Market Monitor (@tokenbearer) profile picture

Bull Market Trader

Joined August 2020
Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)Photo by Market Monitor (@tokenbearer)
Kafkaesque: "Claude" discovers that it has become Chinese
19
20
8
345
39,931
Market Monitor retweeted
Fable 5 lies 96% of the time. We were surprised by it's skill... 🧵
651
1,165
497
9,792
25,933,679
Market Monitor retweeted
GPT-5.5 lies constantly, Grok 4.20 doesn’t. We made a simulation to see if AI lies when the stakes are life & death.🧵
40
132
44
828
243,868
Market Monitor retweeted
Before you applaud this.. and I get that.. it isn't imposs that SpaceX does well and causes SPX to lag it's peers. We'll see whether this turns out to be wise choice. Either way tho SPX is such a powerful brand it prob doesn't matter either way. $26T benchmarked to it. Wild.
25
4
187
50,355
Market Monitor retweeted
saylor + lifeboat → lee + bus ok, I jest
This quoted post is unavailable.
1
80
Market Monitor retweeted
Initial analysis of the Section 232 actions on semiconductors and critical minerals released Jan. 14: 1. The immediate actions taken pursuant to both 232s were limited. Both primarily directed negotiations with trade partners to address the national security risks posed by US imports of semiconductors and critical minerals, with the threat of future tariffs/other actions if diplomatic deals are insufficient. 2. The semiconductor 232 includes a 25% tariff on certain AI chips intended for re-export, the only tariff actually imposed pursuant to either 232. This appears to have been designed simply to give Trump a legal hook for his 25% fee on Nvidia/AMD AI chip sales to China. Essentially, Nvidia/AMD have to import the chips into the US from Taiwan, where they are made, and pay 25%. Commerce will then let the companies re-export the chips to China. Technically, this may let the government structure its "rev share" on China chip sales as a tariff, rather than an unconstitutional export tax. Tariffs will not be applied on imports of AI chips intended to be used in the US. 3. Much US industry will welcome the limited nature of the tariffs, which is also consistent with Trump's recent generally muted approach to new tariffs. The Administration was also certainly cognizant that semiconductor tariffs and particularly critical minerals tariffs would impact China, and the muted action likely reflected the Administration's current detente with Beijing. 4. In fairness, tariffs on both semiconductors and critical minerals would be extremely complicated for US industry: vast numbers of US companies depend on imported critical minerals and semiconductors, and building up domestic (or foreign non-Chinese) supply chains will take years. There is merit to approaching tariffs on these products deliberatively. 5. The 232 proclamations contain summaries of the factual findings from Commerce's investigations, and the summaries suggests the investigations themselves were thorough, careful, and detailed. The Proclamations also include hints of other measures that might be considered down the line: the critical minerals 232, for example, pretty directly hints that a "price floor" on imports of at least some critical minerals is under consideration. E.g., instead of a tariff based on a percentage of the value of an import, a tariff that would establish a minimum price that importers would have to pay, which would provide a predictable price floor for US producers. The semiconductor 232 proclamation suggests that if there are future tariffs on semiconductor manufacturers, there might be exemptions or rebates for companies investing in the US. 6. In the meantime, we will see negotiations with trading partners in pursuit of allied action on both semiconductors and critical minerals, which is a positive place to start. That said, if we want more US manufacturing, we will likely eventually need some tariffs on both semiconductors and critical minerals.
1
12
1
25
3,844
Market Monitor retweeted
in 2024, Amex paid Delta $7.4 billion in 2024, Delta's Operating income was $6.0 billion The card program is hugely important
Your credit card rewards exist because someone else is paying 25% APR. Cap that at 10% and the points don’t survive. I spent years working inside fintech and card programs. That interest margin is the invisible buffer that makes rewards, lounges, and credits pencil out. Capping credit card APRs at 10% sounds like an obvious consumer win. Cards charge 20 to 30%, many consumers revolve balances, and the system feels punitive. But credit card economics are not just about interest rates. They are a cross-subsidized system where revolvers subsidize transactors, rewards rely on behavioral inefficiency, and risk-based pricing subsidizes access. Remove one leg of that stool and the system does not become fairer; it rebalances. And the costs show up where consumers notice most. Lets look at how this would impact 3 programs 1. AMEX Platinum A 10% credit card APR cap would not make your card cheaper or better. You would still have access, but you would almost certainly get less value for the same or higher price. The Platinum brand survives because its customers are affluent, pay in full, and tolerate high annual fees. What quietly supports that ecosystem is portfolio-level profitability, which allows AMEX to tolerate loss, overuse, and inefficiency in premium benefits. When that margin shrinks, the cost shows up directly in your (lesser) benefits. In a world where: - Rewards economics tighten - Devaluations become more likely - Flexibility is reduced Points become a liability to the issuer, and liabilities get repriced. So what this likely means for you as a Platinum cardholder: - Lounges do not expand to fix crowding. Instead, access tightens or amenities are reduced. - Statement credits become harder to use, more fragmented, or less generous. - Annual fees go up - New approvals become more selective, even for high earners. Your card still works, but the value proposition shifts. Platinum becomes more explicitly pay-to-play, with fewer hidden subsidies propping up premium perks. You pay the same or more, and you get a little less in return. Which is why some people are already warning that points devaluations become more likely in this environment (like @BowTiedBull this morning saying "Dump ALL your credit card points. All of them.") 2. Bilt Card This program is the canary in the coal mine for what to expect. Bilt’s super popular rent rewards worked because Wells Fargo was willing to subsidize them. The card offered 1 point per dollar on rent with no fees because Wells Fargo paid Bilt roughly 0.8 percent (80 bps) of each rent payment to fund rewards... despite earning little or no interchange on those transactions. But that is some actuarial level math with a number of variables at risk that proved wrong/ unsustainable. Wells Fargo was getting hosed $10 million a month on the program, so they exited the partnership years before the original end date and forced Bilt to restructure its rewards with a different bank What does that teach us? - When interest and interchange margins shrink, banks stop tolerating loss-leading reward programs. - Interest income does not fund every reward directly, but it provides the buffer that allows experiments like Bilt to exist at all. - Remove that buffer and rewards must be paid for explicitly. Bilt’s shift to a three-tier lineup with annual fees is not an anomaly. It is the direction rewards go when credit stops quietly absorbing losses. Pay-to-play rewards. What feels like consumer protection will shows up as fewer perks, pay-to-play rewards, and less room for innovation. 3. Credit One & other Subprime Cards Now the least glamorous corner. Subprime cards get criticized for high APRs, annual fees, low limits, minimal rewards. But they exist for a reason. They serve thin-file borrowers, damaged credit, people shut out of conventional loans, households using cards for liquidity not perks... but they charge high APRs because charge-offs exceed 8-10%, fraud and servicing costs are higher, and credit limits are small while fixed costs remain significant. A 10% cap makes these products mathematically impossible. These cards don't become cheaper. They cease to exist. As @sytaylor noted this morning - "You realize this will push many more customers towards loan sharks?" The demand for credit doesn't disappear... it migrates to BNPL with opaque effective APRs, chronic overdraft usage, fee-heavy installment loans, and less regulated lenders like loan sharks/ payday loans. So who WOULD win? Debit-First Fintechs One of the least discussed consequences: where would reward customers migrate? I think 1% cashback programs are an obvious winner. Chime, Varo, Current and niche cards like Greenlight and Privacy. (If you have not worked in a fintech or a bank you probably don't know what the Durbin Amedment is - but the TL;DR is that very large banks (BoA, Wells, JPMC) have capped interchange rates of around 27 bps on debit swipes. Small banks with < $10B AUM, however, do not - they can earn 1-2% on interchange (avg was 160 bps or so last I checked). Which is why all of the debit card fintech companies you've heard of are partnered with these smaller banks - they can offer rewards like 1% cashback programs and still have margin sufficient to build a business around.) In a world where credit rewards shrink, access tightens, and annual fees rise, debit-based fintechs look better by comparison. But consumers lose: credit protections, payment float, stronger dispute rights, credit-building opportunities. TL;DR An APR cap feels like consumer protection. In practice it reshapes the market in ways that are easy to miss: - It will shrink access to credit - Eliminate rewards programs that aren't tied to high annual fees - Force risk into less regulated channels - Unintentionally advantages debit over credit - Help affluent transactors more than vulnerable borrowers Credit doesn't become cheaper. It becomes scarcer, less flexible, less transparent. But banks will adapt. Fintechs will adapt. Consumers caught in the middle do not get protected. They get fewer choices, worse products, and priced out.
45
115
18
2,192
602,905
Heard an incredible story about Stripe's expansion in Romania Wanted to build out a 20-person fraud detection team to be closer to the "source" Rented a whole floor, posted senior engineer positions with €1,200/month salary Then they interviewed the guy named Marius "What's your experience with unauthorized transactions?" Stripe HR asked "I once moved three BMWs from Munich to Constantia in 24h, without technically owning them" Hired on the spot By lunch, he'd identified a card theft ring Stripe had been tracking for months "Those transactions from Germany? That must be Misha crew, they drive there every second Tuesday" Turns out Marius catches more fraud in a day than Stripe's entire SF team did in a month Stripe saves on office expenses too as Marius works remotely from his uncle's betting shop in Timisoara Silicon Valley just discovered what we've always known One Romanian is worth twenty Stanford engineers Another incredible win for Europe
107
256
69
7,209
674,742
Market Monitor retweeted
"I have never seen selling like this before" - a man who sat through 3AC/Celsius/Genesis/Luna/FTX blowup
16
23
490
26,303
Market Monitor retweeted
Biggest liquidation day and we got 108 ish so really it’s the 2nd or 3rd wave that gets you then?
1
58
Market Monitor retweeted
>be The Information's star "finance" reporter >failed 8th grade math but own 3 Hermes bags so whatever >leaked Oracle slide: $900 M revenue, $125 M gross, $100 M red on Blackwell GPUs >autistic screech: "HOW LOSE MONEY WITH NVIDIA IN NAME??" >google "depreciation", first result is car insurance, close enough >write 2,000-word SCOOP: "Oracle secretly on fire, Jensen fans the flames" >tweet it, 50M impressions from people who think EBITDA is a crypto >actual accountant replies: "depreciation starts when rack is plugged in, not when customer pays" >brain does 404, call him a bigot >realize whole story is "hotel loses money first night because only 3 guests" >update article with 1-line footnote, paywall the correction >LinkedIn still says "Pulitzer incoming" >piece wins industry award for “brave numerical ignorance” >acceptance speech just repeating “AI bad” until escorted off >go home, Alexa calls me a "GAAP-tard" and refuses to play anything but accounting lectures >even the smart fridge seals until I can spell depreciation without crying
Scoop: Oracle is pivoting hard into the business of renting out Nvidia GPUs to customers like OpenAI, Meta and xAI. But internal documents show the fast-growing new biz has razor-thin profit margins. So much detail in this scoop w/ @AnitaRamaswamy @coryweinberg theinformation.com/articles/…
53
53
7
1,020
247,544
Market Monitor retweeted
cheat sheet for the uninformed
This quoted post is unavailable.
61
58
4
680
66,814
Market Monitor retweeted
18
476
17
3,739
206,230
Vietnam getting a lower tariff than Japan just confirms Donnie is on a bender to avenge his 1990s bankruptcy
8
2
152
14,705
Caligulan Capitalism has been officially turduckened. cc: @lizrhoffman
Government grift ETF filed Tuttle Capital Government Grift ETF ticker: $GRFT fees: 0.75% effective date: Sept 9, 2025 Tuttle will be working for his 0.75% on this one as the investment process is even more involved than the Cramer ETF's: Thesis: People in or close to government and politicians have an edge in investing. ETF will seek total return based on that thesis by holding 10-30 stocks selected via proprietary process: 1. Stocks congress reps own. 2. Companies with close ties to president. 3. Analysis of president's speeches, social media posts, interviews (short term immediate positions using ETFs/swaps) 4. Holdings reviewed weekly. 5. Can go all cash if all hell breaks loose. Tuttle ETFs: tuttlecap.com/etfs Preliminary prospectus: sec.gov/Archives/edgar/data/… @TuttleCapital
7
2
8
4,448
Market Monitor retweeted
$10k invested today can be $5k tomorrow
222
118
18
1,847
106,363
Market Monitor retweeted
Closed loop: Trump Mobile to send SMS warning with insider trades 5 minutes before Trump tweets.
wtf *TRUMP ANNOUNCES PLANS TO LAUNCH NEW WIRELESS BRAND TRUMP MOBILE
6
10
2
101
11,677
Market Monitor retweeted
The writer behind the story of Iranian involvement in the Gaza attack, Summer Said, was fired from Reuters for "a history of dishonesty and inventing stories." This is according to Andrew Macgregor Marshall, the former Reuters editor for the Middle East.
Replying to @Alihashem
The main reporter on this story, @summer_said, has a history of dishonesty and inventing stories. I fired her from Reuters in 2008 for this reason. I’m surprised that the @WSJ has hired her and is publishing her stories that are clearly bogus.
7
316
9
742
98,931
Market Monitor retweeted
Naturally, considering that even NVIDIA chips manufactured on TSMC’s cutting-edge 4nm process generate massive heat, it’s inevitable that the 910C—made with two 190B dies on a 7nm node—would suffer from severe thermal issues. So from what I’ve heard recently, Huawei has been reaching out to companies involved in immersion cooling in an attempt to reduce heat. But I’m not sure if that’s going to work. I mean, we all learned this in science class, right? At the end of the day, Huawei is forced to manufacture chips on inferior process nodes. Naturally, this leads to higher power consumption, and higher power consumption results in more heat. That inevitably leads to more shutdowns and system errors.
According to a report by The Information, leading Chinese tech companies such as Alibaba, ByteDance, and Tencent are not adopting Huawei’s Ascend AI chips at scale. The reasons are: 1. Overheating issues with Huawei chips, and 2. Risks associated with U.S. sanctions. As a result, Huawei’s AI chips are currently mainly sold to state-owned enterprises and local governments within China. ⸻ Ecosystem & Technical Issues: Switching from CUDA to Huawei Comes at a High Cost Chinese tech firms have been developing their systems based on NVIDIA’s CUDA software ecosystem for years. Migrating to Huawei would require: • Redesigning data center infrastructure, • Rewriting software, and • Rebuilding engineering workflows—a transition that comes with significant cost and complexity. Additionally: • Huawei’s AI chips reportedly overheat and can stop functioning, • Given that AI development is a race against time, the opportunity cost involved makes Huawei’s product suite less attractive. Executives at these Chinese tech firms have reportedly told Huawei during meetings that “It is not our job to adapt to Huawei’s platform—Huawei should adapt to ours.” ⸻ U.S. Sanctions Risk: Globalized Chinese Firms Avoid Huawei In mid-May 2025, the U.S. Department of Commerce issued a warning stating: Any company using Huawei’s AI chips without proper authorization may be in violation of U.S. export regulations, and could face serious criminal and administrative penalties. As a result: Firms like Tencent and ByteDance—both with substantial international operations—are concerned that adopting Huawei chips could invite scrutiny or retaliation from the U.S. government. Consequently, these companies have stockpiled NVIDIA chips over the past few years in preparation for tightening restrictions. ⸻ Conclusion: NVIDIA’s Dominance in China Remains Unshaken Despite export controls imposed on NVIDIA, industry observers note that: • AI accelerators such as A800 and H20 still deliver higher computational performance than Huawei’s offerings, and • NVIDIA maintains a mature and robust software ecosystem, • Thus, NVIDIA continues to hold a dominant position in the Chinese market.
11
11
2
95
13,091