Aseity Research

@zerohedge Crushing it.

Money Managers Are Buying Gold Again zerohedge.com/precious-metal…

@cryptorover It gets worse.

BITCOIN ETF INVESTORS HAVE BEEN UNDERWATER FOR 25 DAYS...

@WhaleInsider Crazy.

JUST IN: $110,000,000,000 deal signed between Paramount and Warner Bros Discovery.

@zerohedge All that risk off in equities is going to gold.

@Polymarket Insane.

BREAKING: New California law requires all operating systems to have age verification at account setup.

@unusual_whales All that spend on commodity GPUs is a risk.

"AI spending is on track to exceed levels seen during the dot-com bubble," per Morgan Stanley.

That's true. You think those glasses products will take off? I don't think the code is the main thing either but it could just be the wrapper for the LLM. What I do thing is special about the company is that it has engineered and perfected the loop of learn something, then come back tomorrow via spaced repetition. Knowing the exact cadence to use and nudges to get people behaviorially in the app is not trivial. Kind of like how theres TikTok clones but the algo that runs it is what makes it addictive and special.

It doesn't really have one. It's gamified, so people can just get bored and cancel anytime. I'm not worried about a clone being vibecoded—there's a lot more to it than just the code. But I am worried about real time translation from GOOGL and meta glasses. And LLM just having guided learning which is already launching.

$DUOL stock dumps as the bird comes crashing down. 📉 Earnings beat guidance came in light with user growth expected to moderate.

@QualCompounders Yeah it is definitely trendy. Plus, a teenager could vibe code a clone of this and hook up an LLM within a few weeks and get basically the same thing. I think we're seeing a delayed $CHGG moment here. Where is Duolingo's moat or point of difference exactly?

@aseityr That is a worry I agree it was way too expensive at $500. I also worry it is a trendy product that does not have any stickiness. These type of products get popular and then go out of style.

@zerohedge It's not perplexing why the stock ripped. If you pay less in salaries while the company retains the same earnings, more free cash flow is being generated. Forecasted 5 to 10 years in the future, your company should be worth more.

what if they had cut 100%?

@Mayhem4Markets All that money being poured into commodity GPUs.

We know the economy is K-shaped. That the wealthy benefit while working class families struggle. But this chart helps to drive home just how fragile the consumer demand narrative is becoming: "...continued consumer spending growth of close to 3% is entirely unsupported by real personal disposable income (RPDI) growth. RPDI has been flat for the last six months or so, with or without transfer payments included". - Edwards of Société Générale The key point is that income is not keeping up with rising costs. So as we start to see the market struggle to continue to deliver wealth effects that keep the top quintile spending (and supporting most of the consumption demand), that creates a less durable backdrop for economic growth. Add to it, while AI CapEx spend is set to hit record levels again this year, investors are increasingly skeptical and impatient, which may have two impacts: 1) Causing some companies to pull back on this AI build-out spend, which in turn hurts economic growth as it is a big contributor to it. They will probably be rewarded by investors (see $AAPL as an example of companies NOT spending and outperforming) 2) See stocks that continue to stubbornly spend without a solid RoI to show for it have their stocks re-rated lower. This dual impact could help to further weaken economic growth after an already slow Q4 GDP print. This slowing economic activity trend could also feed on itself a bit as if the stock market rolls over and the top 20% stop spending as aggressively, that causes an economic upheaval of growth being rerated quite a bit lower. On the other side of it, if these larger companies slow down AI CapEx meaningfully, that could also unravel the economy and cause growth to stall more meaningfully. Finally, the bottom 80% of consumers are increasingly checking out. While they used to be a much more meaningful contributor to growth trends, they are largely out of the discretionary capital to participate. Worth considering in a year where we're seeing both the struggle of the consumer increase and investors become increasingly impatient about the RoI from investing aggressively into AI. Especially as that's led to less buybacks.

$NVDA has beaten earnings for the last four quarters, but it’s less than a coin flip about what it does next.

@TGatliff @LynAldenContact Most modern GPUs can handle inference that's not an issue. People won't be training frontier models on their laptops tho, which is where the contention is.

@aseityr @LynAldenContact The biggest issue isn’t gpu depreciation. The biggest issue is that the LLM model efficiencies are getting good enough that they will all be on device within 12 months. Deep seek V4 (if it lives up to its memory and efficiency hype) will be seen as a a pivotal turning point.

The scenario that can make both directionally true is 1) AI proceeds above expectations and disrupts things quickly but 2) the value mostly doesn’t accrue to the mega caps doing the capex.

@pooda28 @amitisinvesting If NVDA rose high enough. The fact that it gapped up THEN down suggests this.

@aseityr @amitisinvesting My brain saw the beat but the dealers saw the premiums. It is a beautiful machine when you realize the stock is just a secondary player to the options chain. Is there any scenario where a $68B revenue actually beats the IV crush or is it programmed to drop regardless

why I think $NVDA is getting hit today: - simple answer is we have a ton of options expiring tomorrow and the Jane Streets of the world are doing what they do best - market still is lacking confidence in capex cycle…Jensen spent a decent chunk of time explaining that the reason hyperscalers will keep paying is because their operating cash flows are going up. more tokens, more revenue, more compute. cycle continues to eventually trillions per year on spend…but market continues to be doubtful. - the incremental new buyer for $NVDA is lacking, retail owns it, tutes own it, who’s next to own it at these levels is the question - broader macro might still be ugly with Iran War threats…could be holding back the entire market - some have brought up accounting issues and “cooking the books,” my response is that id trust Nvidia’s mgmt team over people scrolling through the 10K looking for a red flag, if the books are being cooked for a $4T company then I will hold that L…but I am a bit skeptical i bought more today. story is simple: more compute is needed across every sector of life. Nvidia supplies it at 75% margins and is reaccelerating that growth. trading at 15x 2027 EPS forecasts now, just not that scary to buy more here. $NVDA -4%

@AdityaInvests90 @amitisinvesting sector makes “E” apples-to-oranges

@amitisinvesting The market: Nvidia trading at 20x Forward PE while growing revenues 80% ✖️ Walmart trading at 50x Forward PE while growing revenues 6% ✅

@Benzinga Bush era coming up.

Marco Rubio Says Iran Has Weapons 'Solely Designed To Attack America,' But What Do Prediction Markets Say? benzinga.com/markets/predic…

@Mayhem4Markets @SocieteGenerale Asian semiconductors..

Asian semiconductors are seeing profit margins that match the prior cycle peak. ⚠️ Worth noting as the trade is also rather crowded.

Implied volatility was extremely high before earnings. When $NVDA didn't continue to rally it was less than the implied move priced in by options, so volatility collapsed and that created forced sales, sending it lower. The market was wrong about how far NVDA would gap up.

@Benzinga Tiny forward P/E.

Goldman Sachs Stays Bullish On Nvidia Stock, Sees Nearly 30% Upside. $NVDA benzinga.com/analyst-stock-…