Latest / Deep Values: Value Investing Deep Dives / Why AI Fear Is Crushing Software Stocks?
Transcript
- 0:00Imagine walking up to like a massive, beautifully constructed
- 0:04brick building. The foundation is rock solid.
- 0:06Right. The kind of place that's built
- 0:08to last a century. Exactly.
- 0:10And the bank accounts locked in the safe inside, they are
- 0:13absolutely overflowing with cash.
- 0:15The business operating inside those walls is generating more
- 0:18revenue today than it ever has in its entire history.
- 0:21That. Sounds like a pretty incredible
- 0:22place to be. You'd think so, right?
- 0:25But suddenly everyone inside is sprinting for the exits.
- 0:29I mean, they are tossing the keys away for pennies on the
- 0:31dollar, just trapped in the state of sheer absolute panic.
- 0:35Which makes zero sense when you look at the building itself.
- 0:38Right. And why are they running?
- 0:39Because someone outside yelled that they just invented like a
- 0:42new type of wind, and that wind might eventually blow the
- 0:45building down maybe 3, maybe four years from now.
- 0:48The disconnect between the physical reality of that
- 0:50building today and the, you know, psychological terror over
- 0:54what might happen tomorrow is staggering.
- 0:57It really is. And the financial markets are
- 0:59doing exactly this right now, running from sturdy buildings.
- 1:02So our mission for today's deep dive is to figure out why,
- 1:05pulling from a great breakdown by the financial channel, New
- 1:07Money. There's a lot to unpack.
- 1:09Here for sure, but as always, we want you to be armed with the
- 1:12best tools to navigate this panic yourself.
- 1:15So before we jump in a quick note to access more advanced
- 1:18researches, applications and our latest published personal
- 1:20intelligent multi agent deep value intelligence.
- 1:24Welcome to explore and join Deep values dot AI community.
- 1:28Track total market temperature. Welcome to search and download
- 1:31iOS app Buffett indicator from Apple App Store.
- 1:34Disclaimer All analysis is based on publicly available sources
- 1:38and may be subject to revisions or differing interpretations.
- 1:41All contents from this channel is for educational and research
- 1:43purposes only and does not constitute investment advice.
- 1:47Conduct your own due diligence before making investment
- 1:49decisions. If you like to access wire
- 1:51topics, including life and investing philosophy, please
- 1:53subscribe to our sister channel, Deep Values.
- 1:55Don't forget to like this video. Much appreciate it.
- 1:58All right, So that Paradox you just described, it is playing
- 2:01out across the entire software industry.
- 2:03Yeah, financial media has started calling this moment the
- 2:06Sauce Apocalypse. You know, sauce standing for
- 2:09software as a service. It's a dramatic name, but it
- 2:13fits. I mean, we are watching absolute
- 2:15giants of the industry, companies that form the digital
- 2:19backbone of the modern economy, just getting crushed.
- 2:22We're talking massive market cap wipeouts.
- 2:24Right. Oh, absolutely.
- 2:25Reductions at 3040 sixty and in one extreme case, an 85% drop.
- 2:30OK, let's unpack this. Because the businesses
- 2:33themselves have not collapsed, not at all.
- 2:35If you look at like Salesforce, Adobe, Snowflake, Atlassian,
- 2:39these are the companies caught in this crossfire.
- 2:41Their customers aren't suddenly closing their accounts on Mass.
- 2:43Right, the churn isn't there. Exactly.
- 2:46Revenues are still growing often in in the double digits and many
- 2:48of these entities are printing free cash flow.
- 2:51So we're looking at a scenario where investors are behaving as
- 2:54if beautiful high margin software is suddenly a dead
- 2:58industry. Because the market isn't having
- 3:00a panic attack over bad balance sheets, it's having a panic
- 3:03attack over a profound shift in the future narrative.
- 3:06The AI narrative. Precisely.
- 3:08It's the looming shadow of artificial intelligence.
- 3:12Investors are desperately trying to model out what AI means for
- 3:15the future unit economics of these businesses.
- 3:18And well, the uncertainty of that math is causing them to
- 3:22just hit the eject button. Which means before we start
- 3:24looking at the individual casualties of the sell off, we
- 3:27need to understand the mechanics of the fear itself.
- 3:30I mean, why does the invention of AI suddenly make software
- 3:33look like a toxic asset? So there are two distinct
- 3:36structural fears driving this dread.
- 3:39The first is what we can call the fear of consolidation.
- 3:41Consolidation. Think of this as the Beast AI
- 3:43platform theory. Look at how a modern company
- 3:46runs its customer support today. They don't buy one tool.
- 3:49Right, they cobble things together.
- 3:50Exactly. They might pay for Zendesk to
- 3:53manage the tickets, intercom for the live chat, Salesforce
- 3:56service Cloud for the database. And maybe another tool entirely
- 3:59just for analytics. They're essentially duct taping
- 4:035 or 6 different software subscriptions together to create
- 4:06a single workflow. That's a perfect way to put it,
- 4:09the duct tape of enterprise software.
- 4:11But investors are looking at the rapid evolution of AI agents,
- 4:15these autonomous tools that can process natural language, handle
- 4:19complex ticketing, respond to customers directly, all without
- 4:23human intervention. Wow, so the fear is that instead
- 4:26of renewing 5 different specialized software
- 4:28subscriptions, a business will just subscribe to one single
- 4:32overarching beast AI platform that does everything natively?
- 4:35You got it. The duct tape is no longer
- 4:38needed and those specialized subscriptions simply evaporate.
- 4:41I mean, that is the existential threat right there, the idea of
- 4:43complete replacement. But there's a second fear too,
- 4:46right? And it isn't about the software
- 4:48being replaced at all. It's about the traditional
- 4:50software business model fundamentally breaking.
- 4:53Yes, the per head pricing trap. That's the one.
- 4:55Let's talk about the per head pricing trap.
- 4:57This is really the absolute core of Wall Street's anxiety.
- 5:02The vast majority of these software giants don't sell a
- 5:04flat rate product. Their revenue is tied directly
- 5:07to a per head pricing model. Right, you purchase a basic
- 5:10subscription for your small team or maybe an enterprise lifestyle
- 5:13that covers 500 employees. Exactly.
- 5:15So. The revenue growth of the
- 5:16software company is inextricably linked to the headcount growth
- 5:20of their customers. It is essentially A headcount
- 5:22tax. The more people you hire, the
- 5:24more you owe the software vendor.
- 5:26So consider what happens if AI actually delivers on its promise
- 5:30of making businesses vast, fastly more efficient.
- 5:33A marketing agency using Adobe or a sales team using Salesforce
- 5:37might realize that with generative AI handling the heavy
- 5:40lifting. They no longer need 100
- 5:42employees in that department. Exactly.
- 5:44They can achieve the exact same output with 50 people.
- 5:47Wait, so the moment that agency reduces its headcount to 50, the
- 5:50software vendor experiences a 50% reduction in sellable
- 5:54licenses overnight? That's the fear.
- 5:56The customer might still love the software, but the revenue
- 5:58potential for the vendor is permanently impaired.
- 6:01OK, I hear that logic, but I am going to push back on the scale
- 6:04of the panic here. Let's use an analogy.
- 6:07Think about a family gym membership.
- 6:09Just because someone goes out and buys a state-of-the-art home
- 6:12treadmill, which is the AI in the scenario, that doesn't
- 6:16automatically mean they cancel their multi person gym
- 6:18membership the very next day. Right, because there's more to
- 6:20the gym than just running. Exactly.
- 6:23Maybe they still rely on the gym, swimming pool, the
- 6:25heavyweight lifting equipment, the community classes.
- 6:28Enterprise software is incredibly sticky and multi.
- 6:32A company isn't going to instantly cancel thousands of
- 6:35licenses just because a new AI tool can write a basic e-mail.
- 6:39It's a great point. So are we confusing an eventual
- 6:42threat with an existential threat?
- 6:44The gem analogy perfectly illustrates the flaw in the
- 6:47market's current logic. See, the market operates as a
- 6:50discounting mechanism. It attempts to price in all
- 6:54future events today. Which means they tend to
- 6:56overreact. Oh, massively.
- 6:58What we are seeing is a fear of future efficiency leading to a
- 7:02massive indiscriminate sell off. Today, investors are dumping
- 7:06fundamentally sound, highly profitable businesses because
- 7:10they are terrified of a theoretical reduction in
- 7:13licenses, you know, three to five years down the line.
- 7:16They're basically pricing in a 100% gym cancellation rate based
- 7:19on the mere invention of the Treadmac.
- 7:21Exactly. That's exactly what's happening.
- 7:23Well, let's transition from the theoretical fear of this beast
- 7:25AI and look at the actual behemoths that have been swept
- 7:29up in this apocalypse. We need to start with the giants
- 7:32that possess massive deep seated protective moats around their
- 7:36businesses, specifically Salesforce and Adobe.
- 7:39To companies that are highly profitable, deeply entrenched as
- 7:42industry standards and yet. Caught in the exact same
- 7:45downward spiral as highly speculative startups.
- 7:48It's wild. It really is.
- 7:50I mean, if you are listening to this and you've ever filled out
- 7:53a form on a corporate website only to suddenly find your inbox
- 7:56flooded with sales emails aggressively trying to get you
- 7:59on a phone call. You have sales force to thank
- 8:01for that infrastructure. Yep, you really do.
- 8:04They are the ultimate undisputed customer database.
- 8:07They cater to entire sales teams, marketing departments,
- 8:11and customer support divisions globally.
- 8:13And Despite that dominance, their stock price was halved,
- 8:17slash by 50% from its peak in December of 2024.
- 8:21Which is wild when you look at their actual footprint.
- 8:23They hold anywhere from 20 to 24% of the global market share
- 8:27for customer relationship management.
- 8:28They are miles ahead of Microsoft, who sits in second
- 8:31place. Right.
- 8:32And when you lift the hood and look at the underlying
- 8:34fundamentals, the numbers completely contradict the
- 8:37markets panic. The disconnect is glaring.
- 8:40Over the past three years, Salesforce's revenue, earnings
- 8:43and free cash flow have consistently trended up into the
- 8:46right. There has been a slight
- 8:48moderation in their growth rate since 2022, sure, but.
- 8:51But a slight moderation in hyper growth does not mathematically
- 8:54justify a 50% reduction in the share price.
- 8:57Furthermore, their balance sheet is fortified.
- 8:59Their debt to equity ratio is incredibly low.
- 9:01Like roughly 2/3 of their total debt could be paid off
- 9:04immediately using just the cash and equivalents they have
- 9:07sitting on hand today. That's not a company in trouble.
- 9:09Not at all, and the narrative for Adobe is almost a carbon
- 9:13copy. Adobe, right?
- 9:14This is the creative and marketing engine of the entire
- 9:18Internet. Their stock is down over 60%
- 9:21from its 2022 peak. But look at their actual
- 9:24utility. They completely own the PDF
- 9:27workflow with Acrobat. They own the creative standard
- 9:30with the Creative Cloud Photoshop Premiere Pro
- 9:33Illustrator. Yeah, entire industries are
- 9:35built on top of their file formats.
- 9:36You can't just stop using them. Furthermore, they have massive
- 9:39marketing, commerce and analytics tools sitting on top
- 9:42of that creative base with the Adobe Experience Platform, which
- 9:46allows global brands to run campaigns and harvest data.
- 9:49And just like Salesforce, Adobe's financial results are
- 9:52pristine. Revenue, earnings, free cash
- 9:54flow, all continuing an upward trajectory.
- 9:57Their total debt could be wiped out instantly with their current
- 9:59cash and treasuries. So there is absolutely no
- 10:01fundamental warning sign in their financials today that
- 10:04suggests the company in crisis. Which brings us to something
- 10:08really interesting. What's that?
- 10:10If you run a discounted cash flow analysis, the math screams
- 10:14that these companies are incredibly undervalued.
- 10:17You project their future cash flows, discount them back to
- 10:19today's value and factor in those pristine balance sheets.
- 10:23And the current stock prices look like absolute theft.
- 10:25Right, which begs the question, if the math is so overwhelmingly
- 10:30positive, why is the market completely ignoring it?
- 10:32What's fascinating here is the sheer ruthlessness of market
- 10:36psychology overriding fundamental analysis.
- 10:39Both of these companies possess what we call a deep switching
- 10:42mode. A switching mode, which means
- 10:44once a company integrates Salesforce or Adobe into their
- 10:47daily operations, ripping it out becomes agonizingly difficult.
- 10:51Agonizing and expensive, yeah. If I'm running a 500 person
- 10:54company, moving away from Salesforce means I have to
- 10:56migrate years of sensitive customer data, retrain my entire
- 10:59staff on a totally new interface, and basically disrupt
- 11:02my revenue generating workflows for months.
- 11:05The cost of switching is simply too high.
- 11:07That switching mode protects the businesses revenue today and it
- 11:10will likely protect it tomorrow. But the stock market doesn't
- 11:13trade on today's guaranteed revenue, does it?
- 11:15No, it trades on the future narrative.
- 11:18When the broader macro narrative shifts to you know AI is going
- 11:22to destroy the per head pricing model, the market stops caring
- 11:26about how wide your mode is. The fundamental math gets shoved
- 11:29in the trunk and pure psychological momentum takes the
- 11:31wheel perfectly set. Now, a protective switching mode
- 11:34is an incredible defense mechanism if you are a mature,
- 11:38reasonably priced company, but moats offer absolutely zero
- 11:42protection if your stock price was built on an unsustainable
- 11:44fantasy to begin with. Which brings us to a totally
- 11:48different category of victims in this sell off.
- 11:51Let's examine the companies facing a brutal double whammy.
- 11:54They are dealing with the AI fears plus the incredibly
- 11:57painful transition from being a hyper growth darling to a mature
- 12:01company. This is where the valuation
- 12:03destruction gets truly historic because we have to factor in the
- 12:06euphoric 0 interest rate environment of 2021 and 2022.
- 12:10Let's focus on Atlassian. They're the premier workflow
- 12:13managers. If you work anywhere near
- 12:15software development, you know JIRA.
- 12:17Exactly. You likely use Confluence or
- 12:20Trello. This is the software that tracks
- 12:22what engineers are building, who is assigned to fix which bug,
- 12:25and how agile development projects are moving forward.
- 12:28But the bleeding here is severe. The stock is down in a
- 12:31staunching 85% from its high in late 2021. 85%.
- 12:36Yeah. And even looking at a shorter
- 12:38time frame, since February 2025 alone, the stock has plummeted
- 12:41roughly 78%. To understand a drop that
- 12:44catastrophic, we really have to look at the environment that
- 12:46created the peak. Back in 2021 and early 2022,
- 12:50money was essentially free. Investors didn't care about
- 12:53standard bottom line profitability, what we call gap
- 12:55earnings. Right, they only cared about
- 12:57total market capture. The logic of the era was to
- 13:00acquire every possible user on earth at whatever cost, assuming
- 13:04you could simply turn on the profit spigot later.
- 13:06And during that euphoria, investors gave Atlassian a $15
- 13:10billion market cap despite the company producing 0 GAAP
- 13:14earnings. 0 GAAP earnings, that's just wild.
- 13:17Investors were pricing them purely as a hyper growth
- 13:20phenomenon. Their price to sales ratio,
- 13:22which is the premium investors pay for every single dollar of
- 13:25sales the company generates, was over 40.
- 13:28Which is an astronomical premium.
- 13:30Today, as their growth is naturally slew to more mature,
- 13:33sustainable levels, that multiple has violently
- 13:36contracted that price to sales ratios collapsed from 40 down to
- 13:40just three. Ouch.
- 13:41Then you layer the AI narrative on top of that.
- 13:44Multiple contraction. Exactly.
- 13:46The specific fear targeting Elastian is that generative AI
- 13:49could drastically reduce the need for human software
- 13:52engineers. If AI agents are writing the
- 13:54boilerplate code, running the quality assurance tests, and
- 13:57fixing the basic bugs, development teams will shrink.
- 14:00And fewer human engineers means fewer human beings who require a
- 14:04JIRA license. Right, it is the exact same per
- 14:06head pricing trap we discussed earlier with marketers and
- 14:09salespeople. Just aim directly at coders.
- 14:11And we are seeing a different flavor of this hangover with
- 14:14Snowflake, the massive data organizers.
- 14:17Yes, Snowflake. The entire premise of their
- 14:19business is that if AI is this incredibly hungry machine,
- 14:23Snowflake provides the infrastructure to organize the
- 14:26data it consumes. You would think they would be
- 14:28immune to the panic. You really would.
- 14:30Yet even sitting right in the middle of the AI boom, they are
- 14:33down 60% from late 2021 and down 30% just this year.
- 14:38And remember, during the peak of the boom, their price to sales
- 14:41ratio was over 100. People were happily paying $100
- 14:46for every $1.00 of sales. That's just a recipe for
- 14:49disaster. It was, but Snowflake faces a
- 14:52uniquely different threat narrative.
- 14:54Their primary risk isn't that a standalone AI agent will replace
- 14:57their software. Right.
- 14:59The fear is that the massive cloud infrastructure giants like
- 15:02Amazon Web Services, Microsoft Azure, Google Cloud will
- 15:05eventually build their own internal highly optimized data
- 15:08organizers, rendering Snowflake redundant.
- 15:11Exactly. It's the big tech threat.
- 15:13But I have to step in again and defend the actual survival
- 15:16metrics of these businesses. Yes, neither Atlassian nor
- 15:19Snowflake have standard gap earnings.
- 15:22They are not turning a traditional bottom line profit
- 15:24yet, but look at their cash runways.
- 15:26They are massive. Right.
- 15:27Atlassian is sitting on $1.6 billion in short term cash and
- 15:32treasuries against only 1 billion in debt.
- 15:35Snowflake is in an even stronger position, holding $4 billion in
- 15:39cash and investments against 2.3 billion in debt.
- 15:42Both of these companies have years of runway to adjust their
- 15:45strategies. They are completely insulated
- 15:47from any immediate existential threat of bankruptcy.
- 15:50It feels like investors bought a Ferrari during the hypergrowth
- 15:53phase. They paid a massive premium for
- 15:55top speed. Now the car is cruising at a
- 15:58totally normal, reliable 60 mph and investors are irrationally
- 16:03angry that it requires the basic, boring maintenance of a
- 16:05regular commuter car. You don't scrap a perfectly good
- 16:08vehicle just because it's no longer breaking land.
- 16:10Speed records does dropping from hyper growth speeds to steady
- 16:14growth justifying 85% wipeout. If we connect this to the
- 16:17fundamental math, the underlying operations simply do not justify
- 16:21the severity of the sell off. Revenue growth for both these
- 16:24companies remains steady. And their free cash flow
- 16:26generation is stabilizing exactly.
- 16:28When you run those same discounted cash flow
- 16:30calculations we applied to the behemoths, you see steep
- 16:33discounts to intrinsic value here as well.
- 16:37Current estimates suggest Atlassian is trading at roughly
- 16:39a 67% discount to its true value.
- 16:42Wow. A 67% discount, yeah.
- 16:45What we are witnessing is the market aggressively, almost
- 16:48vindictively, punishing these companies for the valuation
- 16:51mistakes investors themselves made three years ago.
- 16:54The AI narrative is simply providing a convenient excuse to
- 16:58initiate the dump. OK, So what does this all mean
- 17:00for you, the listener? We have traversed the entire
- 17:02landscape of this panic. We started with the macro Fear
- 17:05of the Beast AI consolidating workflows.
- 17:08Then we looked at the highly profitable Titans with deep
- 17:10switching modes like Salesforce and Adobe.
- 17:13And we examine the hyper growth hangovers of Atlassian and
- 17:16Snowflake. So how do we synthesize all
- 17:18these moving parts? The grand take away is that
- 17:20every single layer of the software ecosystem is undergoing
- 17:23a violent repricing. The platforms that manage
- 17:26customers, the tools that empower creators, the systems
- 17:29that track workflows, and the databases that organize
- 17:32information, they are all caught in the exact same psychological
- 17:35storm. It's incredible.
- 17:37It is. But what this deep dive truly
- 17:39reveals is a total breakdown of the efficient market hypothesis.
- 17:44The idea that stock prices perfectly and rationally reflect
- 17:47all available information is falling apart before I eyes.
- 17:50Because markets are ultimately driven by human beings and,
- 17:53well, human beings are incredibly prone to emotional
- 17:56extremes. Extreme volatility and knee jerk
- 17:58reactions create profound mispricings.
- 18:01The market is currently confusing the uncertainty of a
- 18:04is future impact with guaranteed doom for the software industry.
- 18:08Historically, when the market indiscriminately throws out
- 18:10financially sound cash generating businesses because of
- 18:13a shift in the broader narrative, it creates massive
- 18:16opportunities for long term value investors who are willing
- 18:19to run the math instead of following the herd.
- 18:21It is the ultimate test of doing your own homework, which is why
- 18:24we have to reiterate the core message from our new money
- 18:27source material. These are not guaranteed red hot
- 18:31buys. We aren't suggesting you empty
- 18:34your portfolio into software stocks tomorrow morning based on
- 18:36a single deep dive. Just.
- 18:38Don't do that, right? Definitely don't.
- 18:40But if enterprise software sits firmly within your circle of
- 18:43competence, if you truly understand how switching modes
- 18:45protect revenue and how per head pricing models function, now is
- 18:49the absolute best time to dig in and do your due diligence.
- 18:52Absolutely. Then again, leveraging tools
- 18:54like the Deep Values dot AI community and tracking market
- 18:57temperature with the Buffett Indicator app are fantastic ways
- 19:01to run those analysis. It demands the discipline to
- 19:03look past the noisy headlines and focus entirely on the
- 19:06underlying mathematical reality of the business.
- 19:09But this entire entire conversation circles around an
- 19:12underlying premise that Wall Street seems to be treating as a
- 19:15spiral math equation. We spent this deep dive
- 19:18analyzing the market's biggest fear, the loss of software
- 19:21licenses. Right, the per head pricing
- 19:23trap. Exactly.
- 19:24The company downsizes from 100 employees to 50, so they
- 19:28purchase 50 fewer licenses. Wall Street is panicking over
- 19:32the lost subscription revenue. But what do those cancelled
- 19:35licenses actually represent in the real world?
- 19:39This raises an incredibly important and honestly chilling
- 19:42question. Every single one of those
- 19:45cancelled software licenses represents a real human job.
- 19:48A marketer, A software engineer, A customer service
- 19:50representative. Exactly.
- 19:52If Wall Street's fear is correct and generative AI truly does
- 19:56eliminate the need for millions of those human licenses over the
- 19:59next decade, what does that mean for the broader economy when
- 20:02those salaries disappear? It's a huge blind spot.
- 20:05The market is hyper focused on the software companies losing
- 20:08revenue, but perhaps they are entirely missing the larger
- 20:11societal earthquake. The people running out of the
- 20:13building are the ones who actually get swept away.
- 20:15Man. Are we worrying about the wrong
- 20:17apocalypse? It is something to seriously
- 20:20think about.