The only way you will be free to trade the market in front of you is to detach from your biases and impulses
London, England
Joined January 2011
- Tweets3.6K
- Following170
- Followers336
- Likes3.9K
Macro Signals retweeted
Market futures are going absolutely bonkers overnight. 🤯
Of all the companies that could’ve helped turn sentiment around in AI, I don’t think many expected $CRWV to be the one.
But here we are.
Now imagine if we actually get a follow-through day tomorrow.
Boy will that make bears angry.
Macro Signals retweeted
SOFTBANK SIGNALS AI INVESTMENT SLOWDOWN AFTER $7 BILLION LOSS
SoftBank is considering pulling back from aggressive AI spending after swinging to a $7 billion quarterly loss.
CFO Yoshimitsu Goto acknowledged a shift toward caution:
“It is only natural that there is some discussion now about waiting and seeing how things play out for a little while.”
The comments signal a potential pause in SoftBank's AI investment spree, including its exposure to OpenAI, as the AI trade comes under pressure.
One of AI's biggest backers is signaling that even aggressive investors may need to slow down.
Macro Signals retweeted
Tech just broke below its long-run trend vs the rest of the S&P 500
The last two times that happened were Liberation Day and the Iran war, and both marked the bottom
At least twice a year the market sells off on OVERBLOWN fears of something
The DeepSeek moment, the Yen Carry trade, liberation day, etc.
And every time, the market comes back to the same conclusion: AI is a secular trend, and the hyperscalers, the labs and even countries will not stop building and spending to win this race
The demand for AI has not gone anywhere, if anything, it's still accelerating
Once the market remembers that, the price rips in everyone's face as everyone rushes to jump back in
My bet is this is the exact same scenario playing out once again
Which is why I put the cash I raised in September back to work in oversold AI infrastructure names
If you want to see how I'm allocating in this market, check out my real-time portfolio here: link.milkroad.com/t5837z
I took profits and raised cash in my portfolio to >20% in September (I'm usually 95% allocated)
I did it because our Milk Road Macro Index suggested it was time to go risk-off (which was right)
My plan was to use that capital to lower my exposure to AI Infra
However, now that AI infra has sold off SO MUCH and the market looks very irrational (ie. Nvidia down 6% while the rest of the AI market is up big), there are many EXTREMELY undervalued assets in AI infra once again
- Yes the Dems won
- Yes Jensen cheated on his wife
- Yes CDS spreads are high across AI capex companies
I understand all of the fear
But, the fact is the demand for AI continues
- Enterprises are using more AI than ever before
- Agents are being adopted faster than ChatGPT was adopted
- The revenues for AI labs are going higher
- The spot price of compute continues to go higher while the supply of compute continues to remain tight
At least twice/year the market sells off on overblown fears of something, whether it be Liberation Day, The Yen Carry Trade Unwind, The start of a war, The DeepSeek Moment or something else
And every time, the market comes to the conclusion that AI is a secular trend and the hyperscalers/labs and even countries will not stop building/spending to win this race.
And once the market remembers that, the price rips in everyone's face as everyone rushes to jump back in. My bet is this is the exact same scenario playing out once again
So instead of diversifying, I bought more AI infra stocks with my cash. You can see exactly what I bought here: link.milkroad.com/t5837z
Macro Signals retweeted
"The KOSPI semiconductor sector's FY3 P/B has fallen to a historical low of around 0.7x, a level that has historically indicated a bottom of semi cycle and implied encouraging forward returns"
-Goldman Sachs via @MikeZaccardi
Macro Signals retweeted
First things first, the OG @StovallCFRA has shared this for years, so shout out to him for this one.
It took 23 trading days for the S&P 500 to be down 5% from the Oct 6th ATH.
Sam found that most bear markets started with a quick move down 5% from ATHs. In fact, only one bear market ever (34% in late '60s/early '70) took longer than 23 trading days to hit down 5%.
Could this go into a full fledged bear market? Of course it could, but for now we don't expect that and this suggests don't panic just yet.
Macro Signals retweeted
Goldman Sachs raises its forecast for global data center capacity to 273 gigawatts from 217 gigawatts by 2030.
Estimated at 50B per GW of CapEx; the majority is likely attributed to compute costs.
Jensen estimates much higher, at around 60B to 80B per GW, which is more reasonable as a metric for peak pricing power.
Macro Signals retweeted
Across our projects @GoogleDeepMind, we are witnessing a genuine phase transition in the pace of scientific discovery. You will see a glimpse of this in our releases in the next few weeks.
Macro Signals retweeted
$NBIS owns roughly 28% of ClickHouse.
On Hiive, ClickHouse trades at an implied valuation of about $56B in private markets.
That would put Nebius' stake at around $15.7B.
$NBIS market cap sits at about $35.2B.
The market is pricing Nebius like a struggling neocloud, while one minority stake alone could be worth almost 45% of the whole company.
In my eyes, that disconnect won't last forever, especially if ClickHouse ever heads toward an IPO.
Do with that information what you will.
While Nebius is crashing in the public markets, ClickHouse is at an ATH in private markets. Hiive has ClickHouse at a $56B implied valuation. $NBIS
Macro Signals retweeted
Alright lemme take a stab at this as an AI bull who's been staring at the calendar for some hope.
The Calendar:
Nov 9: Senate vote on data center permitting bill (S.5653) which would speed up federal approvals for power and transmission. Sell-side pricing odds at ~25%, so getting it done in the lame duck session would be a surprise to the upside.
Week of Nov 9: Anthropic roadshow. Valuation talk down from ~$2T to ~$1-1.3T so the bar is set low. Potentially can expect narratives from Anthropic (infinite demand for cyber uses) to take hold.
Nov 17: Nvidia earnings. See if management can convince investors that demand isn't peaking, plus previews on memory costs and data center revenue.
Week of Nov 16: Anthropic listing target. Potentially bullish the AI trade if stock pumps (price leads sentiment).
Nov 30-Dec 4: AWS re:Invent. Jassy says AWS can grow 40%+ in both 2027 and 2028 and that capex goes "meaningfully" higher next year during last earnings but with no number yet. A solid capex raise can read-through to other hyperscalers also.
Dec 1: Marvell Earnings. Down ~37% in a month after the $115B all-stock Lumentum deal even though Lumentum guided ~$350M of quarterly growth quarter. Reaffirming customer demand here and smashing the earnings and the guide can start to life the overhang.
Dec 8-9: FOMC. A hike is ~75-80% priced. A hike with "one and done" language may cause a relief rally, and a skip may cause a squeeze up.
~Dec 10: Broadcom Earnings. This would be one of the cleanest Anthropic read-through on the board.
Mid-Dec: Micron Earnings. See if management can argue that DRAM prices will continue to rise while the stock trades like memory already peaked.
The Setup:
To be clear, none of these are huge catalysts by themselves. But narratives don't need a huge event, as much as a small trigger to take hold and flip the story.
The fundamental setup is also favorable in that the SOX is trading at ~13x forward, the lowest since early 2019, that demand already exists where hyperscalers have massive backlogs, and that models are getting better exponentially with compounding improvements.
Generationally cheap forward valuations, huge committed future demand, and exponentially better models. All it needs is a spark.
Macro Signals retweeted
Extended bear market off the table, for now
Given the ongoing expansion in fiscal spending and continued credit growth, I believe comparisons to prior multi-year bear markets such as 2000, 2008, and 2022 are largely misplaced.
Major equity bear markets are typically associated with a contraction in the flow of financial assets to the private sector. That flow is driven primarily by fiscal deficits, credit creation, or some combination of the two.
Today, neither of those sources appears to be contracting. We remain far from a fiscal surplus, and fiscal support is currently accelerating rather than slowing. At the same time, credit growth continues to expand and may be re-accelerating after a modest summer slowdown.
In my view, those conditions remain supportive of nominal growth because they continue to inject financial assets into the private sector.
None of this rules out periods of equity market volatility. Rather, the point is that ongoing fiscal and credit inflows make it more likely that market weakness is ultimately met by a reflexive recovery in risk assets, rather than developing into the kind of prolonged deterioration typically associated with major bear markets.
Macro Signals retweeted
🇺🇸 US Data-Center Backlash May Lower Odds of AI Investment Bust - Bloomberg
bloomberg.com/news/articles/…
Macro Signals retweeted
Anthropic’s October ARR topped $80B, per WSJ, consistent with TickerTrends tracking
We tracked $82.33B as of Oct. 30, up 7.2% from our Sept. 24 weekly release
Macro Signals retweeted
The $NBIS death spiral.
No it isnt levered like $CRWV which will probably go to zero with the bondholders getting a new equity. Congrats to Magnetar for dumping as much as possible since the IPO. Great job.
The issue $NBIS now faces is that to achieve their ARR guidance for 2027 they need to raise about $25-30B in debt and equity. Cost of both sources of capital have quickly risen, and any attempt to raise could exacerbate that further. The other option is to slow growth and dial back the capex. But it is a double-edges sword bc then the multiple collapses on lower rev and EBITDA numbers.
And as I said last evening, their stake in Clickhouse is likely worth no more than $1B or so after tax right now. I don’t care that Clickhouse raised at a $15 billion evaluation back in January. This is a different environment now, and in my opinion, given the ARR companies probably worth about $5 billion about 1/3 of the crazy 60x sales it raised its Series D at in January.
This is what I call the “Nebius Conundrum” bc there is no easy answer. I think there’s another 30% down from here, possibly more.