Paradi Lab

Rough.

Hedge funds have net sold Global AI stocks for five consecutive sessions (Oct 8 - Oct 14).

From a flow perspective, all of the cumulative net buying in this group since the start of 2025 by HFs has now been unwound.

Source: GS FICC & Equities

Memory fundamentals are still very strong (stronger than Q1).

$Micron Tech(MU.US) CEO:

- "We expect memory and storage supply demand conditions to be much tighter in fiscal 2027 and 2028 than they were in 2026."

- "We do not have line of sight to when supply and demand will return to balance."

Back in July, $NVIDIA(NVDA.US) supply commitments rocketed from $119B -> $279B which was "primarily related to the procurement of memory", with ~$175B of commitments for 2028-29...

$Micron Tech(MU.US) also said that >75% of FY27 output is already committed + "a majority of discussions with our customers today are already around 2028."

So as a reminder....ultra long-term demand visibility in the context of the memory market.

Samsung’s Q3 prelim last week was 107.4T Won O.P. on 195T Won revenue:

- So that's ~55% margins company-wide.

- Micron did 87% GM in their most recent quarter

- $Sandisk(SNDK.US) 84.6% GM.

All accelerating/stabilizing margins as of their most recent fiscal Q's.

Mass production from new capacity (from Micron, Samsung and SKHY) is also not expected to come online for quite some time.

In other words, the shortage will intensify next year at least.

Then long-term contracts solidify long-term earnings further via price floors + committed volumes:

- $Micron Tech(MU.US) with 26 SCAs worth >35% of revenue through to 2030 with some now running into 2031.

- $Micron Tech(MU.US) CFO: "even at floor prices, we expect margins meaningfully above any prior cycle peak margins."

- $SK Hynix(SKHY.US) has ~10 LTAs of ~5 yrs w/ pricing structures "designed to respond to price fluctuations."

- Sammy (as of recent reporting) holds ~70% of memory capacity under LTAs until 2031 with $Microsoft(MSFT.US), $NVIDIA(NVDA.US) and $Alphabet(GOOGL.US), with ~25% of contract value prepaid.

- $Sandisk(SNDK.US) also has their "New Business Model" agreements with 8 customers that lock in ~66% of fy28.

Then on buybacks, $Micron Tech(MU.US) will step up returns from December - "over time, we expect to return 100% of our excess cash." Just crazy how much cash these companies are bathing in rn.

Obviously there are risks, mainly with positioning and sentiment towards the AI trade currently - thanks to macro/ARR fears/pacing talks/other stuff I'm probably forgetting rn.

China is ofc another risk with CXMT but I don't see how they'll be a massive problem for HBM / US hyperscaler RDIMM where they'll continue to face US restrictions + massive qualification walls imo.

But ultimately, memory fundamentals are in-tact and very strong.

Stronger than Q1.

FYI, U.S. tech was the most net sold sector this week thanks to the OpenAI ARR scare.

This breaks the trend where U.S. tech was the most net bought sector in the 4 weeks before.

Investors rotated capital away from Semis into Software...or as I like to say "defensive tech".

My prediction from ~6 weeks ago around market choppiness has proven to be correct....but I cannot emphasise enough how crucial the next few weeks are:

1. CPI on Wednesday = rates read-through.

2. Earnings for big tech and semis.

3. Midterms = potential for easing Iran tensions.

My position on where we go from here should be known by now I hope -- a rally into 2027 as we jump over the above hurdles.

But it makes a lot of sense to keep some cash to ride the incoming volatility.

Not financial advice.

With the Singapore Grand Prix in a few hours, I can't help but draw comparisons to the financial markets and the AI trade.

One slow pitstop or a miscalculated overtake can undo weeks of preparation and cost millions of dollars.

So with AI and semiconductor earnings fast approaching, it feels like we need a quick pitstop to get the race back on track.

Rates remain high. Oil prices are elevated. Fresh frontier lab ARR concerns. AI capex. Circular financing.....all create some dark clouds which have resulted in de-risked positioning and low sentiment.

Similarly, consumer confidence is low. New research from Morgan Stanley finds that net sentiment from Americans towards AI is at +14% compared with -20% for data center construction, with water requirements and environmental concerns among the biggest objections.

It therefore seems like AI policy will become an even bigger stumbling block in the years to come...

But the positive is that business fundamentals have remained strong highlighted by recent sold-out comments from $Micron Tech(MU.US) and $Lumentum(LITE.US) CEOs. Plus $Taiwan Semiconductor(TSM.US) reporting 55% YoY revenue growth as the de-facto proxy for the wider AI trade.

My beloved Chelsea won 5-1 yesterday. But conceding even just 1 goal won't be good enough for the AI trade.

With sentiment at fresh lows, I feel like all the key infrastructure companies need to report without any hiccups.

Hi @sama and @DarioAmodei,

You need to disclose your ARR metrics. Urgently.

I don't think it's fair on the CEO's of publicly listed tech companies (NVDA, AMD, MU etc.) to see their share prices nuked by Billions of $ due to incorrect/rumour-filled reporting around your financials.

It's the right thing to do. Everyone wants clarity.

Thanks,

Paradis

PS: It's also not fair on my portfolio.

Winter is coming.

The first three quarters of 2026 have been truly insane for the markets. But it's fair to say that the AI trade has cooled off a bit in recent months.

So yes, the days are getting colder and shorter. Winter is coming. But is it coming for the AI trade?

Well...other than the Leopold deleveraging in July, macro has acted as a major overhang on what would otherwise be an earnings-driven bull market.

Despite September's institutional flows into Tech keeping the market elevated like Wingardium Leviosa (any Harry Potter fans here? 🧙‍♂️), we're yet to see any huge risk be added even with institutional leverage rebounding close to 2026-average levels.

(Slight breakaway - that metric is one I will monitor closely to see if we get to July levels again)

I therefore continue to believe that there's substatial headroom for the AI trade to continue as we head into earnings.

However, investors are largely waiting for major macro hurdles to pass - specifically oil and rates, with the 10Y at multi-decade highs.

All of this matters, and is why I personally choose to ignore investors that say "macro doesn't matter" when it has dictated the course of financial markets since the dawn of time. (Quick rant over)

The most recent payrolls data definitely gave some ease to Fed hike odds, but we're not past hiking-territory yet with rates still skyhigh and CPI coming up soon.

All of this leaves a ton of lingering uncertainty that wasn't there earlier in the year even at the height of the Iran war just as the AI trade and earnings growth lit the initial fire under the market.

However, we're coming into earnings season now with many AI-related stocks (semis, semicaps etc) currently mispriced imo.

Since there's clearly underlying demand for tech stocks (see my previous flow posts)...I wonder if we'll start to see some re-ratings occur as a front-run into earnings.

Or...if the macro landscape is still a bit too treacherous.

I keep my prediction that we'll see an AI bull market heading into 2027.

Time will tell.

Hedge Funds and Asset Managers showed increased demand for Semis this week.

In fact, US Tech stocks were the *only* net bought sector in September as long buys outpaced short sales by over 2:1.

This level of buying was driven by some big catalysts this week, specifically $Micron Tech(MU.US)'s earnings and $NVIDIA(NVDA.US) buybacks.

However, prime broker data shows that Semis demand isn't absolutely skyrocketing yet and there's significant headroom for institutional positioning.

After speaking with many investors over the past week, I would say that there's some anxiety around putting more chips in on the AI trade.

I think this is due to a few factors that are currently in-play:

1. Macro shitshow - Iran, oil, rates situation.

2. Midterms.

3. Constant shifts in the AI story (tech is moving faster than investors' brains).

That said, we're coming up to Q4 earnings in Tech/Semis Land...all while forward tech/semis valuations are relatively low right now.

So I think the issue now centers around whether we can see a re-rating in Semis as we start to see positive Q4 earnings/guidance despite the current macro and midterm hurdles.

If we can get past those hurdles unscathed, then I think my prediction of a tech rally into 2027 looks highly likely as institutions size up for 2027+.

After the Leopold/SALP deleveraging in July...

Leverage being used by hedge funds and asset managers has rebounded close to 2026-highs.

This is quite evident in the AUM ratio of leveraged equity ETFs to the market cap of the underlying stocks -- which has increased in recent weeks.

Most notably for Semis (ex. memory) and Hyperscaler single stock leveraged ETFs. The ratio for Memory stocks hasn't increased in the same way, but it's still high.

I think this could be a potential headwind for stocks and a source of volatility if leverage levels become excessive once again.

Not to be a doomer, but one to keep an eye on...

Goldman Sachs forecast ~108GW of data center demand in 2030 (US-only).

Key beneficiaries:

- $Firstenergy(FE.US) & $Talen Energy(TLN.US): exposure to PJM and upside from generation/transmission opportunities.

- $Xcel Energy(XEL.US): exposure to the growing MISO market.

- $Duke Energy(DUK.US): benefits via ~15.4 GW high-confidence large load pipeline + favorable regulatory environment in the Southeast.

- $Sempra Energy(SRE.US): capture opportunities in Texas via T&D capital investment.

- IPPs like $Talen Energy(TLN.US), $Vistra(VST.US), and $NRG Energy(NRG.US) are best positioned to benefit from rising power prices and incremental data center PPAs across both PJM and ERCOT.

$Micron Tech(MU.US) Q4 earnings prediction:

Revenue: $51.9B vs. $50.0B guide

Gross Margin: 86.5% vs. ~86% guide

EPS: $32.30 vs. $31.00 guide

Maybe my modelling is too bullish? These numbers are significantly above consensus estimates...

Regardless, I aim to drop a primer in advance of Micron's earnings.

Even if Anthropic of OpenAI come out with better agents than Muse, I doubt they'll have enough impact to overcome people's desire for simplicity.

Do you think people will pause their Instagram doomscrolling to jump into Claude's app? Will people stop a WhatsApp conversation to open up a ChatGPT agent? I genuinely don't see it...(unless Ant/OAI agents are some kind of black magic).

Why wouldn't you just simply use Muse which is embedded inside the Meta apps already and can get your job done do a decent/good level? For better or worse, Meta are the absolute kings of this "usability" stuff.

Right now, I think what matters most for these companies is to get AI agents into the hands of the masses - it's important to remember that many people haven't even tried using AI yet.

So instead of monetization, I fully expect Muse's more important milestones to be around adoption/retention, plus more expansion in third-party commerce/payments integrations. This will then broaden out the workflows consumers entrust Muse with which will keep the Meta flywheel spinning.

Just got my first payout from the new Original Content Rewards program!

Reminder - 2x of the money I make from X will always be given away to charity instead of funneled into $Nebius(NBIS.US), $Intel(INTC.US) or $Sandisk(SNDK.US) stock.

Please donate to Cancer Research UK if you're able to! ❤️

I'm grateful that I get to share my thoughts/ideas...and that people take the time to read. It genuinely means a lot to me.

-- Not posting this to show off. Rather, illustrating that some good can come from this fun little side hobby. Plus continued accountability to my original pledge when I first started posting on X.

As of right now, it feels like the top three Mag 7's are $NVIDIA(NVDA.US), $Amazon(AMZN.US) and $Meta Platforms(META.US). I've personally been buying more of each every week recently.

It's also quite fascinating how quickly sentiment for Google has fallen with the stock down 15% since May.

I wonder if this is a good spot to take another look at Google though. Surely they'll release a Muse competitor soon on top of Gemini 4? At the same time, part of me feels like they've fumbled around with AI for too long now - it doesn't instill too much confidence tbh.

You guys need to understand that reaching a resolution with the Iran war is the most important thing in global markets right now. Nothing comes close, not even Muse or DRAM contract hikes! All roads lead back to Iran and gosh do they know it...

The situation is literally like that meme where the backpack is being kept on by a single thread. Roughly 20% of the world's oil supply normally flows through the Strait of Hormuz, along with a massive share of LNG (mainly from Qatar) and current transits are ~5% of the pre-war levels. Let that sink in.

Keeping the SOH shut for so long has caused the biggest oil supply shitshow in history. In fact, it's nearly 3x bigger than the 1973 oil embargo...and it's seeped into daily lives now to the point where I think it's absolutely ludicrous that they've not reached some sort of resolution....until now?!

It's no surprise to see the markets love it whenever news of a potential ceasefire comes through. It's absolutely crucial to end this quickly!

1% chance, 99% faith.

Kinda crazy how no one cares about frontier pacing now. That was ridiculous from Dario and co.

Markets have since digested the fear from last Monday quite nicely and now seem excited with $Meta Platforms(META.US) Muse which is a massive efficiency unlock + HUGEEE compute multiplier. Every lab will build something similar eventually = CPU and cloud compute trade keeps going nuts.

(Wish I had access to Muse in the UK!)

Plus there's strong rumours of OpenAI releasing new models soon -- GPT-6 Sol. And new Anthropic models in the pipeline.

Semiconductor fwd P/E has been tracking inverse oil prices over the past year. Almost perfectly.

⬜️ $SOXX Fwd P/E Multiple 🟦 Inverse oil price

Since the Hormuz shock in Feb:

As oil prices drop, semiconductor valuations rise.

As oil prices rise, semiconductor valuations drop.

AMD, TSMC and OSAT Channel Check:

1. AMD's AI GPU production (MI455/MI450) is expected to remain 100% at TSMC CoWoS in 2027. Production estimates land at a mid-point of 600k units for both SKUs next year per recent MS forecasts.

2. While ASE and Amkor still lead OSAT capacity, their capacity ramps are rumoured to be progressing slower than expected. This means that TSMC may be required to support CoWoS-L wafers for AMD's Venice CPU prodction which would further constrain internal capacity for MI455/MI450 production. Per MS, this would equate to TSMC supporting ~80k CoWoS-L for Venice production.

$AMD(AMD.US) $Taiwan Semiconductor(TSM.US) $Amkor Tech(AMKR.US) $Advanced Semiconductor Engineering(ASX.US)

$Nebius(NBIS.US) INCREASE ON-DEMAND GPU RATES EFFECTIVE OCTOBER 1.

H100: $3.85 per GPU-hour -> $4.50

H200: $4.50 per GPU-hour -> $5.40

B200: $7.15 per GPU-hour -> $8.50

B300: $7.85 per GPU-hour -> $9.50

AMD EPYC Genoa CPU and memory rates will also increase.

Intel Ice Lake CPU pricing remains unchanged.

Honestly - the FOMC decision / rate hike means nothing until the Iran war ends.

After all, the Fed can't reopen the Strait of Hormuz...

Pacing the frontier is ONLY about Anthropic's IPO.

Investors are being buttered up for fewer future model releases.

Why? Margins.

I don't think margins are "bad" per se, but there are always deeper questions on whether they last/grow longer-term.

Ultimately, shipping a new model with better capability at current price points hurts the labs. From a unit economics perspective, that's a price cut per unit of intelligence every time a new model is released. That compounds super fast at the current, highly frequent release cadences.

So slowing model releases means Anthropic can slow price cuts per unit of new intelligence. I think that this is their only clean lever left to preserve margins unless they jack up prices to unforeseen levels which no customer will readily entertain.

This is why we saw Sam and Elon agree with Dario yesterday. It only works if everyone else slows down too. As a margin lever, pacing has to be done industry-wide, and an industry-wide agreement not to compete on cadence is basically an agreement not to compete on price.

Looking at costs: the labs expense training as R&D as they go, so nothing gets amortised in the accounting sense. But the economics are the same. Every frontier model is superseded IMMEDIATELY after the newest one is released. Fewer models being shipped means they each earn more for longer, which means training cost per revenue $ drops without the labs needing to spend anything less than they already are.

This is an investor's dream lol. Again, no coincidence that this is all coming out so soon before Anthropic IPOs...

With OpenAI, I think it's the same logic but from the opposite direction. They don't have the profitability Anthropic has so it makes more sense for OAI to wait to IPO as pacing takes effect. That's why Sam agreed with Dario.

Elon only agreed with Dario because xAI are just so far behind the frontier lol. Also...Elon clearly wants the US to win above all else...so how exactly does pacing the frontier benefit the US in the race against China?

So many tech valuations hinge on forced customer retention scores. I hope the market wakes up and realises that there's absolutely zero moat in usage streaks, notification loops and decent UI. These just end up with low-quality network effects that tech companies harp on about to save face from a relative lack of innovation.

For this reason, Oura's IPO will be tragic. They'll push subscription prices even higher to satiate shareholders, which will work in the short-term, but eventually customers will fall off the edge of a cliff once they've reached the peak of the price curve. My guess is that that's coming soon based on how they've hiked prices over the years.

I genuinely feel like the world needs to wake up to these fads lol. Why do you need a ring or watch to tell you how you feel? I understand tracking steps or calories burnt, but stuff like sleep scores are probably the most useless sales feature that's been fabricated out of thin air over the past decade. Like, do people not know if they had a good nights sleep once they wake up? Who ACTUALLY cares if your sleep score was 87% in March and then 82% in April???

The answer: people who PAY for the hardware/subscription.

It's all backwards. You pay for this thing which means you're forced to care about what it tells you. Obviously there's a tiny subset of the physical elite who do care, but for most (relatively sedentary) customers, why does this stuff even matter other than sybolizing status / fashion accessory.

Whoop are another (probably worse) example with an even worse product because their marketing/hype has been the only thing that's kept them functioning. Them IPO'ing would be genuinely hilarious because the stock would pump when they announce a new sportsperson sponsorship rather than anything to do with product innovation...which would never happen btw.

I hope Apple takes the entire health hardware market by just acquiring Oura lol. That'd be huge aura (not Oura) points for John Ternus.

🚨Leopold/Situational Awareness returns?

Since last Friday, there's been a string of significant Flex calls bought in concentrated AI names: $315M total options premiums / $1.1B of delta / $5.8M of vega.

Positions:

- $Sandisk(SNDK.US): Jan 2040/2200 calls - $57M prem, $198M delta, $1.04M vega.

- $Bloom Energy(BE.US): Jan 250/310 calls - $48M prem, $140M delta, $565k vega.

- $Intel(INTC.US): Jan 105/115 calls - $48M prem, $185M delta, $870k vega.

- $Coreweave(CRWV.US): Jan 105/115 calls - $43M prem, $160M delta, $770k vega.

- $DRAM: Jan 65/70 calls - $43M prem, $173M delta, $855k vega.

- $SK Hynix(SKHY.US): Jan 190/210 calls - $39M prem, $148M delta, $710k vega.

- $AMD(AMD.US): Jan 540/580 calls - $36M prem, $193M delta, $1.08M vega.

It'll be confirmed as Leopold if we start to see some big FLEX calls in $Micron Tech(MU.US) / $Nebius(NBIS.US) / $Taiwan Semiconductor(TSM.US) soon.

Data via CBOE. NMR + GS desks aligned on "single mystery buyer" narrative...

Just some notes from $Nebius(NBIS.US) fireside chat at Goldman Sachs Tech Conference:

TLDR: Yep, it's bullish.

1. Order book extends into H1 2028 which is 2 quarters further out than at Q2 ER.

Arkady said "people are demanding tens of thousands of vGPUs and GPUs now. So we see demand today as unlimited."

Bro...what?!!

At Q2, they said that all of 2027 could be sold today. But Nebius having visibility ~1.5 years away massively de-risks their ~5GW contracted power target and their ~$25B of FY26 capex.

On demand, Arkady also said that demand visibility now extends to 24+ months (vs. 18 months previously).

This is HUGE because demand duration is a huge crux for neocloud bears (kinda makes sense why). Even $Coreweave(CRWV.US) CEO said at the GS Conference that they are "struggling to meet demand everyday."

2. "We do not pre-sell much. We are focusing on free capacity, which we will be selling later"

I think two things are happening here with Nebius

First - this is the opposite of $Coreweave(CRWV.US)'s model where their ~$104B backlog is take/pay dynamics. Nebius are instead choosing to sit on uncontracted 2027 capacity so they can sell into rising prices, kinda like how $Micron Tech(MU.US) and co. were doing pre-LTAs.

Second - keeping some spare capacity keeps room for longer-term strategic partners arriving into 2027 (which will be extremely supply-constrained). I.e. enterprise names coming via the $Palantir Tech(PLTR.US) partnership.

I think both reasons make sense, I'm fairly confident that 2027 pricing will be higher.

3. "We actually have a list of new customers that are looking for any of the older generation chips that come available."

This lines up with $Coreweave(CRWV.US) disclosing at Q2 an A100 contract priced out to 2029 and completely guts the residual-value thesis bears like Burry lean on.

4. $Shopify (SHOP.US) "used open-weight models, trained it with their own data repetitively, and they achieved the quality which is higher than they had with GPT-5 and 6."

Feels like that's the whole enterprise adoption thesis summarized....narrow domain, pvt data, repeated loop, open weights > fronteir, at a fraction of the cost.

Probably also why the token factor and the Tavily acq. matter.

So yeah, Nebius' entire infrastructure goes kinda crazy. More than just GPU rental.

5. Contract mix

Marc: 3-6 month short-term deals go out "at a multiple of the ARR per megawatt" of the core. 1-3 year medium-term deals are "the lion's share" and 5+ year hyperscaler deals were done "with the explicit intent that we are looking for the capitalization benefit."

- So the longer duration $Microsoft(MSFT.US) + $Meta Platforms(META.US) contracts are lower risk collateral for financing the build

- the opportunistic short-duration surge contracts at materially higher pricing are the top-ups.

- and the 1-3 year book with AI natives and enterprises (priced above hyperscaler deals + prepaid) are the core + fastest growing segment.

Just for a summary of the points I found most interesting / different from Q2 earnings.

Cool to also see the "the vision that Arkady has is us becoming a hyperscaler."

I've been sharing the same vision for some time now, and is why Nebius ultimately deserve to trade multiples higher than Coreweave for example.

Someone check-in on Burry!

How can you not love the $Nebius(NBIS.US) partnership with $Palantir Tech(PLTR.US)?!

Palantir naming Nebius as their "preferred sovereign AI infrastructure partner" is so cool because it effectively hands Nebius enterprise distribution on a plate.

It also diversifies future revenue away from $Microsoft(MSFT.US) and $Meta Platforms(META.US) which forms the basis of many Nebius bear theses.

Incredible growth expected for $Lumentum(LITE.US), $Coherent Corp.(COHR.US), and $Applied Optoelectronics(AAOI.US) with new GS optical transceiver forecasts.

- Global optical transceiver TAM revised up by 33% / 81% / 115% in 2026-28E.

- In 2026/27/28E: optical transceiver market to reach $68B/131B/148B.

- Driven by (1) higher rack-level AI server and ASIC AI server shipments and (2) higher estimates on the usage of optical transceivers per Nvidia GPU in rack-level AI servers.

- 800G and above segment to increase at a +69% CAGR to reach $45B/108B/130B.

- 800G/1.6T shipment volumes at 45M/33M units in 2026E, growing to 49M/71M in 2027E, with 3.2T shipment ramping up to 23M/68M in 2027/28E.

- Silicon photonics will account for 60%/80%/80% of 800G/1.6T/3.2T optical transceivers.

Just insane growth on top of what most people had forecasted a few months ago.