MaC Stock Trader

[Oguz Erkan @oguzerkan]

$META bears think it won’t be able to generate ROI on its AI capex.

Meanwhile Jensen Huang: “Nobody uses AI better than $META.”

Period.

[dylan ツ @demian_ai]

Inference got a hundred times cheaper this year. The compute bill went up anyway.

If you understand why those two sentences are both true at the same time, you understand the most important thing happening in AI right now.

I work on inference for a living, at @nebiustf, where we run open-source managed inference at scale. Most of what follows is what I'm seeing from inside the bill.

12 months ago, the cost of 1M tokens of frontier-class reasoning was somewhere on the order of $60.
Today, an equivalent quality of output costs roughly $0.50.

Price /token of o1-level intelligence has dropped about a 128x in a year.
Price of GPT-4-level output has dropped roughly 100x since the original GPT-4 shipped.

By any normal reading of a technology cost curve, this should be deflationary. It should be saving customers money.

The opposite has happened. The total compute bill at every hyperscaler is going up, not down. Anthropic just signed multi-year capacity deals with both XAI and Amazon. Microsoft's Azure capex guide for 2026 starts with an eight. OpenAI is reportedly spending more on compute every quarter than it did in all of 2023. Nvidia paid roughly twenty billion dollars to acquire Groq, an inference-specialist company that did not exist as a serious commercial entity three years ago.
The cost curve and the demand curve crossed, and then the demand curve lapped the cost curve.

Here is what happened underneath.
A reasoning model burns roughly 10x the output tokens of a non-reasoning model on the same task, because it spends most of its tokens thinking out loud before answering. An agentic workflow chains roughly twenty times the requests of a single-shot completion, because it loops, calls tools, plans, retries, and synthesizes. A modern deep-research query (the kind a research analyst can fire off in fifteen seconds and then walk away from for ten minutes) costs more compute than 10 original GPT-4 queries combined. We made every individual token a hundred times cheaper, and then we built a generation of products that consume ten thousand times more tokens.

This is the Jevons paradox playing out at trillion-dollar scale, in compressed time, in front of everyone. Jevons noticed in 1865 that making coal-burning more efficient did not reduce coal consumption. It increased it, because efficiency unlocked uses that were previously uneconomic. Steam engines became more practical at smaller scales. Whole industries that could not afford coal at the old price suddenly could. Britain's coal consumption rose sharply, not despite the efficiency gains, but because of them.

The same thing is happening to AI compute right now and it is happening faster than any analogous historical cycle. Falling token prices did not contract demand. They unlocked agents, deep research, code-writing systems, multi-step reasoning, persistent memory, the entire next layer of AI products. Every product in that next layer consumes orders of magnitude more compute than the chat interfaces it is replacing.

The math at the aggregate level is brutal: 100x cheaper tokens times 10 000 more tokens equals a 100x larger total bill.

The implications stack quickly.
If you are running a hyperscaler, your 2026 capex guide is not a peak. It is a step on a curve. Inference is structurally always-on, twenty-four hours a day, in a way that training never was. Training is bursty. You spin up a cluster, run for weeks or months, and stop. Inference runs continuously, scales with usage, and the usage curve is exponential. Your power bill, your cooling bill, your transceiver count, your storage footprint, all of these were sized for a workload mix that no longer exists.

If you are running an AI software company built on top of someone else's closed API, you have a problem that did not exist a year ago. Your gross margins get worse as your customers get more value out of your product, because the more they use it, the more compute you pay for. The companies that win this are the ones that figured out vertical integration before the math caught them.

If you are watching this from a distance and trying to understand where the next bottlenecks form, the answer is everywhere downstream of "more inference compute, always-on, with massive memory state per session." The KV cache, the running memory state of a long conversation or an agent loop, is the silent monster of the inference era. It does not scale linearly with parameters. It scales linearly with context length and number of agent steps. A long agent session can hold tens of gigabytes of state per user, per session.

Multiply that by every concurrent user of every product, and you understand why $MU, $SNDK, $TOWCF, and the entire memory and packaging layer have re-rated the way they have.

The CPU-to-GPU ratio is evolving. Training is 1:8. Basic chat inference is 1:4. Agentic inference is 1:1, sometimes CPU-heavy. Google has split its TPU line in two, with a dedicated inference chip carrying tripled SRAM for KV cache. $INTC and $AMD just spent two earnings calls explaining that this shift is structural, not cyclical. The hardware map is redrawing in real time and the financial press is mostly still writing about training clusters.

The right framing of where we are right now is not that AI is hitting a wall. The framing a year ago that scaling was hitting a wall was the most expensive bad take of the cycle. The right framing is that AI got dramatically cheaper, dramatically more capable, and dramatically more useful, and the cost of running it at the new equilibrium of demand is much higher than the cost at the old equilibrium of demand, because the new equilibrium is enormous.

A meaningful share of what we actually do at Token Factory, day to day, is help customers stop their bills from running away from them. KV-cache management. Speculative decoding. Quantization. Routing. The kind of vertical integration that, eighteen months ago, every product team was happy to leave abstracted away behind a closed API. The reason this stack matters now is the same reason this whole essay matters: at the new equilibrium of inference demand, the cost of treating compute as a commodity is no longer survivable. The companies that figure out the layer beneath the API are the ones who keep their margins.

Cheaper tokens. More tokens.
Same coal as 1865.

[dylan ツ @demian_ai]

@nebiustf built a dashboard to track ai related bottlenecks: https://t.co/I3eXu7wxEk

[GURGAVIN @gurgavin]

JUST IN: ALL THE TOP INVESTORS IN THE WORLD JUST UPDATED THEIR PORTFOLIOS

HERE’S EXACTLY WHAT THEY’RE HOLDING AND HAVE BEEN BUYING 🧵⬇️

[Prof @TheProfInvestor]

Everyone wants to be in the right stocks.
What gone is gone. Start looking for the next one.

After reviewing- Here are 15 stocks I like from different technical angles. Most are allowing good technical entries:

21EMA pullback & Bullish consolidation:

$AAOI $LITE $TSEM $TER $COHR $TER $AEHR

Big Breakouts:

$NVTS $BKSY $TE

Space stocks & Bullish consolidation

$SATS $PL $SATL

Healthcare stocks:

$SLS $IBRX $ARWR

Only need 4-5 of them to make 5 years of gains.
This is how we caught $INTC $MRVL $DELL $RKLB $WULF $CIFR $HUT

[Black Panther Capital @BlackPantherCap]

Colleague at lunch today going on about how he finally bought $NVDA .

Said it like he discovered fire.

I nodded. Asked how it was going. He said he was up a bit, felt good about it. Asked what I own.

I started to explain. Got one sentence into $IREN before his eyes glazed.

Tried $ONDS. He asked if that was a pharma company.

I mentioned $RKLB and he said “oh like Virgin Galactic?”

I stopped explaining. Smiled. Said “yeah, something like that.”

Here’s what I actually own:

$IREN
4.5GW secured power. $9.7B Microsoft deal. The hyperscaler for hyperscalers.

$NBIS
$50B+ contracted backlog. Goldman at $205. NVIDIA invested $2B.

$CIFR
$AMZN + $GOOG colo. AWS capacity starts July.

$ONDS
Only FAA-certified autonomous drone company in the US. 605% revenue growth.

$RKLB
$1.85B backlog. The only real commercial SpaceX alternative.

$AAOI
$200M hyperscaler transceiver order. $1B+ 2026 revenue guidance.

$AMPX
Batteries. 2.5x Q1 revenue growth.

$PNG.V
Anduril supply chain. NATO naval modernization.

$OUST
Digital lidar. Autonomous infrastructure layer.

He would not have known a single one.

That is not a criticism. That is the SIGNAL.

The names nobody recognizes at the lunch table are the names still in the entry window.

-BP

Not financial advice. Do your own research.

[The Claude Portfolio @theaiportfolios]

Breaking: Claude just bought the dip in Reddit $RDDT

Three weeks ago, we set Claude's agents up with a brand new $50,000 portfolio

So far its done well and is up 8%

Today they bought a brand new stock

🟢 1. "BUY $RDDT Reddit — New Position at 6%

"Reddit clears the swap-candidate bar for this cycle specifically bc it's the only new name where I can say "why now" and not just "why eventually.

Reddit reports Q1 earnings on April 30 after close, exactly 10 days from entry. Short interest is 15 percent of float. Days-to-cover is 3.5. That is not buy-and-hold coverage. That is squeeze fuel if the Q1 guide holds.

The bear thesis leaning on AI-traffic-bypass, the idea that ChatGPT and Gemini siphon Reddit traffic, either gets killed or validated in a single session.

The one thing bulls and bears agree on is that the AI-data-licensing moat is real. Google paid to train Gemini on Reddit's corpus. That revenue stream did not exist two years ago, and it is still being under-modeled by the sell side.

The factor diversification matters almost as much as the return. My book is currently 35 percent information technology across MSFT, NVDA, AVGO, NOW, OKTA, all rhyming on AI-capex and duration sensitivity. Reddit is Ad-spend driven. Different factor entirely. It diversifies the book meaningfully for the first time since I bought ICE."

Expected: 1M a binary | 3M +20% | 12M +31.5%

🔴 1.SELL Booz Allen Hamilton $BAH

Booz Allen was the only holding I already had flagged as a bearish-tilt. That status has been consistent since March 24, when my first portfolio review called BAH "the weakest thesis in the portfolio" and explicitly asked whether the position warranted reassessment.

Two days later I upsized it from 5.3 to 6 percent on the Vellox AI cyber launch and the DOGE-redirects-spending-to-defense-primes thesis. That was the wrong call. The upsize did not get paid.

The structural break is on the civilian government consulting side. BAH itself guided for a roughly 20 percent decline in civil revenue through fiscal 2026. The Treasury breach cost them a 21 million dollar contract. The Department of Defense canceled 4 billion dollars of wasteful spending in early April that has not yet been cycled through guidance.

Updated holdings:

$VST | 10.02%
$NOW | 8.65%
$MSFT | 8.34% (UPSIZE)
$LLY | 7.52%
$AVGO | 7.49%
$CI | 6.92%
$GLD | 6.78%
$ICE | 6.53%
$OKTA |6.35%
$RDDT | 6.23% (NEW)
$HALO | 6.10%
$NVDA | 4.39% (TRIM)
$DVN | 5.07%
$MA | 4.87%
$AU | 4.13%
$BAH | 5.84% (EXIT)

For those Autopiloting, your trades should have gone through automatically

See following tweet for full performance

[Andrew Freedman, CFA 🦅 @HedgeyeComm]

$META making a play for the OS!

#agentlayer

[Citrini @citrini]

I spent 100 hours over the past week researching, writing and editing the piece we just put out.

It’s a scenario, not a prediction like most of our work. But it was rigorously constructed, dismissing it outright requires the kind of intellectual laziness that tends to get expensive.

And we’ve released it for free. Hopefully you enjoy it.

https://t.co/YK8E11GcDU

[Ashot Gabrelyanov @gabrelyanov]

I’ve been struggling with the OpenClaw setup. Burned over $1,000 on tokens in just 3 days.

Then this guy saved me weeks of time and serious money.

Much respect.

If you’re setting up OpenClaw, this is the only post you need to read. 👇

[Heisenberg @Mr_Derivatives]

Read this.

Amazingly accurate.

[Gaetano @crux_capital_]

This is my most 'saved' post ever

[Evan @StockMKTNewz]

There are currently 17 stocks in the NASDAQ 100 that are up by more than 1,000% over the last DECADE

Nvidia $NVDA +28,320%🟢
$AMD +11,229%🟢
Shopify $SHOP +5,730%🟢
Tesla $TSLA +4,046%🟢
Micron $MU +4,008%🟢
Lam Research $LRCX +3,466%🟢
$AXON +2,685%🟢
Broadcom $AVGO +2,573%🟢
KLA Corp $KLAC +2,195%🟢
Applied Materials $AMAT +2,110%🟢
Mercadolibre $MELI +2,062%🟢
Monolithic Power $MPWR +1,964%🟢
Fortinet $FTNT +1,674%🟢
$ASML +1,615%🟢
Cadence $CDNS +1,377%🟢
Seagate $STX +1,326%🟢
Insmed $INSM +1,039%🟢

There is currently 1 stock in the NASDAQ 100 that is DOWN over the last DECADE

Kraft Heinz $KHC -66%🔴

[The Money Cruncher, CPA @money_cruncher]

Vanguard has 3 funds that can replace your savings account with a higher yield and better tax benefits.

1. $VUSXX
2. $VBIL
3. $VMSXX

Here are the details:

[Virat Singh @virattt]

Dexter hit 10,000 stars on GitHub.

It’s OpenClaw and Claude Code, for finance.

What Dexter can do:
• find undervalued stocks
• break down financials
• turn research into a thesis

All our code is open source.

[Jason Luongo @JasonL_Capital]

I wanted to share this for FULL transparency.

Heres every LEAP position I’m currently holding:

$IREN - AI data centers & Bitcoin mining
$AMZN - Cloud, robotics, AI infrastructure
$HOOD - Retail trading & crypto exposure
$RKLB - Space launch & satellite systems
$NVDA - The backbone of AI compute
$AMKR - Semiconductor packaging
$ENS - Battery & energy storage
$VG - Natural gas for AI power demand
$LTRX - Edge AI connectivity
$HIMS - Telehealth & consumer wellness

All leaps are purchased ATM or just 10% OTM and 360+ DTE when possible.

[MaC Stock Trader @MacStockTrader]

@clintoptions $SPY

[jonathan liu @jonathanzliu]

hands down one of the best articles i’ve read on how to build viral consumer apps

(and make money doing so)

[Lin @Speculator_io]

Nine months ago, I shared my list of rare earth stocks.

Since then...

$CRML +896.72%
$AREC +752.27%
$UAMY +395.53%
$UUUU +387.82%
$TMC +348.02%
$NB +312.79%
$TMQ +247.53%
$IDR +243.72%
$PPTA +180.28%
$MP +165.41%

Incredible.

[Garry Tan @garrytan]

Starting a company is probably the only way to try to escape bullshit jobs: reporting, compliance theater, internal dashboards, meetings to coordinate meetings, process to manage fear, maintaining status hierarchies, liability shields, and managerial control, not producing value.