#298: What Software Stocks Should You Buy on the Dip?

0:00Mike:When you want your spring break to feel like... And your kid's pool day to feel like... And your hotel bed to feel like... Ooh, and room service to feel like... Because at Hilton, hospitality feels like...

0:19Emmet:Your cabana's ready. Would you like fresh towels?

0:21Mike:It matters where you stay. Book now at Hilton.com. Hilton. For this day. This episode is brought to you by Indeed. Stop waiting around for the perfect candidate. Instead, use Indeed Sponsored Jobs to find the right people with the right skills fast. It's a simple way to make sure your listing is the first candidate seat. According to Indeed data, Sponsored Jobs have four times more applicants than non-sponsored jobs. So go build your dream team today with Indeed. Get a$75 sponsored job credit at indeed.com slash podcast. Terms and conditions apply. There are some great businesses on sale, but there are also ones that maybe are facing a systemic decline in front of them.

1:11Emmet:Hello folks and welcome to Stock Club, the podcast where we find and discuss stocks that anyone can buy, many of which we believe will go on to grow your wealth. We're approaching our 300th episode, and this is number 298, I believe. And to celebrate, we have something special. Well, we have 30 of them and no more. February's Horizon stock pitch, which is coming on the 26th of February, is a departure from my usual MO because to find it, I started with an almost unheard of academic paper that analyzed nearly 500 US stocks that went on to grow tenfold. And it asked and answered the question, what actually mattered before that tenfold growth?

2:03Emmet:So I ran that research paper through three custom-built AI systems in parallel, they don't touch each other, to screen for companies exhibiting the same pre-multi-bagger characteristics. And the results knocked my socks off because only one name, only one company was common to all three AI systems. And it's an almost unheard of business. I'd imagine less than 1 % of you have heard of it. and I have been considering pitching this stock for the longest time which Mike can attest to so I'm doing it I'm going to pitch the stock and at the end of this month um I will also uh link the published paper that I referenced in Horizon so we have a great offer because no rational person will want to miss this our offer is 50 % off that's$1 ,500 off Horizon and you get nexus one two and three included so hold the line that is our three international stock picking services which commenced three years ago for nexus one last year was nexus two and nexus three is currently in flow and is being is finding some of the greatest investments international investments of our time and uh we are going to give that bundle to you or to 30 of you for just$1 ,500.

3:31Emmet:And of course, you will learn the company that I'm going to pitch that three AI systems all identified exhibited the attributes from the academic paper, which is going to be published as well in Horizon. So Horizon, by the way, is now quarterly enrollment that we've started. So it's a short window, as they say, and you can secure this deal for pennies on the pound or indeed cents on the dollar okay i'm with the show mike how are you this week i'm good i'm good i'm

4:00Mike:curious now after that pitch um oh yeah yeah but uh no tipping away uh you'll be happy to know that my wall street has a new fan uh it's what our 298th episode my mother is finally listening to the podcast so that's all good yeah she texted me uh she texted me this portion fellow really knows this stuff so there you go was that's brilliant that was the other thing as well it wasn't even

4:23Emmet:podcast i was on so hello miss omani um well i'd say that's it she's like i listen to ones my

4:31Mike:young philis and don sure they're the best ones she might learn something from them but uh oh good staring out at a very rainy dreary day here but i think that's pretty much universal for anyone

4:42Emmet:listening right now oh yeah so you're in the south spain i'm in the south of dublin and it is pouring rain here i shut the blinds because i couldn't bear to look at it yeah yeah but we have good

4:51Mike:episode today um we're talking about what's been called the sas apocalypse which i think needs to come up with a better brand name but um this has been in fairness it's been a recurring theme now for a while um this this ai threat to software companies but last tuesday is when it really came to the fore not so much in a it was already happening but in a kind of catalyst i suppose and so we're talking about when anthropic released claude co-work which is an ai assistant ai assistant for automating tasks for legal companies this started out atst ripples across global market markets and a domino effect pretty much ensued obviously first of all were the companies specifically in kind of publishing and legal software uh so thompson reuters and legal zoom i think both of them had their worst day ever on record that's quite logical that's a very prominent and real threat to what they do in their business models but it didn't stay contained in that industry at all and basically spread like wildfire to to anything anything software in the entire market the the kind of consensus was you know if it can replace this quite complex software stock in the legal business an industry that's worth billions what's next so what if this agentic ai tool doesn't just replace legal assistance what if it replaces kind of the core apps that every major enterprise across the world uses.

6:19Mike:So in the space of a couple of days, markets lost about a trillion in market cap. The Nasdaq fell 5 % in three trading days. IGV, so iShares software ETF, was down 11 % in the same time. There's more to it than just that. Mr. Market really didn't like how Google and Amazon were planning on spending on AI, which is kind of illogical in a sense you know we have ai is going to kill software theme on top of wait why are you spending so much on ai uh but that is a different argument we're here to talk about software specifically today and as you can imagine once markets turn red everyone starts acting incredibly rationally and put together so alice from barclays asked in a research note is this the end of software those in jeffries suggested that software companies could go the way of print media or department stores.

7:10Mike:And this is kind of an enduring theme. I imagine it's easy to get clicks and likes and all the rest to just come out with these big bombastic statements, but it definitely is there. This AI threat is not a new thing and it's led to some serious declines in former market leaders and basically a huge re-rating across the board in terms of valuations for some of the top software companies. So this is as a time of recording down from their alt-iron highs. I'm going to run through a couple of big names we'd all know. So CrowdStrike, a stock very much benefiting from AI, is down 27%. Salesforce, down 47%.

7:49Mike:ADP, Automatic Data Processing, down 31%. Paycom, 76%. Intuit, 44%. Workday, 50%. SAP, 37%. HubSpot, 72%, Atlassian, 80%, ServiceNow, 54%, and Adobe, 61%. That's a diverse group of businesses. They serve very different industries, and that's only a few big names. Give me 10 more software companies, and they're going to have very similar drawdowns. So it is a blanket downturn, 100 % motivated by this long-term looming AI threat, but then very much exacerbated by starting valuations. So the sell-off is so extreme because software companies have always carried such high valuations. They're always valued on revenue rather than earnings for the most part.

8:36Mike:Because they're highly scalable, you build a product once and you can sell it a million times, you know, with more or less the same fixed costs. That's obviously incredibly high margin revenue then because of that. You've got long-term contracts. It kind of ensures stable, transparent revenue for the years and years to come. Why wouldn't anyone want to invest in that? And we saw that, like as in software has dominated the markets for, I'd say, coming out of the GFC in 2009, 2010, that whole period up until 2022, 23, was completely dominated by software stocks and the SaaS business model and this kind of revolutionary cloud technology that just made everything so easy.

9:13Mike:And I think Marc Andreessen had the famous phrase, software is eating the world. And it very much was for a long period of time. And that's why valuations, especially now, it's funny, I was looking through all these businesses. A lot of them peaked at the end of 2021. So that post-COVID run-up where everything just kind of went absolutely nuts and valuations went completely haywire. So those drawdowns aren't specifically from the AI threat alone. but they're very much exacerbated by that. And I think if you're looking in general, a stock can't carry a premium valuation with this now impending threat looming.

9:52Mike:You know, it's other very multi-characteristics for the industry as a whole are coming under threat too. Like I mentioned scalability, which was kind of underpinning the whole software movement. The per seat model now is under threat because not only are there major tech layoffs across the board, but there's also higher efficiency from AI tools. So there's less developers or designers or account managers or whatever else now needed to the same level of work in the future. Who knows? Like the speed in which AI is moving is terrifying, really. So it brings all that into question. Do you know what I mean?

10:25Mike:Like how much of that work can be completely automated by AI? Because that is, business leaders have come out and said that, you know, that is their goal. And it's very dystopian, but it's true. So now the argument for high margin software companies being valued on multiples of revenue is coming under fire. Hence why we're seeing such a universal decline across the board, irrelevant of the industry. You know, we're bringing CrowdStrike into this. It's one of the huge beneficiaries of that. So that's where we're standing now. Obviously, I believe that the threat to, say, Adobe is a lot stronger than the threat to ADP.

10:59Mike:You know, I don't think AI is ready to completely disrupt payroll. But that being said, maybe ADP is an unassailable moat, will be challenged in the coming years. is that the notion that a lot of these companies will adapt to AI and be the entry point for existing clients, maybe that makes sense at first glance, but maybe it unravels when you think about the AI native applications that are inevitably going to come to market. Do you know what I mean? It's a bit like the introduction of licensing software versus SaaS, which is ironic in a way, but the barriers to entry have come way down for new entrants and we could be at the precipice of a whole new way of doing things.

11:34Mike:Do you know what I mean? that like isn't yeah i don't want to be extreme because i don't believe it is the case but for a lot of non-mission critical software businesses that they could be they could be up for grabs you know that revenue that they're creating someone can come in and do it a lot cheaper and then i guess that's what that's all we're here to discuss really is to distinguish between yeah between the baby and the bathwater because at the minute everything is being thrown out but there's no doubt that there will be a number of winners in the software market There'll be a number of AI beneficiaries.

12:03Mike:There'll be a number of ones that just don't really feel the same threat. And I think that's what we're here today to pretty much discuss is there are some great businesses on sale, but there are also ones that maybe are facing a systemic decline in front of them. And that's what we're looking at. So to kind of define that and identify the industries, identify the companies and see which way I suppose the wind is blowing in terms of software because it is a very interesting concept right now in terms of the market. Big game, big meeting, big crowd. However you're gathering, Sweetgreen Catering brings the bowl everyone wants.

12:43Mike:Introducing the buffet-style Sweetgreen Bar, perfect for sharing, customizing, and keeping everyone happy. Because when the food's this good, people show up. Learn more at sweetgreen.com slash catering.

13:05Mike:akamai cloud gpus for agentic ai bring ai inferencing closer to users everywhere get started at akamai.com gpu it is and in fact when you pitched the idea to me yesterday about

13:20Emmet:let's have this debate i really locked in i'm looking forward to our conversation here and now because it's it's not binary and binary i mean it's not that everybody there we don't know we don't know how the future is going to play out and even when you see the stock market shed a ton of value in a matter of hours a fraction of trillions of dollars in a matter of hours it just goes to show that nobody really knows nothing changed in reality in those few hours only perception

13:47Mike:and understanding or at least this is what we're saying like you know like as an illegal assistant that can automate a couple of paralegals tasks and ADP loses 10 % of its market value like you know 10 billion dollars.

13:59Emmet:It's crackers but it's crackers and you happen across a few names which I hope we talk a little more about but I saw a very interesting post on LinkedIn only this morning by a woman who I don't know called Clara Shee which is the best summary that I lean towards agreeing with and I say lean towards because I have not 100 % bought into any specific argument I'm all ears and my mind is open but no one actually knows um so I wouldn't mind just giving our listeners a quick summary of Clara Shee's uh LinkedIn post she said I wonder how many times she's heard a joke she said that's what she said uh her surname is Shee so uh Clara Shee said and I quote, Claude Code has transformed how I manage both my 1000 plus person organization at work as well as home life.

14:55Emmet:But far from SaaSpocalypse, I believe agent solutions will actually usher in a SaaS golden age. Okay. That's quite an interesting one. It's what caught my eye. She says, first, every business relies on a set of deterministic workflows, e.g. legal approvals and guardrails, e.g. customer must sign by 11.59 p.m. Therefore, the overall system needs to be governed by determinism, even if AI is injected in places to make individual processes and the system better. For example, in B2B sales, there are AI use cases for researching prospects, automating outreach, etc. But boards won't accept an AI-generated forecast that the VP sales can't explain a VP sales won't accept an AI generated forecast and there was repetition she made that point twice SAS is essentially a system of governance and accountability for employees and each seed level leader and she goes on to say second the core value of SAS has always been coordination not productivity of a single employee VPs oversee managers and reps HR coordinating across managers and employees and then she says this is why classic SAS features such as access control are so important.

16:10Emmet:What Ben Thompson describes as SaaS capturing institutionalized processes. And third, AI models are changing fast. You don't want to be locked into a single provider. It would be irresponsible knowing that a game-changing new LLM drops every few months from a different lab. And then she says, while I'm bullish on the category, not all SaaS will thrive. the entire sector is being taken down to the studs and reconstructed for and with ai this is a good thing things people have hated about sas will get shaken out one unused seats will churn or get negotiated into ai credits two pricey seats from bloated vendors that struggle to become more efficient will get replaced by leaner competitors and three complex implementations will be largely automated with AI it'll be easier to audit and hold SI consultants accountable for faster high quality implementations and she goes on but first thing I'd say is when I read it I realized AI didn't like this LinkedIn post which was refreshing because it was kind of had it had a few human errors but a human tone of voice I've read so much AI output I regret to say that I can smell when it's a human who wrote it and there's an organic nature to this where she didn't say hey hey Gemini may tell me something smart to say.

17:29Emmet:But what's happening, I think, Mike, right now, it's just not the end of software. No, it couldn't be. No, it couldn't be. It's the end of lazy assumptions. It's the end of lazy assumptions about software because us investors are being forced now to really differentiate between tools that are easy to replicate and systems that are extraordinarily hard to replicate. And you mentioned, Dobie, there a few minutes ago, we had a full podcast about it uh i would say about two months ago um and i for one believe that the business doesn't look like uh it's in as much trouble as the market has priced it which contrary to what you've said i think it's an example that the baby went out with the bath bath water um and i'm starting to think paradoxically that ai is making the distinction

18:18Mike:of really valuable companies clearer yeah not blurrier I think I use the term lazy and I think that's a very appropriate word because goes both ways in the sense of what she said she do is um yeah in terms of software spending I think is going to come really under the microscope now because technology teams are going to have a certain budget and so much of that is going to have to be dedicated to ai so they're going to consolidate vendors for sure and the non-critical stuff is going to get washed away so those kind of hangers on won't be around anymore and and i think this is going to play out over the next two three years of we're going to find out through people to people through people voting with their wallets what is truly mission critical and what isn't and that's and that's going to be a top distinction for especially stock pickers is figuring out that because there will be software companies that fall in the wayside from this just surely church no doubt about it from a dollars and cents perspective there isn't enough money in tech budgets for a software that only three designers use and you know they're paying the minimum entirely the minimum uh spend is 10 seats or something like that that is going to be axed and they're going to figure it out.

19:41Mike:And those kind of non-critical platforms are going to go by the wayside. But that doesn't mean that software itself is going to go. And I think that's what she's really getting down to. What is mission critical? And that's the most important question I think you're going to have to ask yourself when you're looking at these stocks moving forward.

19:58Emmet:Yeah, you're right. I mean, AI, it just has suddenly and almost overnight reduced the friction involved in creating software. I think I was telling you yesterday how I uploaded a picture of my friend Donal to Claude asking it for a caricature for his birthday WhatsApp group and I returned the code for me to upload any photo to get a caricature and of course I don't know what to do with that I just looked at the code code around I don't want that's not what I want I want a funny picture of Donal so I just went over to GPT but the point is that coding agents can now write usable code it expected me to copy and paste that into something where suddenly I've got an app and And non-developers like myself can spin up these quick and dirty internal tools.

20:42Emmet:And one of the terms that I learned the hard way through my journey in my Wall Street and founding the business 12, 13 years ago was this term technical debt. That as a business works through its life, it builds up a stack of technologies that is akin and analogous to monetary debt. you have a pile of systems and software that is redundant half used fully supported partially supported we don't know what it does we don't pull the plug in case it turns out like this technical debt is a big thing and when suddenly you have eejits like me and i'm not going to do but eejits like me who are like i'm going to write the code myself i don't you know they are going it is going to build a whole pile of technical debt i mean prototypes that once took months to assemble are now being done in days hours maybe even moments and um and i think this is what's called

21:37Mike:vibe coding is that what vibe coding is mike yeah vibe coding is essentially not for non-coders

21:42Emmet:they give it's a great name vibe coding bro i'm gonna do some vibe coding then i'm gonna get some surf i it actually is a great name isn't it i'm gonna write an app and i'm gonna it's gonna get my donal my friend donal's avatar and then i'm gonna add surf dude but that's basically the audience but do you know what i mean like that is it and and i think

22:06Mike:it's definitely dumbed down a lot of processes yeah but it's also simplified ones that don't need that oversight and then i think yeah when it comes down to critical stuff i tell you what because an open ai is paying their salaries via traditional sas providers who have all the regulatory stuff and all the compliance and all the security built up they're not vibe coding their own salary payment stuff and and that's when that's when we go back to the same discussion of mission critical like is in this this vibe coding thing is scary and and then the other thing is as well it's just moving so fast do you know what i mean yeah we've seen this even from feedback from developers in the space of a year yeah the capabilities of what we keep coming back to antropics cloud of its coding system what it could do a year ago to what it could do now is is probably where the fear comes from it's like okay well this is blowing past all expectations in the space of a year and it's only going to get faster what happens in five years do you know what i mean is the art of software development going to be completely automated do you know that that's

23:12Emmet:where people are thinking entirely and like to your opening point is in it it's understandable why this has unsettled investors, especially in horizontal SaaS categories where products at first glance look replaceable. If software can be created for cheaper for free by flutes like myself, the argument goes that the pricing power will collapse. And why rent something when you can build it yourself cheaper? And you think that problem, our problem with that argument is it confuses creation with responsibility. and what it does not do any piece of vibe code is make it easier to run software inside a complex organization where generally failure of a piece of software carries consequences and that's just b2b like if you take duolingo which is i think a great example of a publicly listed business which is b2c business to consumer you have to suspect that um a business obsessed on reinventing the right way to learn a new language, which I think is probably one of the most complex and time intensive things you can take on.

24:22Emmet:It has probably figured a few things that AI has not figured, or at the very least, it's using AI to make their tools better. But when I look at Duolingo's graph, as I am right now in the middle of my screen, last May, it was about$400 and change. Today, it's about$116 and change. Its share price is sitting at five year lows. It's down 77 % last year. But when you look at the underlying metrics of the business, which is retained users and revenue and all those kind of things, quarter and quarter just gets better and better and better. But if I can loop this back to B2B, so that's B2C over there.

25:01Emmet:Most enterprise software isn't purchased because it's really clever. It's purchased because it makes an ugly task easy. It's auditable. It's compliant. and it's supported you can get onto someone if something isn't working and i guess you could also say it's familiar now if you walk into your your chief accounting officer your cfo in a fortune 500 company and you make a pretty good argument say that you should replace an internal financial system with some new tool that a whole bunch of vibe coders back and floor hc have done and they're great lads they really know what they're doing you can lead with the argument going we can take this line out of our annual opex but then you need if you're going to be dutiful you need to explain to that decision maker that they have you're going to have to migrate decades perhaps of historical data you're going to have to convince everyone else in the place if it hasn't been done through use and organic uptake you're going to have to retrain staff you're going to have to ensure this really flawless reporting and generally there's a regulator sitting outside the window who wants to know everything is being done the way it should.

26:11Emmet:And so I don't think, and this is why I don't even know what I think, but what I do think is that that decision, the CFO going, yeah, do it. I don't think it's now. I don't think that there will be a rational CFO who in 2026 says, yeah, okay, come on, let's just do it the AI way. I don't think so. I suspect in five years when we're quite a bit further down the path when you consider the speed of development and the rate of progress and all these LLM tools, it will be more of a no-brainer. But I also believe in the intermediate time, specialist software companies, whether it's Atlassian or whether it's Paycom or whoever else, they will also be looking at AI.