0:10
SpaceX’s Quest for Warp Speed Rattles Smaller Firms in Its Wake
Before its initial public offering, the prospect of a megacap SpaceX drew capital from retail and institutional investors to both publicly listed rivals and adjacent firms. It was great for companies that Wall Street viewed as a proxy for SpaceX or that, at least, offered a vehicle for investment in the industry it dominates.
Post-debut, however, it has created something of an accordion effect for individual pure-play (or near-pure-play) space stocks. In the months leading up to the SpaceX IPO, many stocks rose as new investors sought a piece of the industry before SpaceX officially launched it into the stratosphere, much like what happens when the bellows of the wind instrument expand.
But then the accordion gets squeezed: Since SpaceX debuted, money has rotated from these firms to the newly public industry leader.
For example, satellite manufacturer AST SpaceMobile, a smaller competitor to SpaceX’s Starlink satellite internet services, has climbed 34% over the past year but taken a 43% nosedive over the past month amid portfolio realignments triggered by the SpaceX IPO.
Observation satellite firm BlackSky, which provides high-quality satellite imagery for commercial, government and defense clients, is up 18% in the past year but down 50% in the past month.
Space exploration company Intuitive Machines: up 74% in the last year, down 43% in the last month.
Rocket Lab, the company that’s widely considered the most direct competitor to SpaceX, has fallen 42% in the last month but still boasts a 130% gain in the last 12 months. Like Musk’s firm, it designs, makes and launches both rockets and satellites.
Where things go from here depends a lot on how you feel about SpaceX and the broader sector. The firm’s Wall Street debut has been tumultuous: Last week, it finished down 15%, and it has experienced single-day double-digit swings over its first two weeks, though its shares are still trading at $157 a pop, up 5% from their $135 listing price.
Looking at fundamentals, Musk’s company lost nearly $5 billion last year on $18.7 billion in revenue, the latter of which represented a notable 33% increase.
That’s not unusual among peers: BlackSky, Rocket Lab and AST SpaceMobile are still operating at a loss because they, too, are investing heavily to build capacity as they add lucrative government contracts and other future revenue sources. Of note, Planet Labs, which operates the largest commercial Earth-imaging satellite network, reported its first profitable year (on an adjusted basis) in the 12 months through January 2026.
Some believe the next few years will mean a blastoff, while others are a little more hesitant to make that call.
SpaceX bulls advocate looking to the future: Morgan Stanley bankers advised investors earlier this month that SpaceX could reach $3.4 trillion in revenue by 2040, according to a report in The Wall Street Journal. Yes, trillion. Goldman Sachs, another of the company’s IPO bankers, projected $470 billion by 2030.
To put those numbers in context, McKinsey estimates the entire space economy will be worth $1.5 trillion in 2035.
Bears have questions. Meeting the ambitious forecasts will require developing and scaling unproven technologies such as solar-powered orbital AI data centers and enabling rapid reusability of SpaceX’s Starship fleet. Then there are Musk’s purported plans to colonize the moon and Mars.
Morningstar analysts, who have proven among the most articulate skeptics, believe SpaceX stock is worth $62 and said that even maintaining its current, much higher price “implies all the company’s projects will pay off according to our most optimistic scenario, which depends on rapid Starship reusability and compelling commercialization of orbital data centers.”
Of course, if the bulls are proven right, the broader sector is likely to benefit, a prospect highlighted in surging interest from both Wall Street and Main Street.
“We are already seeing evidence of this trend through the emergence of dedicated SpaceTech ETFs and increasing investor engagement with specialist space-focused investment vehicles,” said Mark Boggett, CEO of Seraphim Space.
Before 2026, investors had just one option if they wanted to put money in a space-industry-dedicated exchange-traded fund. Today, they have over half a dozen choices.
In the short term, meanwhile, the decline in SpaceX shares this week isn’t necessarily a reflection of fundamentals, positively or negatively.
“There appear to be several factors at play, driving the pullback,” Eugenia Mykuliak, executive director of trading platform B2Prime, said in a statement. “First, the fact that SpaceX has announced plans for bond issuance has led to investors asking questions about the financing costs and how much funding the company will need going forward.
5:14
Flipping houses isn’t paying like it used to
25.5% return on home‑flipping in 2025, the lowest level since the 2008 recession.
297,045 single‑family homes and condos changed hands last year, the fewest since 2020, a 3.9% drop from 2024.
Higher mortgage rates, elevated home prices and a tight inventory are squeezing margins, while material costs—some inflated by tariffs—are adding to the strain.
In Q1 2026 the profit margin nudged up to 25.4% from 24.7% a year earlier, marking the first quarterly gain in two years and hinting at a modest rebound.
5:52
US homes aren’t getting any younger
US homes are getting some snow on the roof, metaphorically speaking. The median US home is a record 44 years old, meaning that the typical house was built before the first-ever Blockbuster, according to the Harvard Joint Center for Housing Studies. And with new construction struggling to keep up with demand, houses are getting even older—and in need of renovation.
Reno 911: Aging homes are especially common in the Northeast. According to Redfin, in 2024, the typical home in Buffalo, New York, was built 69 years ago. When a house is decades old, not every upgrade gets your blood pumping and your Pinterest fingers scrolling. Rachel Drew, director of Harvard’s Remodeling Futures Program, told the Wall Street Journal:
- Nearly 50% of all improvement spending is now for necessary maintenance, like replacing a furnace.
- In 2023, US homeowners spent an average of $9,030 on replacement projects, like windows—up 59% from 2009, even adjusting for inflation.
Not only do older homes need maintenance more often, but the repairs themselves are also more expensive: Between 2022 and 2024, inflation-adjusted costs for structural repairs jumped about 14%, the WSJ reported, citing data from the Federal Reserve Bank of Philadelphia.
If you’re forced into a home improvement project (or you just want to spruce things up), it’s worth keeping in mind that future buyers may not share your affinity for shiplap or bumblebee-themed wallpaper. Zillow reported last year that the home improvements with the best return on investment are:
- Replacing your garage door (with a 349% ROI).
- Replacing your entry door (216% ROI).
- Adding a manufactured stone veneer (208% ROI).
- Installing fiber-cement siding (114% ROI).
- Minor kitchen remodel (113% ROI).
On the flip side, renovations like swimming pools, marble countertops, and koi ponds may actually drive some buyers away. But, it’s your home, so no judgment if you want to prioritize koi over ROI.—BC
Become smarter in just 5 minutes. Subscribe to Morning Brew today.
8:02
The Crooked Smile
Quick reminder that the cost of annual subscriptions will be rising from $360 to $450 at the end of June (next week!). Monthly subscriptions will remain at $45. As ever, all existing subscribers are grandfathered at their original rates.
I know that Ray told Tammy in the final scene of The Firm that he loved “her crooked little mouth” but allow the 🐿️ a tiny bit of poetic license…
Right now, Stephen Jen’s “dollar smile” is looking decidedly crooked after Kevin Warsh’s debut press conference violently yanked up the right side of Holly Hunter’s second best “feature”.
For those in need of a quick refresher, the “dollar smile” model dictates that the greenback strengthens in two scenarios: global panics (the left side) and periods of US economic exceptionalism (the right side). Between them lies the dollar trough, where the assets this rodent likes most tend to flourish.
We were supposed to be sliding comfortably into that trough. Instead, Warsh’s 130-word hawkish ‘pivot’ last Wednesday has seen a kink in that smile emerge.
warned me that the ‘Market Gods’ would have something to say about the title to the early May ‘Benny & The Squirrel’ conversation with :
Will the removal of the crutch of Fed forward guidance finally breathe some life into the sleepy world of FX volatility? Well, it’s a start. Actually, I think seeing the back of the Fed hand-holding will put a lid on positioning extremes by market participants that I see as the dry tinder for the next risk deleveraging unwind.
Renewed FX vol has certainly done much of the heavy lifting in terms of the recent tightening of US financial conditions (red box below).
But what should we make of this collective narrative pivot around Warsh from the ‘Trump Administration water boy’ to ‘the next Paul Volcker’ that has become a staple of financial commentary over the past week?
To be honest, what did you expect? Warsh only said pretty much exactly what every Fed Chair has proclaimed upon taking office since this rodent was in short trousers.
Fun fact: Paul Volcker’s first year (‘79/’80) saw roughly 7 percentage points added to the Fed Funds rate. He certainly followed through! This eventually prompted the dollar to move - soon to be reversed aggressively.
This is what that looks like alongside the famous cigar chomper’s successors:
Even if following the traditional script, Warsh’s words were enough to blast the dollar out of the ‘2 big figure’ trading range that has existed since ‘Liberation Week’ last year.
My guess is that markets were primed for Trump’s pick to offer at least something to the doves and position a Fed willing to follow the established pattern of “looking through” any transitory impact on prices related to the Iran war.
Well, Kev’s bold rhetoric has certainly been effective at curbing inflation expectations:
Which are now at close to their lowest point (across the curve) since the Trump Administration took off.
Even if cynics might possibly argue against the predictive properties of the inflation swap market.
The latest BofA FMS also points to the effectiveness of Warsh’s words in terms of getting money managers to (i) take the hike risk seriously (without scaring off expectations for a ‘soft’ or ‘no’ economic landing scenarios); (ii) reel in the US dollar underweights from their 2025 extremes; and (iii) pare back their “foolish” (not my view, obviously!) recent experimentation with overseas stocks.
The move was certainly enough to finally shake this rodent out of his long duration fixed income hedge () at the beginning of the week.
It was a good entry in April but I should have taken the hedge off when I started adding my Brent crude oil position at the beginning of June, as it is - to a certain extent - a similar bet. More on rates strategy and hedges at the end of this note.
However, what if this dollar move is actually telling us that the barrel counters have a point after all and that energy markets are complacently sleepwalking into a second-round crisis created by a 2-3MMbpd net deficit in crude that has not disappeared?
My base case is that this is an underpriced scenario even without a renewal of tensions in The Gulf - something which (as at Saturday morning Australia - with futures markets shut for the weekend) seems far from settled.
It goes without saying that my P&L in crude is agreeing vocally with those inflation swap traders. No bueno.
The recent move in inflation expectations (at least over shorter timeframes) is certainly implying that my oil call is wrong.
12:46
AI Chips Are Hard. Memory Is Harder.
The biggest misconception in today’s semiconductor industry is that designing AI chips is the pinnacle of engineering difficulty.
It is not.
In reality, the technology that has proven far more difficult to master—and far more resistant to new entrants—is memory.
The rapid emergence of custom AI chips from cloud service providers has reinforced this misconception. Companies such as Google, Meta, Microsoft, Amazon, Tesla, Open AI and countless startups are now developing their own AI accelerators. This has created the impression that semiconductor design has become democratized.
Yet if we look at the memory industry, the story is completely different.
After decades of consolidation, the world’s advanced DRAM market is still dominated by only three companies. The number of suppliers capable of mass-producing high-bandwidth memory (HBM), the most strategic component of AI infrastructure, remains exactly the same.
The contrast is striking.
If AI chips are truly the hardest products in semiconductors, why are dozens of companies able to build them while almost nobody can build memory?
The answer reveals one of the deepest truths about semiconductor manufacturing.
When people say a company is “developing its own chip,” they usually mean a logic chip—a CPU, GPU, AI accelerator, networking ASIC, or custom processor.
These chips certainly require extraordinary engineering expertise.
Architects must design instruction sets, optimize dataflow, reduce power consumption, balance compute and memory bandwidth, develop software stacks, and verify billions of transistors before tape-out.
None of this is easy.
However, one enormous advantage exists.
Most companies do not need to manufacture these chips themselves.
Instead, they rely on foundries such as TSMC to transform digital designs into physical silicon.
This separation between design and manufacturing fundamentally lowers the barrier to entry.
A startup with a brilliant architecture can leverage the same advanced manufacturing technology used by the world’s largest semiconductor companies.
The foundry ecosystem, combined with mature EDA software and licensable IP from companies such as Arm, has dramatically accelerated innovation.
Memory does not work this way.
Unlike logic semiconductors, advanced DRAM production is almost entirely based on the integrated device manufacturer (IDM) model.
The companies designing the memory are also responsible for inventing the process technology, developing manufacturing recipes, optimizing equipment parameters, and continuously improving yield.
There is no equivalent of a “TSMC for DRAM.”
A company cannot simply finish a DRAM design and send it to an external manufacturer.
Instead, it must master every layer of production itself.
That includes capacitor formation, cell architecture, leakage control, lithography integration, process chemistry, defect management, refresh optimization, reliability engineering, and manufacturing yield.
Many of these technologies are accumulated through decades of production experience rather than written documentation.
Some of the industry’s most valuable knowledge exists only as tacit know-how inside manufacturing teams.
Money alone cannot buy it.
The greatest challenge in memory manufacturing is not designing a working cell.
It is producing billions of identical cells with almost perfect consistency.
A modern AI processor may contain over one hundred billion transistors.
A DRAM chip contains an even larger number of memory cells that are nearly identical.
Every one of those cells must behave correctly.
Even microscopic process variation can introduce leakage, retention problems, or bit failures.
Unlike logic chips, where redundancy is relatively limited, memory products are extraordinarily sensitive to manufacturing defects because every defective cell directly reduces usable capacity or long-term reliability.
As process nodes continue shrinking, maintaining acceptable defect density becomes exponentially more difficult.
Success depends not only on transistor scaling but on manufacturing discipline across thousands of process steps.
This is why DRAM manufacturing is often described as a yield business rather than simply a design business.
If conventional DRAM is already difficult, High Bandwidth Memory represents another level entirely.
HBM combines advanced DRAM fabrication with sophisticated three-dimensional packaging technologies.
Modern HBM requires through-silicon vias (TSVs), wafer thinning, precision stacking, thermal optimization, advanced signal integrity, and increasingly sophisticated bonding technologies.
Every additional memory layer increases manufacturing complexity.
Every interface introduces another opportunity for yield loss.
The challenge is no longer limited to producing excellent DRAM dies.
Manufacturers must also integrate multiple dies into a single package while maintaining electrical performance, mechanical reliability, and thermal efficiency.
The result is one of the most comp
18:01
Weekly Market Metrics | What to expect next week
Hello and welcome to the Weekly Market Metrics! I’m Sandeep Rao, and we’re in Week 26 of 2026.
The Strait of Hormuz has opened, and oil wasted no time in taking the elevator down. As of now, Brent crude is at its lowest since before the Iran war — around $73 a barrel, nearly 39% off from the $120 peak. For India, the key thing is the Rupee. We import close to 85% of our crude, so when oil drops this hard, the current account improves, and the Rupee gets stronger. The Dollar-Rupee is around 94 right now.
But the monsoon is a separate concern. The deficit has widened to 43% nationally, the advance has stalled near Mumbai, and both IMD and the US National Oceanic and Atmospheric Administration are flagging a moderate to strong El Niño this year. Central India is at a 63% deficit. If July and August don’t deliver, kharif crops — summer-sown staples like rice, pulses, oilseeds — will feel it, and food inflation could concern the RBI.
And if you follow the startup ecosystem, you could not have missed Kunal Shah being named global head of WhatsApp. Founders stepping out to run someone else’s company is rare — but with Kunal, the predictable move was never on the table. There’s a short Rainmatter conversation with him and Dinesh Pai that just dropped, possibly his last on record as CRED’s founder and CEO.
Markets did reflect the general positive backdrop — Nifty is up roughly 0.2% on the week, trading around 24,050 as of Thursday’s close. This was a short week on the eve of Muharram.
Not much changed this week, which is exactly what you’d expect after last week’s rally. Nifty 50 ended almost flat, up just 0.2%, while Nifty Bank continued to inch higher with a gain of 0.9%.
The one exception was Nifty Mid Select, which corrected 1.3% after being one of the strongest performers over the last few weeks. Even then, it remains up nearly 19% over the last three months and close to 5% for the year.
The broader takeaway hasn’t changed. Nifty Bank continues to be one of the strongest large-cap indices, while Nifty 50 still remains down almost 8% for the year. One quiet week doesn’t change the trend.
On YTD numbers, Microcap 250 continues to lead at 10.5%, followed by Smallcap 250 at 6.2%. Next 50 is up 3.6%, while Midcap 150 has eased to 1.7%. Nifty 50 remains the clear laggard at -8%. The past week hasn’t changed the broader picture — 2026 continues to reward investors willing to move beyond the headline index.
Weekly
A truncated week with just 4 trading days. Nifty gained just 43 points, or 0.2%, to close at 24,056. It continues to trade above both the 10-week and 20-week SMAs, holding on to last week’s breakout.
Nifty is currently hovering around the 24,000 mark. Last week’s breakout has held, but there still isn’t enough evidence to say the index has decisively moved higher. For now, 24,000 remains the level to watch, while 24,600 continues to be the next major resistance.
W26
W25
Even in a relatively low VIX environment, Nifty’s average daily range rose to 242 points this week from 149 points last week — a reminder that low VIX doesn’t always mean small intraday ranges. Monday was the quietest session at just 95 points. Tuesday and Wednesday recorded the widest ranges at 351 and 301 points respectively, while Thursday was still fairly active with a 223-point range.
Daily
On the daily chart, Nifty had a volatile but largely directionless week. After a quiet Monday, Tuesday saw a sharp sell-off, while Wednesday and Thursday recovered most of those losses. The recovery, however, stalled near 24,200, with Nifty unable to close above it. The index ended the week almost flat but, more importantly, continued to hold above 24,000.
Nifty continues to hold above both the 21 EMA and the 50 EMA. It briefly touched the 100-day EMA this week but couldn’t close above it. The 200-day SMA at around 24,900 remains the next major hurdle overhead.
Hourly
Nifty largely stayed above the 50 EMA this week. Tuesday’s sell-off briefly broke that structure, but Wednesday’s recovery quickly reversed it. Thursday saw a move to fresh weekly highs before some profit-booking into the close brought the price closer to the EMA. Overall, the hourly trend continues to remain positive.
Nifty Weekly Expiry — Tuesday, 23 June
The ATM straddle opened at around 90 points and steadily decayed to nearly 57 points by around 11:30 AM. But the sharp afternoon sell-off, which pushed Nifty into a 351-point daily range, saw the straddle briefly spike back to around 108 points before theta took over again into the close. Realised volatility ended up much higher than what the opening straddle had implied, making it a much harder expiry for non-directional sellers. Buyers, on the other hand, must have had a good day. Let me know in the comments how your expiry went.
The 24,000 level continues to be the key support. Just below that, the 23,800–23,650 gap zone remains an important support area if the market sees a deeper pullback.
23:15
Mercedes, GIC to Back Autonomous Driving Firm Momenta’s IPO
The Momenta.ai booth at the IAA Mobility 2025 expo in Munich, Germany, on Monday, Sept. 8, 2025. The Munich car show is taking place as trade tensions linger with Beijing following the European Union’s decision last year to impose tariffs on EVs imported from China. Photographer: Krisztian Bocsi/Bloomberg
23:41
South Korean Regulators Are Lamenting Leveraged ETFs
Gary Gensler, former chairman of the US Securities and Exchange Commission (SEC), during the StubHub Holdings Inc. initial public offering (IPO) on the floor of the New York Stock Exchange (NYSE) in New York, US, on Wednesday, Sept. 17, 2025. StubHub Holdings Inc. shares are indicated to open at $24.50/$25.50 each, as much as 8.5% above the initial offering price of $23.50. Photographer: Michael Nagle/Bloomberg
24:13
Money Roundtable: Hidden Inflation, Social Security
Bloomberg's David Gura and Romaine Bostick join Lisa Mateo and Tom Keene on "Bloomberg Money." They discuss this week's US inflation report, Apple's price hike, and the future of social security. (Source: Bloomberg)
24:33
CFTC Conducting Broad Probe Into Prediction Market Polymarket
A Polymarket advertisement in a subway station in New York, US, on Thursday, Feb. 5, 2026. Kalshi and Polymarket, which have been assailed by critics for encouraging financial risk taking by making betting more accessible, are now using the promise of free groceries to win over New Yorkers.
