Episode Transcript
The Finance Ghost: Welcome to episode 269 of Magic Markets. We are in mid-April. It has been a pretty wild time in the markets this year. We've been dedicating a few shows recently to talking about some of the macroeconomic stuff, but there's another big theme going on out there, which has been somewhat forgotten, actually, because of all of the geopolitical unrest and the energy shock and everything else.
And that is the SaaS-pocalypse, as I've seen it described. Basically, the Software-as-a-Service or technology stocks are getting absolutely hammered this year, thanks to what's going on in AI.
And we thought, perhaps it's time to actually have a look at that, Moe. Because you've been playing around with Claude, haven't you? You've been playing around with the source of all this pain.
Mohammed Nalla: Indeed, Ghost. I was lucky enough to get two months free on Claude, the paid version. The reason why I'm saying I was lucky enough, is that I was contemplating, do I test Claude against some of the other AIs there? Because at the end of the day, there's only so much money I want to be spending on AI.
And I've seen the horror stories of people whose AI builds just escalate massively, because they're doing a lot of this engine stuff in the background. So, the fact that I got two free months meant I got to test this out. And look, it’s still early days, right? I've literally only started doing this since last week.
My first main reservation around Claude was the fact that it has usage limits, where some of the other AIs I use, I don't run into token usage limits. With Claude, sadly enough, there's a weekly usage limit and I haven't broken that yet. But on the day that I was intensively trying to test this out, I ran through my five-hour usage limit within two hours.
And that, for me, is a massive constraint. Because unless I'm willing to upgrade, and I guess that's the business case from Claude or Anthropic's perspective, and unless you're willing to upgrade, you burn through your capacity quite quickly. And just in a nutshell, my experience has thus far been rather disappointing.
If I compare Claude's output to stuff I'm getting out of OpenAI, out of Gemini, out of Copilot - and I've tried all of these, right - I would say Claude is distinctly lower-tier. Caveat out there, I haven't really tested it intensively for the stuff that it is quite well known for, which is the hardcore coding type of stuff. I still want to try and do that and see - maybe it shines in a specific use case.
But at the end of my experiment last week, I just got so frustrated after burning through those usage limits time and time again, that I eventually just parked it and I said, look, let's actually see what I can do in the coding space. I'm going to run that over the course of the next couple of weeks and give our listeners an update on that. But I was left unimpressed.
The Finance Ghost: It's karma for all the times you've listened to podcasts on 2x speed. Claude has basically more than “2x speeded” your five-hour limits. I don't really understand how that works. So, if they say five hours, why do you get two?
Mohammed Nalla: It's a good question, right? It's quite opaque and you’ve got to go and read the terms and conditions, but they have your terms of use. And if I use this in effectively North American office hours, that is your peak usage - and so you can actually burn through your five hours in two hours.
If I did this over a weekend, for example (that's low impact on their servers and so forth) then I could probably stretch it beyond five hours, or maybe at least get to my five hours. So that degree of complexity and opacity around usage, around when can I use it, isn’t great.
And then just the fact that I've got these token limits, where on other paid off offerings, I don't have token limits. I've paid for a service, I'm going to get it and I'm going to know where I need to be on their paid tier.
Now with Anthropic, that was still somewhat uncertain to me and I think the gap from where I needed to be, versus what the current paid program was giving me, was more than 2X - and so it eroded some of that use case.
And again, this is interesting to our discussion today because I'm looking at this and I'm saying, you know, maybe it's user error. I know you're going to throw that one at me, so I'll throw it at myself, right?
Maybe it's user error, but I see people out there with these phenomenal use cases with Claude Bot and they're trusting them or trusting Claude, I guess, with some very sensitive information with the ability to transact, albeit with a yes, a prompt and so forth.
Just given the output I have received, I would be absolutely terrified doing that. Not just specifically on Claude. I just don't think AI should have access to certain very sensitive information.
But on that point, that again ties into some of what we're going to try and discuss, certainly in the SaaS space where there has been an absolute bloodbath.
The Finance Ghost: Black boxes can be quite scary when you're trusting your life with the thing. I've just gone back to driving a manual car, which is quite nice, I’ve got to say, as a petrolhead.
I kind of like to work the same way. It's nice to get a little bit of assistance, but there's a limit for sure.
The market doesn't think so though. The market thinks that these Software-as-a-Service companies are toast, or at least the golden days are toast. That's an important nuance here. I don't think that anyone is saying “well, Adobe and Salesforce go bankrupt”, but what we are seeing in a lot of these names is that share prices have come off really hard in recent times.
So that's what we're going to dig into today. I mean it's quite remarkable to go back. I'll just give you something to just give context to the conversation here. The 10-year share price CAGR on Salesforce is 8.3%, Adobe is 9.3%, S&P 500 ETF is 12.8%. So over 10 years, despite the golden era of Software-as-a-Service, you would not have beaten the market if you were in Salesforce and Adobe, thanks to the recent sell off.
Isn't that remarkable?
Mohammed Nalla: It's amazing. I'm glad you looked at that stat because I looked at another one, and I was saying, okay, where are we on the technicals? Where are we relative to the 200-week moving average? That's a very long-term moving average. It defines the long-term trend.
And for context, the S&P 500 is around 30% above its 200-week moving average. So, it’s still solidly in a bull trend, even with some of the correction we've had. But if you look at most SaaS names, they're down 30% to 50% below their 200-week moving averages. They broke those a while ago. They confirmed those breaks. So they are under a lot of pressure.
Now, for context, one of the stocks I'm going to talk to is going to be Microsoft. Microsoft right now is getting a lot of chatter on social media. People are like, “Oh, it's broken its 200-week moving average”. And they're right. But when you're testing a 200-week moving average, it's not a straight line through.
It's going to test it. I see it as Microsoft testing its 200-week moving average. For context, that's still doing 30% to 50% better than some of the names in that SaaS space.
And then just on the other side of that, you've got Alphabet, which is 94% above its 200-week moving average. It’s shooting the lights out, beating the S&P 500. And even Amazon, which underperformed Alphabet, is still around 40% above its 200-week moving average.
And so, that also outperforms the S&P 500. Because I was trying to figure out if software has done so badly, what are the big names that have driven the index? Obviously, you've got Nvidia, but those other names, Google or Alphabet, have done quite well. You've got Amazon, that's done quite well.
So, this concentration on software is really bifurcating along the lines of who has a moat, where are you in the ecosystem? And you can actually see that once you strip it down to individual company performance.
The Finance Ghost: And the market's taking big bets here. They really are saying, “Look, for this to be true, we need to now assume that Alphabet is going to win the AI era”. Which means that Google, which was the SEO era, is now going to just beautifully transition into this “AEO” era, where people are searching with artificial intelligence and everything's optimised for that. And then ads are served inside those models, etc.
It's really big bets that the market is taking. I always go back to a sensible thing to remember in all of these cases: the market tends to do two things. One, it underestimates long-term change, and it overestimates the rate of change.
So, it assumes that things will go really, really quickly - when usually they don't. And then often people underestimate how much things can change, when in reality things can actually change a lot over the long term. And you can see that come through in some of these valuation multiples.
I'll just give you a couple. So, as I said, I looked at Salesforce and Adobe, which are kind of the obvious two choices in the SaaS space. I'm so glad you did some different names, Moe. It's going to make it really interesting,
If you have a look at these two businesses, Adobe's gross margin is in the high 80s in terms of percentages, and Salesforce in the mid-70s. Now, gross margin is maybe not the best indication necessarily (of profitability) because it's not a traditional “We buy stuff and we sell stuff”-type model.
What goes in gross margin versus what goes in operating margin is probably a point of debate, but those are obviously very high margins.
If you look at Adobe, operating margins are in the mid-30s in terms of percentages, and it was back in the mid-20s in the period after the global financial crisis.
So, what happened is over about a decade ago. Adobe managed to put about 10 percentage points on its operating margin. And that was the transition from software being bought off the shelf into the software as a service SaaS model, cloud, etc, etc.
Salesforce is actually quite funny, because if you look at Salesforce, the operating margins there were frequently in the low single digits because they just basically spent everything they made. They're very well known for the incredible amounts of stock-based compensation as well.
But recently, Salesforce has pulled back on that stuff to show that they can actually make money, and their operating margin is in the low 20s. So, Adobe's numbers actually quite a lot better than Salesforce.
But it hardly matters - because when the market panics, it panics properly. And now both of them are trading at a price/sales multiple of around 3.7x.
Now to give that context, before the pandemic (so not even pandemic craziness), Adobe's multiple was up at about 15x. Now it's at 3.7x. When last was it in that range?
Guess what? It was around the time of the aftermath of the global financial crisis, when Adobe's margins were 10 percentage points lower, and people were paying a similar price sales multiple of 3x to 5x, which is really quite remarkable.
So, the market is basically saying, “Well it looks like we might actually see these companies going back to those days”. What would drive them there? If AI made software much cheaper, perhaps new competitors can emerge, perhaps people are vibe coding, per se.
I'm not so sure I believe that story too much, but I guess there is some truth to the argument, or there might be some truth - that AI will make software cheaper. And it's not difficult to find people who are willing to complain about the price of something like Adobe. It is very expensive.
What the market is saying is, we're only willing to now pay a price sales of 3.7x. By the way, on Salesforce, you have to go into the global financial crisis to find that multiple. Nevermind the aftermath, you have to go right into the GFC. That was almost two decades ago.
It is incredible. The market is either extremely right here: these companies, their golden days are over, their days of those margins are gone, or the market is extremely wrong, and those who buy at these levels are going to make a lot of money in the years to come.
It's a bit of a gamble at the moment. Right? You really are betting on technology.
Mohammed Nalla: It's quite a tough one. To your vibe coding comment, we're at the user optimism stage where a lot of people are willing to expend time and energy to try and vibe code.
And once you try and do it enough, you're probably realising that, well, “I'm not going to actually go and vibe code myself a new tax software, not going to go and vibe code myself a new Adobe Photoshop, because I don't need to do it”.
In fact, to the Photoshop point, maybe that's why Adobe is also under a lot of pressure. AI image generation and the impact on creativity - I think that's a big angle on that, right? So, let's strip that out of it. I think with the broader vibe coding, your risk is really someone out there going and let's say in a garage, vibe coding an entirely new platform.
But then, remember, if they're doing this as a business, yes, maybe they've done it with significantly lower cost, but then they go into the same bloody waters as all of these other SaaS companies - that have deep balance sheets and established processes and sales processes.
So, it's not as easy as, “Hey, you can just plug and play with something that you vibe coded in your garage and go out there and compete with some of these bigger players”. So, to that extent, I think some of the fears are overblown.
However, I think the divergence that you're seeing in the market … You've mentioned some of the margins, some of the multiples when you look at a player like Microsoft (and that's where I'm going to go with this one, right) - when you look at a player like Microsoft, their operating margins a little bit higher.
And that's because Microsoft trades - it's a software company as well - but then they've got the whole cloud, they've got a whole bunch of other stuff in the ecosystem as they try and pivot.
Similar to what you said with Google. It was an SEO business, now it becomes an AEO business. And that's because your large incumbents, like a Microsoft, like an Alphabet, they can actually invest. They can throw billions and billions of dollars into this, to try and pivot their businesses to evolve.
Microsoft's done that successfully over several decades. For me, the question on a stock like Microsoft, specifically, is: why is the market punishing it? The market's looking at things where they're saying Microsoft's throwing hundreds of billions of dollars into capex spend here.
I think it was close on 40 billion in the last quarter and a lot of that spend is being allocated towards the short-lived assets like GPU, CPUs, the whole data centre demand story as well. And the market's looking at that and they're saying “That's great, but where's the actual spinoff on the other side, in terms of realising revenue from this?”.
And I think that's the story behind a player like Microsoft. It's rather a question of, “Hey, you're falling behind some of the other large players in the space”. Amazon Web Services, Google with their cloud offering. And so that's the punishment story on Microsoft.
For me it’s very different to what you're seeing in some of your more traditional SaaS names. Because I view Microsoft as partially having transitioned away from being a pure SaaS business. It is still exposed to the SaaS industry, but I see Microsoft as more than SaaS.
And again, longtime listeners will know we've covered Microsoft in detail on Magic Markets Premium. I'm not going to give away what we've done on that show, but if you're a subscriber go and check that out. If you're not a subscriber, it's only R99 a month. So go and have a look at that as well. Great value.
Ghost, I'm going to pause there because I want to talk about two additional companies. One is Intuit. That's one we've also covered on Magic Markets Premium. They've got a moat, for example, your tax software TurboTax is a very large portion of that business. There's QuickBooks that sits in there.
So I want to talk about Intuit and then I want to talk about another business called ServiceNow, which I think a lot of listeners may not be exposed to, but they are still an important player in the SaaS space up in North America.
I'm going to pause there, let you maybe chime in in terms of maybe your views on Microsoft and anything else that you had a look at.
The Finance Ghost: Microsoft has been, what I always call it, “My ride or die”, since the start of Magic Markets really, but it’s more die and a bit less ride at the moment. But that's fine. I'm kind of just sitting back and just letting this thing do its thing. And if it keeps coming down, I'll buy more.
Microsoft has stood the test of time through a lot of different technological eras. And I get it that there's always a risk, and maybe this time it's different, and maybe something will go wrong for them, or whatever the case may be. But I think when you're in the enterprise layer, then stuff like data security and all of that becomes really, really important.
To your vibe coding point - it's one thing when you're targeting SMEs and maybe they're willing to take a chance and they're very, very cost sensitive and everything else. But that old story of “Nobody gets fired for hiring IBM”, that still applies today.
And that's basically the Salesforce model. To say, “Well, we have all your customer data”. This is actually the most sensitive data in your organisation. Do you really want to go and feed that into a general AI model? Or would you rather use an AI layer in our world, and do your analytics accordingly?
For Adobe, it's saying: well, we're going to aggregate all these image-generating models, we're going to make sure that it's commercial safe. We're going to protect you as the creative and make it easier for you. That is probably how they both survive.
I think the question is whether they survive with the fees that they've been able to charge for the past 10 years. That's where the question comes in.
Whereas Microsoft can charge almost whatever they want. I can't cancel Office, realistically, all of that stuff. Yeah, I suppose there's Google as an alternative. I know some organisations do run on it, obviously. But it's a difficult thing - the cost of moving across… switching costs are very significant.
Mohammed Nalla: Yes, okay, maybe they can't survive with the fees that they're charging I mean, Adobe can get very, very expensive. But remember, when you're looking at your company as margin, you're looking at two aspects.
And in terms of protecting that margin, the same way someone can go and vibe code in their garage, guess what? These big players can also utilise AI to make their own back-end a lot more efficient, and maybe they don't need to charge those fees to try and defend some of the margins. So, I think there's a little bit of push and pull there.
Let me move on from that into two distinct names. As I mentioned, Intuit and then ServiceNow. The reason why I've looked at these names specifically is, I looked at how Microsoft was more than a SaaS name. Now I'm looking at Intuit - and for me, Intuit is distinctly a SaaS name. Yes, absolutely.
But like I say, they operate in these very interesting markets, and specifically TurboTax. And the reason why this is topical is, we're in the middle of tax season up here in North America.
TurboTax has, I wouldn't say a monopoly, but they own a significant portion of the market share up here. And they just seamlessly integrate, in terms of being able to pull through all of your tax information from various slips. And then, producing a return and then liaising with the tax authority in Canada, the CRA. In South Africa it would be the SARS, and in the US, the IRS.
And that's something that right now is just too sensitive. There's a big moat around that. I don't see people necessarily trusting AI with that. If they do, they need to seriously question how much they're trusting tech, because AI at present is really public information. So, I have some question marks around that.
It's why, I think, Intuit has a bit of a moat, certainly around the Turbo Tax product. And yes, there are disruptors even in that space, but I would give that a little bit of a green flag.
The second point though, is things like QuickBooks. Already, QuickBooks operates in the kind of small to medium enterprise space. Companies larger than that usually go for different services and so forth, but in the SME space, they have a distinct niche. You mentioned barriers to entry.
For you and I to move our books away from QuickBooks… Just as an example, when QuickBooks transition from desktop onto online and they stopped supporting desktop, I was very grumpy because I like to buy my software and use my software until it's obsolete.
And even though it was a painful process, I went from QuickBooks back into QuickBooks, rather than considering some of the alternatives. So that gives you an indication - just in terms of familiarity - of what that does for your moat when it comes to sensitive company information, and when it comes to your tax data.
Now, before I get into the performance of some of these big stocks, ServiceNow is a stock that some people might not be familiar with, I certainly wasn't exposed to it when I was down in South Africa. But what ServiceNow does is, they really provide some of your business process software in the back-end.
So, if you think of integration of HR systems, your IT payroll goes into that - that's what ServiceNow does. And this is where I think the potential for disruption can become quite large. Because if you are a medium-to-larger corporation, maybe you want to start exploring building some of that capability in house.
We've seen all of the disastrous SAP integrations down in South Africa with some of the large corporate names there. And when you have painful experiences like that, guess what? There is no moat. You’re actually incentivised to try and say, “Can we build some of this capability in house?”
And where previously this would have entailed hiring a very well-paid team of developers, right now you could probably get away with… I wouldn't call it vibe coding, but it's like, you have entrepreneurs and intrapreneurs. This would be like the intrapreneur version of vibe coding within larger corporates.
And that is where I think players like Anthropic with Claude (or wherever it might be out there in the AI layer) start to chip away at some of that in more medium-enterprise business. I don't think this is limited to consumers. In fact, I think consumer is where some of the SaaS names are protected.
I think if you go up the value chain there, if you go to mid-tier corporates who don't have the big budgets to spend on the big names, that's where I think some of that disruption comes through.
Again Ghost, I'm going to pause there, because I want to look at the five-year and the one-year performance. Maybe just some comments? Am I actually on the right track in your view?
I think that makes a reasonable thesis for why we're seeing a little bit of a dispersion even though the overall SaaS sector has been under pressure.
The Finance Ghost: And that's exactly where the opportunity comes in, right? It's the baby with the bathwater. So, everything gets whacked. It's amazing that Salesforce and Adobe are now at identical price/sales multiples.
You can't tell me that that's because they are identical businesses. They have totally different margins, they're in completely different industries, and yet the market is just like, “We don't understand this anymore. We'll pay 3.7x for you both.” - and it’s dropping all the time.
So that's the opportunity to look and say, okay, which of these tech companies is actually in serious trouble?
And it's the stuff where the moat is relatively narrow, and the service is actually a real grudge purchase. It's not necessarily very sensitive data, whatever the case may be, that stuff is in trouble, as it should be. It's a weak business, unfortunately. That's how life goes.
The stuff that should be fine is the stuff where the data really matters, or it's a mission-critical process, or moving across is very risky. Something can go wrong. And then it becomes a real business risk where people say, “Well, we can't really move everything across to this, we're going to have to take it slow”.
So, there are going to be casualties. Absolutely. But the money will be made by picking the ones that are not going to break. And then getting long on those names and then sitting it out.
Mohammed Nalla: Indeed, Ghost. I think to wrap the show, I’m going to ask you some questions here, right? One-year and five-year performance.
Over the last five years: let's pick an assortment of the names we've spoken about.
So, we’ll use S&P as our benchmark, right? The we’ve got - let’s put Alphabet in there. No Googling, I can see you going to Google! [Laughs].
The Finance Ghost: Asking AI! I would never Google! That's so old-school.
Mohammed Nalla: Alphabet, Microsoft, Amazon, Adobe, obviously Intuit as I've discussed. ServiceNow.
I think that's a decent assortment. There are lots of other smaller companies we just don’t have the time to discuss here. Like PayChex. Yes, that's the name of a company. Paychex. Over the last five years, who's the top performer?
The Finance Ghost: I'm guessing that because of what's happened in the last year, probably Alphabet. Because they've had such a crazy run, it's just taken them out of how everyone else has done.
Mohammed Nalla: That was the easy one. So, I'll give you that. That's Alphabet. Definitely the top. Over five years. In fact, even over one year that shot the lights out,
One year, over 100%. Over five years, 176%. So that's been Alphabet's mega performance. Now maybe a nuanced question is on the five-year, of those names I've mentioned, how many and which ones have outperformed the S&P 500? That's a trickier one.
The Finance Ghost: Yeah, that is a tricky one. I'm actually not so sure that Microsoft is on the right side of that anymore. I haven't looked. You're going to depress me if they haven't. I suspect that Amazon has, because they've had a few useful tailwinds. So there. Tell us, Moe. Ruin my life around Microsoft. I just know that the answer is going to suck.
Mohammed Nalla: I'm going to ruin your life around this entire sector, right? Because actually none of them other than Alphabet have outperformed the S&P 500 over the last five years. The S&P 500, let's call it 60%, more than 60% up over that time period. If you look at Microsoft, 45%. It's a strong performance, sure, but certainly not what you would have expected in this tech hype world.
And then the loser. Absolutely on the other end of the spectrum, the worst performing stock of the subset we've mentioned is… Drumroll. Do you want to guess?
The Finance Ghost: No, tell me.
Mohammed Nalla: It's one you've covered. Adobe.
The Finance Ghost: It's got to be. Yeah, yeah, has to be.
Mohammed Nalla: Adobe down 55% over the last five years. Absolutely horrible.
The Finance Ghost: Ugh!
Mohammed Nalla: If you look at that Microsoft this year, is pretty much flat, down low-single digits, - 2% there and thereabouts. I didn't have a Microsoft position. I've recently added a Microsoft position. I believe this 200-week moving average should hold on a technicals.
Yes, it's speculative. If we actually see that support hold, I'm probably going to lean into that simply because I believe in backing a strong horse. I believe Microsoft has the wherewithal to actually ride this out over the longer term.
But some of those other names have really just all clustered to the downside just over the last year. Take your pick. It's Intuit, it's Adobe over the last year, it's even Salesforce. You look at any of those names. They're all down over the last year, around 35%, 40%.
So that's absolutely massive. And they've given back where they had gains, they've given a lot of those gains back and are now firmly in the negative even over a five-year time frame. So I think the long story here is to be selective.
And when we say, buy the best company in its sector, that certainly makes sense. Because even if you're doing that, as I've indicated, only one of those has actually outperformed the S&P 500 over the last five years.
The Finance Ghost: Now, I'm familiar with the head and shoulders technical pattern, but if you look at Microsoft over one year, it's more like the spiky hair pattern. Because it basically starts at one ear and went up, up, up towards the middle. Lots of little spikes, and it's just dropped off the other side, like the shape of a head.
And now it's back to the other ear, essentially back where it started a year ago. So, there's that, which is not ideal. That's not really what you want to be seeing.
At least I didn't buy the top of the head. I've had it for a long time, so, it’s a messy time. I'm certainly holding on to the names that I have, and we'll see what happens.
Mohammed Nalla: I want to comment on that head and shoulders. Yes, there was a head and shoulders on Microsoft if you go back to 2024, it formed the head then. Then it had a shoulder, and that's when it broke. And it broke and tested that 200-week moving average back in around March last year.
So, we're now close on those levels. We're testing the 200-week again.
Is that concerning? Yes, absolutely. The more times you test a level, it maybe increases the probability of breaking it to the downside. That being said, I believe in management, the strength of the balance sheet, and I think that they can pivot from that.
I think that 200-week moving average historically has been reasonably reliable. Every time Microsoft has tested those levels, it seems to have bounced back.
Hope is not a strategy. I'm going to watch this one closely. If we actually see it breaking, and signs of that building to the downside, I'll probably cut the position. But for now, let's see how that actually lands.
I know you are still a Microsoft shareholder. I'm a more recent Microsoft shareholder, so I'm down a little less than people who got in at the peaks. And I'm happy I didn't chase this, because I'd actually exited Microsoft when we had those very extended levels. Then I felt like an idiot because it actually went even further.
And now I don't feel like so much of an idiot because you redeploy the capital to where you're probably getting less risk, or where you can sleep easy at night.
I think that's probably where we've got to leave it Ghost.
The Finance Ghost: Yeah, I think that's exactly where we have to leave it. So, thank you to our listeners for being here. We're always happy to hear from you about what you're doing, actively trading this trend, or if you are just sitting and waiting out the volatility, or if you run away out of fear, let us know.
And of course, join us again next week where we can have another fun chat about the markets.
Mohammed Nalla: Indeed. Hit us up on social media. It's @magicmarketspod, @FinanceGhost and @MohammedNalla. All on X. Or go and find us on LinkedIn. Pop us a note on there. We hope you've enjoyed this until next week, same time, same place. Thanks and cheers.
The Finance Ghost: Ciao.
