Nathan (@NateB2) profile picture

Yet another computer geek

Somewhere in the Pacific NW
Joined April 2008
Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)Photo by Nathan (@NateB2)
Retail traders celebrating a 2.2x forward price-to-earnings multiple on memory producers are misreading cyclical accounting. When SanDisk trades at 2.18x, SK Hynix at 2.55x, and Micron at 2.82x while Costco commands 44x, retail investors conclude the market is irrational. Claiming that SanDisk can buy back half its market capitalization in twelve months misunderstands the physics of semiconductor capital expenditure. Three mechanical realities explain why memory trades at low single-digit multiples: The Denominator Collapse: The forward earnings in a 2x multiple are calculated against peak cyclical profitability. Throughout 2025 and 2026, memory suppliers enjoyed gross margins exceeding 50% on severe HBM and enterprise NAND shortages. However, as confirmed by TrendForce contract price forecasts flipping negative in 1Q27, pricing leverage has peaked. Memory fabrication involves massive fixed-cost depreciation on advanced cleanrooms. When wafer ASPs drop 10% to 15%, operating income experiences roughly 80% downward operating leverage. The denominator in that 2.2x multiple evaporates over four quarters. What looks like a 2x multiple today transforms into 20x or negative earnings during cyclical downshifts. The Share Buyback Fallacy: Retail models assume peak operating cash flows can be directed toward equity retirement. In heavy manufacturing, cash flow generated at cycle peaks must be preserved to fund multi-billion-dollar technology migrations. To maintain competitive bit densities, producers must continuously capitalize next-generation cleanroom tooling, including 300-layer NAND stacks and HBM4 advanced packaging. When contract prices decline and operating cash flow contracts, boards immediately suspend buyback programs to defend liquidity and debt covenants. Consumer Annuity vs Commodity Depreciation: Comparing memory producers to Costco fundamentally confuses business models. Costco trades at 44x because membership fee revenue operates like an annuity, supported by negative working capital and stable consumer demand. Memory producers manufacture an undifferentiated commodity subject to boom-bust inventory gluts. Peak cyclical earnings represent temporary windfalls, not durable compounding. The Cyclical Inversion Rule: The classic rule of memory investing remains unchanged: never buy memory equities when their price-to-earnings multiple looks historically cheap. Commodity semiconductor producers consistently print their lowest forward multiples at the cyclical peak, immediately before margin compression accelerates. Buying memory at a 2x multiple is the standard entry point of a cyclical value trap. $SNDK $MU $WDC
83
32
14
197
49,567
Oregon is raising its statewide lodging tax by 83.3% on January 1, 2027—from 1.5% to 2.75%. Because apparently, Oregon’s tourism industry needs another obstacle. At a time when Oregon ranks 49th in job growth, lawmakers are increasing the cost of hotel stays and making the state less competitive for visitors. On a $180 hotel room, that’s another $2.25 per night, before local lodging taxes. And don’t be fooled by the “nature conservation fee” label. It’s still a tax increase, with revenue distributed among multiple state agencies. Tourism is one of Oregon’s economic bright spots. Why keep piling on costs for the hotels, restaurants, workers and small businesses that depend on it? 49th in job growth, and somehow the answer is to make visiting Oregon more expensive. @GovTinaKotek @travelportland @TravelOregon
47
156
8
530
7,312
JD Vance assured us we lost the midterms because we didn't hate Israel enough, yet he's the ones dropping in the polls. Interesting.
Marco Rubio surges past JD Vance to lead 2028 Republican primary race: poll trib.al/j2m4flI
83
131
1
1,273
9,199
Nathan retweeted
I once loved Portland. I felt safe, and the city was fun. Now it's just sad, empty, and dirty. No more holiday shopping or festivities. Just heartbreaking.
Replying to @PDXFato
I first moved to Portland from SF at the beginning of 2000. At that time, Portland was beautiful, and it was a refreshing change from the Bay Area: friendly, safe, clean, affordable. Unpretentious yet endearingly quirky. By the time I left Portland proper at the end of 2017, I could scarcely recognize the city. Impossible to convey how much the city changed for the worse in less than 20 years. Now I live in Hillsboro and am active in Washington County's conservative community. But looking back at those years in Portland, it's almost as if a foreign army invaded and seized control of the city. Maybe because that's what happened.
19
12
159
3,612
I’ve worked on the Riemann Hypothesis. When the news broke that an Anthropic researcher "solved" it, I called hype. After reading the actual Lean Zulip chat, my intuition was spot on. Here is what actually happened: It’s a Logic Verifier, Not a Proof. The AI generated a 61GB, obfuscated code dump. As mathematician Joël Riou noted, it's "messy" and ignores standard Mathlib spelling. It's not a readable mathematical proof; it's a machine-verified logic tree. Lean Deserves the Credit, Not AI. As Max Nowak said in the chat: "This is not thanks to AI, this is thanks to interactive theorem provers like Lean existing." The AI brute-forced a codebase, and human-made Lean kernels checked it in 7 hours. It's Market Hype. Formal proof expert Assia Mahboubi literally called for a quarantine on such announcements, saying: "what we are contributing to is not science but market and finance." Earlier, I tested an AI on this. It built a fake React slider app and declared "99.9% Mathematical Singularity." This time, it built a 175,000-file Lean dump. The scale is different, but the hype is the same. Until a human can read it, understand it, and explain it, it’s a logic verifier outputting "True"—not a mathematical breakthrough. note my test was back in 2025 not this year #RiemannHypothesis #LeanProver #AIHype #Mathematics
80
103
26
740
107,953
Nathan retweeted
Once your mom and dad pass away, nothing will ever be the same in this life. Absolutely nothing. Value them while they are still alive.
353
1,827
71
21,828
501,023
I think a lot of people who are newer to the markets (within the last 10 years) are expecting this drop to end and the market to rebound like it always has from their perspective. The reality is this market was drastically overvalued with some of the tech valuations in batshit insane territory. This market can go lower and may just be getting priced more sanely. And that can last a while before it starts heading back up again.
BREAKING: The Volatility Index, $VIX, extends gains to +35% on the day and rises above 38 for the first time since April 2025. This puts the $VIX up +159% since October 6th.
65
6
5
144
53,169
Remember when IPOs used to be about giving retail investors a chance to invest in generation defining businesses Now the businesses get built, venture investors take all the upside and then utilize retail as the exit liquidity Incredible shift in capital markets
147
987
65
9,699
201,277
I’ve taken two companies public. The Anthropic IPO is one of the worst executions I’ve ever seen. This is especially true given the 6X demand they touted. And the primary reason is simple, greed. Insider selling was a terrible idea. And it’s not over w the staggered lock up. The secondary reason, either incompetence or lying. No way did you really have 6X over subscription and manage to break issue this quickly. Either you didn’t have quality demand or you massively misallocated. My guess is the former given the questionable ethics of this company. All in all under a huge spotlight both management and the banks failed to deliver.
55
77
4
784
76,276
Nathan retweeted
I’ve been working on a novel proof in Lean for 2 years. It is extremely tedious work. Even AI assisted. Getting down to the same three, core axioms as you build is addictive. All I will say about this “proof” released today(and plenty of others for that matter) is that it is very likely full of bollocks at least somewhere. It’s just far too easy in Lean to say vacuous things dressed up in fancy constructions — which, thanks to post training — is exactly how AIs like to work in Lean. The worst part is, there are so many definitions it’s likely impossible that any human or team of humans can be really sure that a compiled proof is the proof they think it is. What I will say is that, in my experience, Claude is often the most guilty of this vacuousness. ChatGPT has been the most reliable. We shall see.
22
13
5
167
35,398
Can we get a follow up on the status of the buying opportunity here?
History says last week's drop was a buying opportunity. On Thursday, the S&P 500 declined -4.4%, its largest daily drop since April 2025. Before that day, the index recorded 56 one-day declines of at least -4.0% since 1952. Following such drops, the S&P 500 has gained +1.0% on average over the next week. By comparison, its average weekly return across all periods has been +0.2%. Over the next month, the index has averaged a gain of +2.0%, compared with +0.8% across all periods. Over the following 3 months, the S&P 500 has averaged +6.7%, more than triple the historical average of +2.2%. History suggests the market is set for a meaningful rebound.
43
27
2
476
16,676
Everyone thinks Claude just solved the 167-year-old Riemann Hypothesis, until you actually open the PDF. Hours after Anthropic's $ANT dropped 27% from its post-IPO peak, their mathematician Levent Alpoge posted a 4-page note claiming Claude cracked it. Inside Footnote 2 of alpo. ge/remark.pdf, he admits Claude's writeup hit 250,000+ pages, which Claude cutely called a "book". Theorem 1 literally opens by noting that because real numbers are totally ordered, if a number isn't 1/2 then it's either greater or less than 1/2. He wouldn't post the PDF because he calls it a "massive unreadable travesty of 'slop'", and admits in Footnote 4 he's "paraphrasing a 'slop' summary, which might make zero sense". Instead, he dumped 14 split tar.gz Lean files and wrote that he still hasn't mastered the math around them. He even messaged Dario to clear the announcement, then added in Footnote 1: "There is an obvious cynical view, to do with the timing, but it is false." Yeah, 250,000 pages of unverified Lean slop just happened to land the exact morning ANT slid back toward its $83 IPO price.
Well, looks like @__alpoge__ & @claudeai are claiming a proof of the Riemann Hypothesis alpo.ge/remark.pdf
128
567
125
5,305
483,817
An underrated element of understanding Trump is that he just really likes and thus wants to be admired by dictatorial strongmen, which informs all his actions. Seems bad!
Trump greets Putin: "Who is he? He's my friend!" before practically cooing with the softness of a lover, "Nice to see you, nice to see you..." ​This is even embarrassing to watch. Where is the USA that stands against bloody dictatorship and defends democracy? Shame! 🤢
15
46
2
450
10,473
I’m seeing a lot of investors defending tech stocks by pointing to their forward P/E ratios. It’s a dangerous way to think about valuation, particularly in semiconductors. At the top of a cycle, earnings often look their strongest, margins are unusually high and analysts assume the good times will continue. Ironically, that’s when a stock can look its cheapest. A stock trading at 20 times earnings isn’t cheap if those earnings are about to halve. Suddenly you’re paying 40 times, and that’s before the market starts cutting the multiple too. I’ve seen this mistake repeated over decades. The real question isn’t what multiple you’re paying today. It’s how much of today’s earnings are actually sustainable. That’s the difference between looking at a valuation and understanding one.
83
30
7
270
34,763
Nathan retweeted
Anthropic 这些创始人,把虚伪二字体现的淋漓尽致 昨日刚上市,7 个人合计套现了将近 20 亿美元,平均每人 2.8 亿 更恶心的是,他们此前还承诺,要把自己持有的 Anthropic 股权的 80% 用于慈善 听起来挺伟大吧? 结果我翻了下招股书,发现后面还藏着一句: 这项承诺完全自愿,可以随时撤回,而且不具有任何法律约束力 哈哈哈哈哈哈 这帮人打着为了人类的名义,一直干着反人类的事情,还利用慈善的名义给自己脸上贴金,真他妈令人作呕 🤮🤮🤮
107
207
13
2,098
201,373
One similarity today to 1929 and 2000 is the lack of a proximate cause in earnings.
219
302
45
4,349
478,692
Nathan retweeted
Nothing went wrong, the whole plan was to rush Anthropic to IPO, pull off a pump and dump, and suck up all available liquidity left in the market while opening the Emergency Exit door for all pre-IPO investors to cash out before it’s too late This also means OpenAI is doomed
Anthropic IPO shares were predominantly allocated only to existing investors and large AMs that will place them in their index strategies and ETFs and won’t lend shares in the market for short selling. 13% to retail investors. More than 50% on institutional orders got zero allocation but I have been told those have not been spotted buying what they didn’t get in the open market. From one side GS and MS made sure $ANT shares would not be flipped, but from another side I have also been told that they heavily supported the price today to make sure the close will be well above IPO price buffering heavy insiders selling at the beginning of trading While they and the company will claim this IPO has been a success, in reality $ANT will have poor organic demand and liquidity that will remain dependent on the bookrunners for the time being. Considering $ANT unlock schedule, its price will have a hard time to perform especially after the additional forced buying that will be triggered by index inclusion in ~15 days will be exhausted. Furthermore, the reason why both SP500 and Nasdaq are heavily down today signals that there wasn’t extra liquidity in the market to accommodate $ANT IPO that it might end up being the very last one of this AI bubble. Let’s also not forget that in little over a month Dems will take control of the Congress and the administration of key states like Texas with a strong mandate from voters to curb heavily the so far wild data centers buildout and financial speculation around that.
117
911
49
5,188
126,036
have no fear $ANT bulls, you've got cathie wood on your side
Investor @CathieDWood on Anthropic: "Our confidence, here, has not been shaken at all." cnb.cx/4B5hgsP
93
53
10
903
31,269
The $ANT IPO making a few thousand people insanely rich and then proceeding to tank the entire rest of the market we all depend on would be... just exquisitely on brand.
16
58
3
1,018
22,528
It can’t be overstated how silly this is. Taxpayers will spend $30 million just to build one store in East Harlem—which will undercut nearby small businesses that operate on razor-thin margins. Just give people food assistance. This is the most inefficient way to address poverty.
Zohran Mamdani when asked about subsidizing public grocery stores: “I take a step back, and I think about when society deems something necessary for the public, then we subsidize its creation and its operation. We know that so much of what we rely on from the public sector is not being paid for at price. Whether it be the subway or a public school or a library or sanitation, we have determined this is necessary. And when I think about what is necessary for a dignified life in New York City, and anywhere frankly, it is very easy to come back to food because you cannot live without it. And when you were saying, “Explain the math to me,” in some ways, that’s the question that New Yorkers are asking themselves every day when they go to the grocery store: “Explain the math to me of how the same thing that I used to buy a couple of years ago has now skyrocketed in price.”
111
207
13
1,958
52,786