3 to 2028 EPS. The neocloud option is easier to execute than a full hyperscaler build, and it monetises spare capacity.
Micron (MU) +4.5% Thu
Announced it would increase planned US investment to more than $250 billion through 2035, plus a separate $3 billion commitment to the domestic semiconductor supply chain. The stock settled 4.5% higher. Memory is the bottleneck, and Micron is building capacity to meet it.
SpaceX (SPACEX) −7% Day 1 in Nasdaq 100
Joined the index Tuesday and immediately sold off. The company makes limited revenue against its $1.75 trillion valuation. A wave of analyst initiations (BofA, Citi, Deutsche, Goldman) were mostly bullish, but the quiet period ending and the index inclusion created a "sell the news" dynamic. If sentiment turns, SpaceX alone could drag the Nasdaq by hundreds of billions.
PepsiCo (PEP) −3.3% Thu
Beat on revenue, missed slightly on adjusted EPS. In a market where expectations are sky-high heading into earnings season, "beat on one, miss on the other" gets punished. Consumer staples are supposed to be the safe trade in a rotation — Pepsi's report suggests even that isn't straightforward.
AstraZeneca (AZN) −5.9% Thu
Wainua drug failed a late-stage trial for rare heart disease. Partner Ionis Pharmaceuticals cratered 24%. Pharma is binary at the late-stage trial level. One failure and months of momentum evaporate. AstraZeneca's pipeline is deep enough to absorb it, but the market doesn't think in portfolio terms on trial-result days.
-
Middle East/Hormuz impact, defence ETFs, healthcare value (RMD, CSL, COH), US earnings season setup, memory chip stocks.
-
Dividend stocks coming back into focus with a defensive tilt. Healthcare names Cochlear, ResMed, and CSL in the spotlight.
-
ASX under pressure from renewed Middle East conflict and global chip selloff. Energy sector rally, BHP strike, Telstra outage, RBNZ rate hike to 2.5%.
-
the call: 06th July | EOL WAF MP1 IPG XRF RMD OCL LLC ALX KOVCovered EOL, WAF, MP1, IPG, XRF, RMD, OCL, LLC, ALX, KOV.
-
Stock of the Day: Worley (WOR)Deep dive on Worley as an energy services play in the current Middle East environment.
-
the open: ASX to fall & Fed in focusPre-market briefing covering ASX outlook, Fed minutes preview, and Warsh's first policy meeting as chair.
"Are we in an AI bubble? We think the answer depends on whether AI can turn today's scarcity into tomorrow's abundance."
"It's premature to say whether this is 'the big one', i.e. a major structural reversal in the AI trade. It does seem fair to say that the era of the 'easy' AI trade — the one with the smooth parabolic rally — is indeed in the rearview mirror."
"It's a big wake-up call for markets. The expectation had been that we were likely to start to see the flow of oil coming back into the markets, and we saw inflation expectations being dialled down."
"The next correction could very well be much larger than people expect because of the crazy amount of speculation that's occurring."
"One thing is certain: future policy is heavily contingent on the political situation in the Middle East."
Morgan Stanley flagged that equity repo financing costs hit 200 basis points above the fed funds rate at quarter-end — the highest since December 2024. Even after normalising to 89 bps, the structural demand for leverage hasn't gone anywhere. Dealers hold about $211 billion of equity financing exposure on their balance sheets. Leveraged ETFs are concentrated in tech and semis. The put-call ratio — downside insurance — is at its lowest level on record. Fund managers are at their lowest cash levels in a decade.
In plain English: everyone is fully allocated, leveraged into one direction, and not hedged. Morgan Stanley's Tobias said it directly: "the marginal buyer has become one that's been increasingly reliant on leverage." Sam Earl at Barclays called it a straightforward supply and demand problem — demand for financing has surged while dealer balance sheet capacity hasn't kept pace. The next quarter-end will test this again.
Trump bombed 170 targets across Iran on Tuesday and Wednesday. The market went red. By Thursday afternoon, chip stocks were rallying 3%, Meta was announcing a 14-gigawatt compute buildout, and SK Hynix had priced the biggest foreign IPO in US history. That is roughly eighteen hours of attention span for a genuine military escalation between two major powers.
That should worry you more than it reassures you. The market has been trained by a decade of "buy the dip" to treat geopolitical risk as noise. Usually, it's right. But the current setup has a specific vulnerability: everyone is in, nobody is hedged, and the margin of safety is thin. Fund managers at their lowest cash in a decade. Put-call ratios at record lows. Equity repo costs spiking. Leveraged ETFs concentrated in the exact same names that drove the rally. If something actually breaks — a sustained Hormuz closure, a genuine funding squeeze, or a hyperscaler earnings miss — the exit door is narrow and everyone will try to fit through it at once.
The AI trade isn't over. SK Hynix at seven times oversubscribed. Global memory sales at $74.6 billion. Meta doubling its compute footprint. Micron committing $250 billion. The demand is real. But the demand is being met by levered buyers at stretched valuations with no hedging. That is a different kind of risk than "is AI real?" The answer to that question is obviously yes. The question that matters now is: what happens when the 14% chip selloff doesn't bounce back in two days?
For us, the playbook is clear. We're happy to buy the index on dips using a lookback structure over the next three to six months. Australian healthcare — ResMed, CSL, Cochlear — is where we see the best risk-reward right now. These are quality companies at prices that were unthinkable eighteen months ago. Defence exposure via a broad ETF makes sense while NATO countries ramp spending to 3-5% of GDP. And we'd be adding energy selectively — Hormuz is still half-open and the supply picture can deteriorate quickly.
But we're also telling clients to hedge. We've got a four-month short strategy designed for exactly this setup: low cost, a bit of sleep at night, protection in case these stacking risks actually converge. Earnings season starts next week with the banks. If the hyperscalers don't deliver revenue — not promises, revenue — from all that capex, the index is in trouble. And the Nasdaq's newest member is a $1.75 trillion company that makes next to no money. That alone could cost the index half a trillion dollars on a bad week.
Be prepared, not reactive. The risks are stacking up. That doesn't mean they all hit at once. It means when they do, you want to have already made the decision, not be making it with the index halfway down.
