June 2026: Peace, Space and Semiconductors

June was one of those months where you had to hold two completely contradictory thoughts in your head at the same time and not flinch. Markets are up. Risk is rising. Both are true.

The semiconductor index has now done 180% in twelve months, which is either the most rational outcome in the world given what's happening with AI capex, or a sign we're getting into priced-for-perfection territory. Probably a bit of both. We've been active in US memory names through the rally, but we've also been trimming. There's a point where the thesis is still intact but the position sizing gets uncomfortable, and a few names are there.

Then there's the Fed. Jerome Powell spent four years basically telling markets what he was going to do two weeks before he did it. Kevin Warsh's first act as chair was to stop entirely. No forward guidance. Shorter statement than anyone expected. And a cluster of FOMC members flagging possible hikes before year-end, the first time that's happened since the hiking cycle ended. Rates jumped, the dollar strengthened. Anything rate-sensitive got sold. We don't think that dynamic resolves quickly.

SpaceX went public. Largest IPO in history. Priced at $135, ran to $200 in days, then gave back 23% in a week. Six hundred billion in market cap, gone. Retail was still net buying on the way down, which tells you something. The float structure alone is worth understanding before anyone touches this — a firehose of new supply hitting the market across a series of lockup expirations out to 180 days.

And sitting under all of it, quietly, the US-Iran deal. If that falls apart.... who knows what July will bring

---

KEY THEMES

What drove the month

---

Month in Review

The Key Moments

---

The SOX Soars

106% year-to-date. Sit with that number for a second.

While the S&P 500 has ground out a respectable 9–10% this year, the PHLX Semiconductor Index has more than doubled. If you had meaningful exposure to semis in January, you've had a genuinely great year, in an index that was already elevated and that most commentators were calling frothy.

The short answer is AI. The longer answer is that the AI buildout has turned into something far more capital-intensive than most people anticipated eighteen months ago. Microsoft, Google, Amazon, and Meta are collectively spending north of $300 billion on data centre infrastructure this year. Every dollar of that spend eventually becomes demand for chips, memory, networking silicon, and power management equipment.

Nvidia is the obvious headline. But the strength has spread much further: memory, networking silicon, power management chips, testing equipment. Even Intel has caught a bid on renewed server CPU demand. This isn't a one-stock story, it's a structural cycle playing out across the whole supply chain.

Now, the obvious question: is this 1999? Some of the parallels are uncomfortable. Valuations are elevated. The AI investment thesis has attracted genuine believers and genuine speculators in roughly equal measure, and it's not always easy to tell them apart. But the revenue is real. These aren't dot-com companies burning cash on projected eyeballs. Nvidia's data centre revenue is enormous and growing fast. The hyperscalers buying chips are profitable businesses making considered capital allocation decisions.

Two risks worth taking seriously. First, Taiwan. The concentration of advanced manufacturing in a geopolitically sensitive region is a tail risk that gets discussed, set aside, discussed again, and set aside again, but it hasn't gone away. Second, demand timing. There's a real question about whether enterprise AI deployment will scale fast enough to justify the infrastructure being built ahead of it. If it doesn't, you get a correction.

For investors already in, the 180% gain over the past twelve months is a prompt to review position sizing. Not necessarily to sell.... but to be deliberate. A holding that's grown to twice its intended weight is now running two portfolio risks, not one. The sector is volatile by nature. But the underlying demand story is durable enough that most long-term investors should want some exposure through the cycle.

The Philadelphia Semiconductor Index has dramatically outperformed the market in the last year

---

New Sheriff in Town: Kevin Warsh as Fed Chair

Jerome Powell spent years telling markets exactly what the Fed was about to do. Kevin Warsh's first act as chair was to stop doing that.

At his debut FOMC meeting in mid-June 2026, Warsh held rates steady at 3.50–3.75% — exactly what markets expected. The surprise was everything else. The policy statement was shorter, stripped of the guidance language markets had come to rely on. No forward guidance. No "data-dependent" phrasing. Just a statement of the current position and a clear signal that the committee was watching inflation, not managing sentiment.

His first meeting was a statement of intent: the Fed will respond to data, not manage market expectations. Two-year Treasury yields jumped, Thirty Year Yields fell &the dollar strengthened, all signs of long term confidence in the "intent"

The easing bias was removed and replaced by a hiking bias from the voting members (Warsh declined to vote) with the hawkish lean reflecting genuine concern about inflation. Core PCE has been above 2% for years. Warsh made it clear he considers that unacceptable rather than inconvenient. Part of the pressure is structural, the Iran deal removed the oil supply shock, but services inflation remains elevated and wages are still running hot.

What Warsh announced alongside the rate decision is arguably more interesting. Five task forces to review the Fed's communications strategy, its data sources, its inflation framework, its productivity metrics, and its inflation measurement approach. That's a root-and-branch review of the analytical infrastructure the Fed uses to make decisions. It signals that Warsh isn't just adjusting the tone, he's rebuilding the framework.

For investors, the practical implication is clear: the Fed just became harder to read. That means jobs numbers, CPI prints, productivity data, and energy market moves matter more directly to asset prices than they have for years. More volatility seems likely in the near term as markets recalibrate to a less hand-holding Fed.

---

SpaceX SPCX priced at $135 on June 12, raising $75 billion in the largest IPO ever recorded. The stock "sky rocketed" to $200 within days, peaking at a $2.4 Trillion dollar valuation and making Elon Musk the first Trillionaire and 4400 long suffering staff multi millionaires.

Then came the three-day rout: −23%, $600 billion in market cap erased, back to $154.60. A $20B bond raise and a $6.3B Reflection AI computing deal announced in the same week. KeyBanc slapped a sector-weight rating on it before the ink dried. Retail is still net buying.

The float structure is the story. Starting at 4.3%, the tradeable float expands across a series of lockup expirations, reaching 100% by the 180-day mark. That's not a gradual drip. It's a firehose of new supply pointed at whoever is holding the stock at each milestone.

Elon Musk got to work quickly, acquiring Cursor for an all stock deal worth $60b and launched (i know, another rocket pun) a huge $25B Corporate Bond issuance, which took the wind out of the stocks sails

Factbox: Five Dates that could Make or Break SpaceX

---

May-June Recommendation Highlights

Five wins and one loss from the positions that closed across May and June. The semiconductor and Al infrastructure thematic did the heavy lifting - AMD AMD, Micron Technology MU, Lam Research LRCX, and TSMC TSM all entered over the last 6+ months, on the same broad read and all delivered. We exited nearly all recommendations in the chip sector early this week, as the valuations and technicals look as overbought as we've ever seen.

Electro Optic Systems EOS on the ASX side was the standout surprise as it played catch up with the hot "drone" thematic that has seen the likes of DroneShield DRO and recently listed Boresight Ltd BST have significant rallies

Palantir Technologies PLTR and Bentley Systems BSY were the ones that didn't cooperate, closed at a modest loss after the Al valuation premium compressed.

---

Bulls vs Bears In July (and Beyond)

The US-Iran MOU is the only thing that matters right now. Everything else is noise. If that deal falls apart — and the signals out of Tehran suggest Iran may have already walked away — oil snaps back hard, the risk premium re-enters the market, and the rate cut thesis collapses entirely. Watch Phase 2 talks out of Switzerland carefully.

Copper is the one I keep coming back to. The hyperscaler build-out isn't slowing. Data centres are eating copper at a rate the market still hasn't fully priced in, and there's no real substitute at scale. The supply side is constrained. That's a multi-year thesis, not a trade.

Gold and silver I'm less excited about. They're not yield assets, and in a hawkish environment that matters. Rate cuts are now pushed to the end of 2027. There's a real chance of hikes by October in the US. Owning zero-coupon insurance in that environment has a real cost.

Locally, Australia's May CPI and employment numbers are the events to watch. The market has priced out further RBA hikes. If CPI comes in hot, that changes quickly. It could be a genuine shock to rate-sensitive parts of the ASX (property trusts, consumer names, highly leveraged growth stocks.)

We've been carrying medium levels of cash levels for about six weeks and trying to remain agile and patient. With our recent exit in the US Chip Stocks, and some general tidy up of "non-performers" we wont be adding at these levels, and sit at high levels of cash across the board.

The next 6–12 months look more complex than the last 6, more data-dependent, more Fed-sensitive, more headline-driven. In that environment, fortune favours the patient, not the brave, and I wouldn't be surprised to see a panic sell-off sometime before the next newsletter