IPO Anthropic похоронило рынок. Санта Ралли все. До каких уровней будем падать? — Transcript
Full transcript
- 0:0013 times.
- 0:01That is exactly how many times the word
- 0:03"catastrophe" appeared in the Anthropic
- 0:05report.
- 0:07And the IPO that we watched yesterday,
- 0:08while not a catastrophe, of course, was
- 0:10not that either.
- 0:12We also can't call this IPO a super
- 0:15success.
- 0:17From the initial offering price of $ 83
- 0:20, the opening happened much higher at
- 0:22the $ 113 mark.
- 0:25But it was a one-second pump, after
- 0:28which the market simply started to dump
- 0:30, and the stock price lost over 20%from
- 0:33that peak.
- 0:35And I wrote in my Telegram channel that
- 0:38the IPO would become a great pump and
- 0:40dump.
- 0:42And as I emphasize here, everything
- 0:43would be dumped very quickly, I mean,
- 0:45very quickly.
- 0:46Basically, that is exactly what happened
- 0:47. We didn't get any growth in the market
- 0:50at all. All the growth was pre-market,
- 0:52meaning they created artificial demand.
- 0:55The investment banks did a great job,
- 0:57they did good work, but the market is
- 0:58what it is, the market is unfavorable.
- 1:00That’s it, the idea has deflated. And
- 1:02accordingly, we can see the stocks
- 1:04continue their decline. And the
- 1:07three-day success, like the one SpaceX
- 1:09had, Anthropic, as we are observing, did
- 1:12not get.
- 1:13What are Anthropic shares, and why are
- 1:16we seeing what we see in the market?
- 1:19Let's figure it out in this video.
- 1:22Don't forget to like, subscribe to the
- 1:23channel, and subscribe to the Telegram
- 1:25channel, the link to which will be in
- 1:26the description.
- 1:29And also, friends, I want to remind you
- 1:31that next week the "Coding Camp Light
- 1:322.0" master group kicks off, where I
- 1:34will teach you how to write trading
- 1:36robots and automate your trading using
- 1:38artificial intelligence, as well as
- 1:40provide several ready-made solutions,
- 1:42including a trading robot, a grid
- 1:43trading robot from Veles Capital, and a
- 1:45"False Breakout" semi-automatic tool,
- 1:47which will be a bonus.
- 1:50There are still a few spots left with a
- 1:5241%discount, friends.
- 1:53The total cost is $ 500, but for the
- 1:55first 10 people, where there are still a
- 1:57couple of spots left, the cost will be $
- 2:00295.
- 2:00So if you are interested in the topic of
- 2:02algo trading, friends, the link to this
- 2:03post will also be in the description.
- 2:05And in the description, there will be my
- 2:06contact information where you can sign
- 2:08up for this training. But let's get back
- 2:11to Anthropic and look at what this
- 2:13company looks like in terms of numbers.
- 2:17Revenue is 17 billion, and
- 2:18year-over-year growth was, just think
- 2:20about it, 12 times.
- 2:22These are crazy rates.
- 2:24I don't think there are other similar
- 2:26companies with such revenue growth rates
- 2:29.
- 2:30Well, maybe some micro-caps going from $
- 2:33100,000 to $ 2 million, perhaps.
- 2:35But when we are talking about billions
- 2:37of dollars, I don't think there are any
- 2:39similar companies.
- 2:40The number of users paying $ 100,000 or
- 2:43more also increased 12-fold.
- 2:46The adjusted operating margin is 15
- 2:48percent, but the real operating margin
- 2:51is 3 percent.
- 2:54And the net loss for just the third
- 2:56quarter alone is 23.3 billion dollars.
- 3:01Also, friends, we must understand that
- 3:03the main revenue growth came from the
- 3:05release of new models.
- 3:07Accordingly, older ones become cheaper,
- 3:09and these growth rates are exclusively
- 3:11due to new models. And here, several
- 3:13questions arise. Who will need these
- 3:15models? That is, who is their target
- 3:17audience?
- 3:18Because a huge number of tasks can
- 3:20already be solved now with a
- 3:21subscription of, I don't know, 20
- 3:22dollars.
- 3:23But if you are an active user, then 100
- 3:25dollars.
- 3:27Unless you are, of course, doing
- 3:28business with video, text, and so on;
- 3:30just for normal operations, well, 20
- 3:32dollars, well, I don't know, 100 dollars
- 3:34should definitely be enough.
- 3:37Well, maybe 200.
- 3:38But the majority of the population using
- 3:41various models covers their needs with
- 3:44the most minimal subscription.
- 3:47If we are talking about the fact that we
- 3:50will have new releases while old ones
- 3:52decrease in price, allowing us to solve
- 3:55current tasks more efficiently and
- 3:57quickly, the question arises: who will
- 4:00pay for further revenue growth—firstly,
- 4:02and secondly, to what extent is
- 4:04Anthropic even capable of maintaining
- 4:07such a pace of development and creation
- 4:09of new models,
- 4:12because there is already a question of
- 4:14where to get the data to train all these
- 4:17models.
- 4:18And this crazy pace of development that
- 4:21we see is, in fact, slowly approaching
- 4:23some kind of plateau.
- 4:25And how to deal with this, AI-related
- 4:28companies do not yet know. But let's
- 4:31talk about something else as well. 90
- 4:33billion in liabilities. Besides the fact
- 4:36that the company is deeply unprofitable
- 4:38, besides the fact that the company
- 4:40doesn't know at all when it will become
- 4:42profitable and, most importantly, how—
- 4:44that is, what must happen for this
- 4:45company to become profitable.
- 4:48Well, for example, if we take Uber,
- 4:50which probably won't ever become
- 4:52profitable either, but there was a very
- 4:55clear understanding that they needed
- 4:57fully automated taxis.
- 5:00As soon as that happens, as soon as
- 5:01there are many of them, the company
- 5:03could become profitable.
- 5:05At least, that's how it was all
- 5:07explained to us.
- 5:09In the case of OpenAI, in the case of
- 5:11Anthropic, no such point is visible.
- 5:14Well, perhaps other than the creation of
- 5:16a full-fledged AGI.
- 5:19But when will that be, will it happen at
- 5:21all, and to what extent AGI in our
- 5:23understanding and what the technology
- 5:25will actually represent—how different or
- 5:27how similar that will be—the financial
- 5:30performance of these companies will
- 5:32depend on all of that.
- 5:35But for now, these are companies that
- 5:37are simply burning through hundreds of
- 5:39billions of dollars in CAPEX, with very
- 5:41dubious expectations regarding when they
- 5:44will actually start generating net cash
- 5:46flow.
- 5:48So, 90 billion dollars is the total debt
- 5:50for the coming years for Anthropic,
- 5:52which includes 121 billion dollars from
- 5:55Google—these are computing resources
- 5:57that cannot be canceled.
- 6:00110 billion dollars is Amazon, and
- 6:03around 48-49 billion dollars is
- 6:05Microsoft.
- 6:07That is to say, these are computing
- 6:09capacities that cannot be canceled, even
- 6:11if they are not used.
- 6:13Basically, it is the leasing of these
- 6:15capacities.
- 6:16Again, given the condition that the
- 6:19company is burning through 23 to 25
- 6:21billion dollars per quarter without any
- 6:24understanding of when it will actually
- 6:26become profitable.
- 6:30If we talk about the business model of
- 6:32Anthropic and companies like OpenAI in
- 6:34general, the major players in the field
- 6:36of artificial intelligence, a
- 6:38significant bet is placed, friends, on
- 6:41theft—on plain theft—that people will
- 6:43create some services or products now,
- 6:45and then we will simply dump them, make
- 6:48them better, make them cheaper, reduce
- 6:50the cost, and essentially wash away the
- 6:52value that ordinary people create using
- 6:54these companies.
- 6:58A very similar story happened with
- 7:00Wildberries, where for many years the
- 7:02company developed at the expense of
- 7:04small and medium-sized businesses, and
- 7:06then, once the infrastructure was built
- 7:09and all processes were streamlined, the
- 7:11company said: Why do we need you now if
- 7:13we can just take the most profitable
- 7:16goods, buy them directly from
- 7:17manufacturers, and sell them ourselves?
- 7:21What do we even need you for?
- 7:23And unfortunately, this is the problem
- 7:25with big corporations. That is how it
- 7:27works.
- 7:28It is not very fair to ordinary people,
- 7:30but it is the main model for large
- 7:33platforms.
- 7:35Even Apple with their App Store, despite
- 7:38not creating ready-made solutions for
- 7:40successful applications or building
- 7:42equivalents, takes a huge margin that
- 7:45eats up a significant portion of the
- 7:47profit that companies who created a game
- 7:50or an app could have kept for themselves
- 7:52.
- 7:53And that is a problem with all platforms
- 7:55. And you need to understand that
- 7:57Anthropic and OpenAI are platforms where
- 7:59you and I create some kind of
- 8:01intellectual property.
- 8:03And if our business or some idea of ours
- 8:06suddenly takes off, with almost 100%
- 8:08certainty, this solution will appear
- 8:10within these platform companies in the
- 8:13near future.
- 8:15Moreover, all the internal
- 8:16infrastructure remains with them—the
- 8:18queries we enter, the things we test,
- 8:20and so on.
- 8:22This is all intellectual property that
- 8:24stays within these companies.
- 8:26And yes, it is unethical, not very
- 8:28ethical, but that is capitalism—what can
- 8:30you do.
- 8:32And this is probably the main idea that
- 8:34, in general, will prevent these
- 8:36companies from falling too hard, because
- 8:39over a horizon of five, maybe ten years
- 8:41, a huge number of ideas will be
- 8:44onboarded by these very companies and
- 8:46platforms.
- 8:48Whether we like it or not, friends. But
- 8:50despite this, Sam Altman is sad
- 8:52somewhere, realizing that there is no
- 8:55more money left in the market in the
- 8:57near future.
- 8:58We had SpaceX, now it's Anthropic, and a
- 9:01huge amount of capital has simply been
- 9:03washed out; there's just not enough
- 9:05money in the market to cover the tens of
- 9:07billions of dollars these companies are
- 9:09absorbing.
- 9:11And they expect or plan, at least, to
- 9:13enter the market in 2027, but there is a
- 9:15possibility that this IPO will be
- 9:17delayed, especially considering the
- 9:19dynamics we are observing in the market
- 9:21right now. And in just a few seconds, we
- 9:24will talk about all of this in more
- 9:26detail. As for OpenAI's valuations, the
- 9:28last funding round was at a valuation of
- 9:31$ 157 billion, where they raised $ 6.6
- 9:34billion.
- 9:35Perhaps for a number of months, I don't
- 9:37know, days, weeks, they won't need money
- 9:39, but most likely, they will definitely
- 9:41need cash sometime within the next year
- 9:43.
- 9:44And if they don't go for an IPO, they
- 9:46will have to do another private round. I
- 9:49think OpenAI will follow this exact path
- 9:52because super-favorable conditions
- 9:54probably won't be coming anytime soon.
- 9:58Although, perhaps I'm wrong, and we will
- 10:00see some kind of hyper-positivity now.
- 10:03But at the moment, there is no
- 10:05positivity at all. And we see, friends,
- 10:07how the Nasdaq has dropped by about 20
- 10:09percent, I think. Let's measure it. I'm
- 10:11curious myself. It’s down about 15
- 10:13percent from its peak values.
- 10:16And by the looks of it, this decline is
- 10:19continuing because we see how our trend
- 10:21line was broken and how a very heavy
- 10:23support level is breaking as we record
- 10:25this video.
- 10:27And there are no signs of a quick
- 10:29recovery, at least not yet.
- 10:31Perhaps, once again, friends, some super
- 10:33news will come out in a couple of days,
- 10:34and we will head back onto a growth
- 10:36trajectory.
- 10:37But at least, neither from a fundamental
- 10:39nor a technical perspective can we see
- 10:41that happening.
- 10:42And it is very likely that this
- 10:44correction will continue to the levels
- 10:47of 22,500 to 23,000 on the Nasdaq—that
- 10:50is the next, essentially, support range
- 10:52.
- 10:53And perhaps somewhere here we will see
- 10:55some kind of pause.
- 10:56And in principle, if we see movement
- 10:59toward these values, then I would
- 11:01seriously consider buying here.
- 11:04Including Anthropic shares, because yes
- 11:07, it is a machine that burns cash.
- 11:10But again, friends, from the perspective
- 11:12of their business model, their influence
- 11:14on society, and their role in shaping
- 11:16the agenda, because we don’t really
- 11:18Google anymore, we practically don’t
- 11:21Google, we talk to Claude, or Cursor, or
- 11:23someone else, or ChatGPT.
- 11:26Tell me what happened there, tell me how
- 11:28Xi Jinping, Putin, and Trump interacted
- 11:30, what they talked about.
- 11:32And you are provided with information
- 11:35that helps you form an opinion. Meaning
- 11:37you no longer go and read various
- 11:39publications, but instead consume the
- 11:40summaries created for you by artificial
- 11:42intelligence. And through this, your
- 11:45worldview is formed.
- 11:47And all of this is very dangerous, and
- 11:49very important, and highly influential.
- 11:52And I think, friends, that the
- 11:53correction will end in the foreseeable
- 11:55future, that is, within some period of
- 11:57time.
- 11:58I am not a psychic, I don’t have a
- 12:00crystal ball. But, friends, I don’t
- 12:03expect a repeat of the 2000s, where the
- 12:05stock market fell by 50 percent.
- 12:08And I probably wouldn’t even expect a 30
- 12:10percent drop like the COVID crash.
- 12:13Maybe 20-25 percent, I think, or
- 12:15possibly even 18, which actually
- 12:17coincides with this forecast.
- 12:19So, this is a kind of historically
- 12:22healthy market correction.
- 12:24Let’s see how much we have here.
- 12:26Exactly, friends, yes, 25-27, so it
- 12:28turns out that by both historical
- 12:30standards and the technical picture,
- 12:31this is a sort of ideal scenario.
- 12:34And within the framework of this
- 12:36scenario, I would happily start buying
- 12:38something. But at the moment, there is,
- 12:39of course, no money in the market. There
- 12:41is no "risk-on" sentiment, because what
- 12:44do we have? I mean, everyone was waiting
- 12:47for the IPO, it happened. And so what?
- 12:49We see how the market fell very sharply
- 12:51yesterday.
- 12:52The Yalta 2.0 meeting, which was not
- 12:55even close to being Yalta 2.0, was just
- 12:58a pathetic imitation.
- 13:01Sure, they met, that's fine, they talked
- 13:03, exchanged pleasantries, and took some
- 13:05nice photos.
- 13:07What’s next?
- 13:08Are there any solutions? None.
- 13:11We see the war in Ukraine continuing, we
- 13:13see that there is no progress, and that
- 13:14the parties 'positions have not gotten
- 13:16any closer.
- 13:18The situation in Iran has not been
- 13:19resolved either.
- 13:21Yes, Trump is delaying those strikes,
- 13:22but the military grouping around Iran is
- 13:24growing.
- 13:25And perhaps in 2027, or maybe there will
- 13:28be a New Year’s gift from Trump. So, we
- 13:30don’t see any solution here either.
- 13:33It was all just talk, nice photos, and
- 13:35that world division that would have
- 13:37helped solve these existential issues
- 13:40and conflicts, or brought the sides a
- 13:42little closer together—none of that
- 13:44happened.
- 13:46So what are we left with? In essence, we
- 13:48are left with only the pre-New Year
- 13:50rally. But what should be the trigger
- 13:52for this rally?
- 13:54A very big question. But how do we make
- 13:56money in a market like this, friends?
- 13:58And when it's not about investments or
- 14:01long-term buys, the only option for
- 14:03earnings is some kind of short-term
- 14:05speculation.
- 14:07Well, it is simply impossible to make
- 14:08money in such a market in any other way
- 14:10. You can't just buy something and
- 14:12expect it to grow.
- 14:14Perhaps it will grow in 10 years or a
- 14:16few years, but the trends here and now
- 14:18are negative.
- 14:20I have an absolutely free service—it's
- 14:22signals from a market maker and a crypto
- 14:24screener, where you receive very hot,
- 14:26very interesting ideas throughout the
- 14:27day.
- 14:29And for example, the token—this is one
- 14:31token, as it turned out yesterday.
- 14:35It was a small discovery for me too that
- 14:37tokens with minor differences are quite
- 14:39numerous and can be traded on different
- 14:41exchanges.
- 14:43But, friends, here we received a signal
- 14:46around 25.56, meaning there was a signal
- 14:48here that this token could be bought
- 14:50after a correction.
- 14:53So, a small correction actually happened
- 14:54, and you could have earned quite well
- 14:56here. Right here, by the way, a signal
- 14:58arrived at this point that large
- 15:00liquidations had occurred here, friends
- 15:01.
- 15:02And here are the liquidations themselves
- 15:04, and here is the signal that arrived at
- 15:06the price of 0.0319, friends.
- 15:09So, well, somewhere around here, upon
- 15:12the closing of this bar, we actually
- 15:14received this signal.
- 15:16Another very interesting entry point is
- 15:21that is, these are ideas, they aren't
- 15:22exactly entry points, friends; they are
- 15:24ideas where you must find the entry
- 15:26point yourselves. And yesterday there
- 15:28were two such trades based on this
- 15:30signal. The first ticker is You can see
- 15:33the execution on the screen. It turned
- 15:35out to be a very beautiful trade. And
- 15:37there was one more trade for Lunc.
- 15:40The trade was a bit more nerve-wracking
- 15:42because they almost hit the stop-loss,
- 15:44but they didn't, and actually, both
- 15:46take-profits were hit on this position.
- 15:48Why do these entry points work so well?
- 15:53Because, friends, it's pure When I was
- 15:54just starting to build the crypto
- 15:56screener, I tried taking parameters
- 15:57head-on and watched how they worked.
- 16:00But since everything differs in terms of
- 16:02liquidity, placement on various
- 16:04exchanges, and the number of
- 16:05participants, it is logical that there
- 16:07cannot be universal values.
- 16:10So I went further and started applying
- 16:13tests, specifically for each instrument
- 16:15and each individual exchange.
- 16:20Furthermore, for every instrument and
- 16:22exchange, several parameters are taken
- 16:24and put through a multitude of tests;
- 16:27only if there is a statistical edge are
- 16:29those parameters added to the crypto
- 16:32screener, and as soon as a specific
- 16:34token hits them, you receive a signal
- 16:36here.
- 16:39In other words, the crypto screener
- 16:40isn't just a tool where you pick or
- 16:42create your own parameters.
- 16:44Today, it is a complex infrastructure on
- 16:46the level of a high-quality hedge fund.
- 16:49I won't be modest, because it really is
- 16:51a sophisticated product, and you receive
- 16:54the results in a ready-to-use format.
- 16:57This means you don't need to test a
- 16:59large number of settings yourself;
- 17:01signals come directly to you in Telegram
- 17:04, and you just look for where an entry
- 17:06point is forming to trade it.
- 17:10As of today, there is still one list
- 17:12left, friends, and that is liquidations
- 17:14, which currently have the most activity
- 17:17, because I spent a long time optimizing
- 17:19the open interest list.
- 17:22Now, open interest by itself is gone;
- 17:24there is only open interest plus
- 17:26liquidations, because I haven't been
- 17:29able to find an alpha that provides a
- 17:31statistical edge using just open
- 17:33interest—but open interest plus
- 17:35liquidations, again, once passed through
- 17:38historical data and testing, provides a
- 17:40certain alpha, and that list exists and
- 17:42works quite well.
- 17:44As for liquidations, they are still
- 17:47being optimized. They are the most
- 17:49numerous, but I think that by the end of
- 17:51the month, we will find additional
- 17:53parameters that will reduce the number
- 17:55of these signals and significantly
- 17:57increase their quality, just like with
- 17:59the other tools.
- 18:01As I said, this screener is absolutely
- 18:03free; you can connect it via the
- 18:04Telegram bot, which you will find in the
- 18:06description.
- 18:07The only condition for using this
- 18:09screener is registering on one of the
- 18:12crypto exchanges using my referral link
- 18:14and, of course, trading.
- 18:16Because yes, friends, I don't charge you
- 18:18money, although a paid option does exist
- 18:20if someone doesn't want to become a
- 18:21referral.
- 18:23There is a window where you can
- 18:25subscribe to this crypto screener for a
- 18:27fee, but for everyone else, in order for
- 18:29it to be free, registration via my
- 18:31referral link is required.
- 18:33That way, I earn from the commissions
- 18:34paid to me by the crypto exchanges.
- 18:37Everything here is honest and
- 18:38transparent, so, friends, links to
- 18:40register on the crypto exchanges will
- 18:41also be in the description of this video
- 18:43.
- 18:45In short, connect to the Telegram bot,
- 18:47go through the few steps necessary to
- 18:49set it up, and start using this tool.
- 18:53I believe that by the end of the month,
- 18:55the beta tag will be removed from this
- 18:57tool, and all lists will be working
- 18:59properly to generate real alpha.
- 19:03Just a second, as of today, we have a
- 19:05refined list for open interest plus
- 19:07liquidations, a refined list for false
- 19:10breakouts, as well as impulses and
- 19:12simple liquidations.
- 19:15This is the final list that is currently
- 19:17being worked on.
- 19:18So, there are four idea lists that are
- 19:20already refined and provide a good
- 19:21statistical advantage.
- 19:23And, generally speaking, the final fifth
- 19:25list, liquidations, is the one where the
- 19:26work is in full swing right now. Anyway
- 19:28, friends, go ahead and use it; it’s a
- 19:30great tool. And I think that in a market
- 19:32like this, it is arguably almost the
- 19:34only way to make any money at all.
- 19:36That’s all I have for now. Andrey
- 19:37Mikhailis was with you. See you later.
- 19:39Bye-bye.
About this transcript
This page contains the full transcript of IPO Anthropic похоронило рынок. Санта Ралли все. До каких уровней будем падать? by Wall Street с Андреем Михайлец, generated from the public captions YouTube serves with the video. The transcript has 3,264 words across 550 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.