저자: Kevin · 외국인
요약 보기
소프트웨어 업종 바닥이 확인되고 있으며 Q3~Q4 저점 예측이 맞았다고 주장. 세일즈포스(PEG 0.9, 공정가치 $458), 팔란티어(공정가치 $222), UiPath(목표 $40~45)를 긍정적으로 봄. 토큰 비용 절감형 '하네스' 효과로 소프트웨어사 수익성이 개선되고 고객이 이익을 얻는다고 설명. 다만 세일즈포스의 가격인상 부재, 전문서비스 4% 감소, 마진 하락, 성장 둔화 가능성 등 리스크도 언급. 영상 중간에 강의/구독 할인 홍보가 다수 포함됨.
다듬은 글
Reminder, the Jhole coupon is valid until the end of today. But, friends, we need to discuss the software market bottoming out. It's finally happening . Literally in all the headlines: Wall Street is learning to love software again. The headline reads : "Forget the SaaS apocalypse, the SaaS renaissance has begun." Does anyone want "SaaS- ense"?
In any case, we are finally seeing the bottom in the software industry. If you've been following this channel for the past six months, you know that we predicted a bottom in Q3 and Q4, especially after what we called the " SpaceX effect," i.e., the sucking of money out of the economy and pumping it into Nvidia.
The top of the funnel, right? Software became so cheap that at one point, Salesforce stock, which we have in our portfolio, got a buy signal from us when it was around $165 because it had a free cash flow yield of around 13%. And Marc Benioff literally raised $26 billion in debt to buy back $25 billion in shares at a price of about $191.
We looked at this situation and decided that the PEG ratio here is around 0.9. This is the ratio of price to earnings growth. We can buy them for 165, cheaper than the price the CEO just bought them for. And although the bottom is not yet final, when it comes, the growth will be quite aggressive, given the low valuation.
Salesforce is just one example of the fact that the bottom in software is near. But it was n't just Salesforce shares that soared in price . I mean, you see that when we bought at 165 in that lower range, the price dropped even lower. It fell to 146. But since the third quarter, around July, we have seen almost perfect growth not only in Salesforce, but also in another company we have in our portfolio, Palantir.
We went in about 118 and sent a notification about it. Palantir exploded in Q3, but now you see the software market expanding. You see Atlassian recovery. You see UiPath recovery. You see a ServiceNow recovery, a small Workday recovery, although not as significant. Oh, sorry, that was Monday.com. Workday is being restored.
Axon has hit rock bottom. And even QuickBooks from Intuit, you know, is trying to break out of there. This is one of those companies that is a little behind. It is still recovering from most of the losses after the earnings report. So, what's going on? Has the software market bottomed out yet? And isn't it too late to get into software?
And what does the software do to try to minimize some of the costs that are frankly going to the software? We know about tokens because when you look at the earnings of companies, whether it's Salesforce or UiPath or Palantir, tokens are increasingly costing these companies money as they integrate them into their services.
But there are some really interesting things happening with tokenomics that, if you don't know, you should learn about. First, let's take a quick look before mentioning these changes in tokenomics. Let's quickly look at the potential valuation, even now, for Salesforce and Palantir, or even UiPath, because I know a lot of people are asking , "Kevin, why should I care if I've already missed the mark and the software market bottomed out?" Is it too late?
In my opinion, the answer is no. And if I look at our stock data tab , UIPath is at $32, that's 75% upside potential from current levels. Honestly, I think it's closer to $40–$45. I have Salesforce at a fair value of $458, which still means a doubling of the price. We'll look at some Salesforce numbers in a moment, I'll show you the positives , the early shoots, and also some red flags.
And then we have Palantir, which I think we currently value at a PEG ratio of 25, which gives a fair value of about $222. Honestly , I think this is an underestimate by 20–30% . So I think Palantir has about 50% growth potential. This is one of the tools we are still building and improving. We've only had this tool for a few months.
We are very pleased with it. But beyond that, let's focus on what's happening with this innovation in tokens. And it all comes down to what is called the " harness effect." This is really about how software companies can not only make money, but make more money. And, of course, what we love most is doing everything we can to prepare analytics and make money.
That's why you should not only sign up, but also take advantage of the Jhole coupon that expires today. Get all nine courses, every private live stream, every alpha report, and every trade alert. Tax experts say this could be a tax write-off. And, of course, many of those who are already taking the course , as well as some of the newcomers, combine both of these offers together.
Many are now simply combining home AI with this. This is our own artificial intelligence. He gets better every month. We are preparing many updates for it , but we are removing lifetime access, which means that if you subscribe today, you will keep lifetime access, but no one else will be able to subscribe to it in the future .
So, you can put this together if you are a beginner. If you are already a course participant, you can write to us at [email protected]. We will provide you with a special code so you can add this and be guaranteed to keep lifetime access. Okay, now that we've got that out of the way , let's move on to talking about the harness effect .
So, Visa reportedly spent 3 months and engaged 100 engineers to create a tool they call the Visa Vulnerability Agent Harness. It coordinates work between Anthropic, OpenAI, or open-weight models and reduces the clutter that models create through hallucinations. This gives better access to tools and data, and reduces the time it takes for software to adapt to situations.
How does it work? Well, let's turn to the research and analyze in a little more detail what they say there. So, first, think about...someone in our chat during the stream made a great analogy. They said, "Think of teaming as creating a to-do list for your AI." So these harnesses clear the context. They route the right information into the right models.
They connect tools from which you can pull data, whether it's stored workflows, access to APIs, databases, Salesforce, or whatever else. They organize all of this for you so that you become almost independent of a specific model. You become result-oriented. This is a huge advantage for software companies as they can reduce token costs while still providing equivalent value to their customers.
And this is actually very, very interesting. Look at this. Oh, sorry, I got a little distracted. Pile, that doesn't sound very good. Good luck and be careful. You're in the Coast Guard, I see. And you received a call from management to prepare an alarming suitcase with documents. Unconfirmed threat in California.
Oh, they're after Gavin. In any case, sorry. Eh, focusing on this here. Here in this fragment. Take care of yourself. This snippet here tells us, uh, this study from July, this study from July tells us this: when you use an orchestration layer, which is one of these " bundles," essentially, when you use one of these "bundles," you can ultimately reduce costs by 40%.
Your time to complete the task is reduced by 44%, although in reality it is seconds, so nominally it does not matter much to a person. And token usage is down 38%. Now, why is this important for software profitability ? And you can pause the screen and look at the gist right here if you want. I highlighted it for you.
But why is this important for software? Well, this is very important for software. Because if software companies see their margins being impacted by token spending, in other words, let's say UiPath says, "Hey, we're integrating into our RPA, robotic process automation." We are integrating tokens and are going to start charging customers for tokens.
At best, the overall software bill looks lower if you can provide them with cheaper tokens. This is relevant for Palantir. This is relevant for Salesforce. This will be relevant for UiPath, or anyone else. Uh, this is also relevant for companies, for example, that use artificial intelligence for customer support.
And if they can provide that kind of customer service with cheaper tokens with equivalent quality , then that's the key point. In fact, that is exactly the argument they are making here. Quality is on point, they actually say that the quality remains on point to complete the tasks. And as token costs have decreased, the overall quality per dollar increases by 82%.
You have to ask yourself, who benefits from this ? And that's why I'm taking out the board. Who will win? If you use an orchestration layer, does the token provider win ? Will the infrastructure layer win , like, you know, NVIDIA or Neocloud or someone else? OK. Does the software company and ultimately the customer win, who wins when you lower the cost of tokens and the quality per dollar increases by 82%?
Well, these are not infrastructure players and not token providers. In fact, their margins are getting squeezed because you are now converting tokens into a commodity in the background. Who will win? Software companies are delivering value now, and customers are winning, and customers can be other businesses, other corporations, who are benefiting from these cost savings.
Okay, great. This all makes logical sense. So, is it too late to buy software? And is everyone really going to win in software? No, not everyone is going to win. Lorraa , which Salesforce integrates with, appears to be a big winner. They have their own GPT platform that organizes documents for cases, and they are achieving incredible success in annual recurring income.
They grew from $1 million in annual recurring revenue to a hundred million. This is madness. There is a huge potential for annual recurring income, which is why we are canceling lifetime access to Homes AI. We want to move to the ARR model, to be honest, because in the future we hope to take the startup to an IPO, and ARR is suitable for this.
So what does this mean for all software companies? You still have to pick and choose. Here, for example, is Notion. The Information writes about how Notion is betting on AI, but the problem is that Sensor Tower reports a drop in mobile app usage. Is AI replacing Notion, or can it actually improve the company's performance?
And here it all depends on the specific company, the situation is quite complex and potentially problematic. So, everything is not so clear-cut. If we look into the Salesforce documents, we can see where the red flags are hiding. And keep in mind, we own Salesforce stock. We invested in Salesforce, we invested in Palantir.
We are in the black for them. We are truly delighted with them. We spent months studying these software companies with our course participants. I, of course, have an optimistic view, but I want to be clear. I will also be critical. So I'll start with some red flags. Look at this. If we look at the red flag here, we see that Salesforce is reporting that the bulk of revenue growth is driven by new business, which includes not only new customers but also additional subscriptions from existing ones.
That is, pricing was not a significant driver of revenue. This is a double-edged sword. We can say: we will attract people now and then raise prices, or we can claim that we did not want to raise prices so that customers would not cancel their subscriptions and leave us. They didn't want an exodus. So you're not raising prices, you're just trying to sell people more AI features without raising prices for existing customers.
This could be a sign of weaker pricing. This is important. In addition, they say that there is a decrease in the cost of professional services. What you can see right here is about a 4 % decline in professional services. Why? Because the demand for large-scale, multi-year transformations has decreased.
In my opinion, this is a warning sign. This is a sign that you're essentially taking this existing sponge and saying, "We're not going to raise prices because we don't want to lose them, but let's offer them more AI and squeeze this sponge." And to me, this aligns with their announcement regarding Claw Force, which allows you to manage Salesforce through Claude without an interface.
So Claude gets direct, regulated access to 360 data, Tableau, Slack, all your Salesforce junk, without leaving the chat. I call this " palantuizing" your data. You have one level where you can interact with all this data and get what you need, exactly when you need it. This is good because everything is in one place, and it reduces the time spent searching, updating, and processing data, allowing you to focus on decision-making.
Of course, there's a limit to how many decisions you actually have to make, but, you know, that's fine with me. This makes you more efficient when you need data to make such decisions. Okay, cool. So what does this mean? This means that if you are not actually growing, then do you justify these massive price increases?
Apparently, that's the case for now . This doesn't mean it will last forever. At some point, companies can, you know, if you work through Claw Force. You can end up just rebuilding the Salesforce stack, which is literally the reason all these companies were eventually sold off. Everyone thought that people would just “ vibecode” and get rid of Salesforce and they wouldn’t need it anymore.
This is not the case at the moment. But this seed of discontent still exists. The good news is that the company has become very cheap, and although they barely beat forecasts, the stock has soared, operating margin was 34.1% versus 33.6%. It was a success, but the RPO, the work performance obligation, increased by only 14%, which was a slight shortfall.
Subscriptions exceeded expectations, but only by 28 basis points . So the forecast for the third quarter also turned out to be better—11.46 billion, but with an excess of only 35 basis points. It turns out that these were not very good results in terms of expectations versus reality. I think the reason for such large candles on the charts is that they have become too cheap.
They became oversold during the SaaS apocalypse. So let's take a look at the company's valuation. Do you know what such a company is actually worth? Well, let's take a look at the income statement . So, if we look at the report, we see that gross profit is down a little bit. We went from a gross margin of 83.1%.
If you subtract one, this is a decrease from 83.1% to 81.4%. Gross margin decreased. Therefore, income is growing by only 10.8% year-on- year. This is stable, better than a cut, but their expenses increased by 18%. So, we see a decrease in profitability. And again, I say this with a tendency to be critical, really critical, of the things I encounter.
I should just be an enthusiast, right ? But I also want to see "red flags" because it's smart. So you know when it's time to exit, anyway, they get 31% of the net profit, which is very good. When you earn 31% of net income, $3.5 billion out of $11.3 billion , you deserve a PEG ratio of 2.6. And if we make an estimate by multiplying the PEG of 2.6 by the 10.5% projected EPS growth over four years and by their projected year-end EPS of 16.71.
We will get a score in the mid-400s. And, honestly, it could very well reach the mid-400s, and then jump even higher on a wave of momentum. If this happens, that's when you should a) stop investing, and b) perhaps set a trailing stop because it's time to get out. But why go out now? You are still at the very beginning of the process.
This thing is just starting to grow rapidly, right? This is true for many software companies, whether it's Atlassian. I did a lot of analysis of Atlassian: great balance, great growth. Of course, you can't invest in all companies. For example, ServiceNow is another one of them. That's not bad. But it's not my favorite.
I preferred Salesforce over ServiceNow. But does that mean I'm pessimistic about ServiceNow? No, not at all. What about Team? The same. I'm not pessimistic about them. The problem with Team is that if you analyze their historical performance, their margins have been very low and only recently have they started to increase.
Hopefully this will continue thanks to artificial intelligence, but we need to watch out for "red flags." So what other “red flags ” do we see in, uh, these software companies ? Well, let's take a look here. I wrote this under Legora Reply Zero. Zero is a platform for accountants, a financial platform, 5 million subscribers.
Agent Force is engaged in customer support on a large scale. Legora uses Salesforce for auditing, management, compliance, blah blah blah. There remains a concern that these companies will be able to optimize and replace Salesforce in the future, and the move to a “ headless” architecture for integration with Claude suggests that these problems are not going away .
And if a recession starts, you know, maybe that's the moment when companies say, " Okay, it's time to optimize costs , where can we cut?" So if I look at this, I 'll see a limiting factor , even a few. One is, of course, the interface limitation we talked about earlier: how many problems do you actually need to solve?
This is a limiting factor for AI. But another thing: at a certain point, the growth from the introduction of artificial intelligence products also slows down. If we come back here, I want you to see the growth from AI right here. It is now growing faster than the core business. The main inherited business is here.
So I'll highlight it in orange. Agent Force applications, Agent Force itself grew by only 7.6% year-over-year. 360 AI's " headless" architecture and data grew by 20%. So now you get more benefit from selling these AI products to existing customers. As long as this is growing, you can continue to see the stock rise.
That is my expectation. Once this growth starts to slow down, that will be your biggest “red flag.” So, this is the income statement. Uh , that's where the income comes from. If you look at the balance sheet, we have long-term debt, $45 billion, mostly financed at about 1.5% above Treasury bonds. This is your spread, the yield spread.
And about 25 billion was used to, uh, buy back its own shares on the dip at a price of about $191 per share. They have enough cash to pay their bills and their cash yield is still good. They currently have a yield of about 7%. And even if the shares are fully valued at around $450, they would still yield a 3.5 % cash yield, which is significantly better than the roughly 1% that Palantir yields or the less than 1% that CrowdStrike yields.
So, SaaS recovery—bottoming out in Q3-4. Yes, it's definitely happening, and that's great, I'm really happy about it. I am grateful for this . Can this continue? Is it too late to join Palantir or Salesforce? In my opinion, the answer is no. I think these companies can still grow. But if you're going to buy these stocks, you should be aware of the hidden risks.
Some of them need your attention. Of course, we've already talked about this here. This is exactly what I'm paying attention to. You may have your own reservations that you are guided by. Here's how you can invest in it while remaining cautious. It's still an investment, right? Don't get too attached to them.
Of course, if you haven't already , join us for fundamental analysis at mekevin.com. And remember, this coupon is valid until tonight. We are also removing lifetime access for this product. If you get lifetime access today, you'll keep it forever. Even if the app is n't working perfectly in your district right now, the cool thing is that with lifetime access, in 3 months when it improves......you'll have it for life.
You have already made your investment. Everything is fine. You get all the updates, which is cool and exciting. So join us at meetke.com for Alpha membership and meetreinvest.com for the Homes AI app. trick, financial analyst and .
Comments 0