Nvidia Loses a Record Breaking $600 Billion in Market Cap

[00:00:01] This is The Canadian Investor, where you take control of your own portfolio and gain the confidence you need to succeed in the markets. Hosted by Braden Dennis and Simon Belanger Welcome back to The Canadian Investor Podcast. I'm here with Dan Kent. We're back for our news and earnings episode. And it's pretty jam-packed because unless you've been living under a rock, you probably heard the news of DeepSeek,

[00:00:30] the Chinese large language model, LLM, that is making a lot of waves and sending some pretty prominent stocks down. So we will be having a big segment on that. But before we get started on that, Dan, you want to talk a little bit about the Bank of Canada? I think it's safe to say that tomorrow, so when people will be listening to this, don't know what they did, but I think it's safe to say that they'll be cutting rates by a quarter basis point, right?

[00:00:58] Yeah, I think last week when we kind of were talking about it, I think they were pricing an 83% chance. I don't know what the odds are now. I didn't look at it, but there's really been nothing that indicates it's going to be, you know, nothing has changed, I would say, to put that towards less than 83%. Yeah.

[00:01:24] Trump is, he's going crazy right now. I'm really interested to see what he does, Tara. When is he not, is the question, but yeah, no, it's going to be interesting, right? I think he floated the deadline of February 1st now, or deadline, he's floated the date of February 1st. We'll have to see whether it happens or not. And we saw what they did with Columbia too. Yeah. I don't know if you saw that.

[00:01:54] That's kind of what I was saying about the craziness, just like an immediate, and then within what, like 24 hours, Columbia kind of changed her mind and kind of changed her tune because it's, yeah, he's, the U.S. has a lot of power. It's pretty crazy. They do. And at the end of the day, it's our biggest trading partner. Clearly, the U.S. has some dependency on Canada as well, but not as much as we do on them.

[00:02:18] It'll be interesting whether it's really economically related or they're just leveraging their economic power over Canada to try and get some concession on other fronts. One that comes to mind that's been talked a lot about is defense spending, the 2% of GDP. So maybe Canada will agree or politicians will agree to speed that up and try to get defense spending closer to 2% of GDP.

[00:02:45] I think they've committed to that, but it's a few years down the line. So we'll have to see. I really don't know, but I agree with you. The Bank of Canada will probably err on the side of caution and try to, and cut rates to try to mitigate any kind of tariffs that could be imposed. But again, I think people just have to take a step back because we really don't know what can happen. I mean, you did those tariffs and then they were lifted within a day or two, right? Not even our hours afterwards.

[00:03:15] So we'll have to see what D does. Yeah, I mean, we were talking about this before the podcast, but the Globe and Mail put out an article that they're looking at, and I would view them as a pretty good source, that they're looking at pandemic level relief type relief for companies or workers that are hit hard because of the tariffs. So obviously, the government here thinks they're going to come into place and they're going to be pretty severe because they're already planning for it.

[00:03:43] So it's going to be pretty interesting to see what happens. Yeah, yeah, exactly. Now we'll move on to the next topic. Before we start that, just a quick mention, I did move to a new place. So if there is a little more echo than usual on my end, I do apologize. I'm still setting up my podcast studio and I don't have the soundproofing all set up. We've just been moving. I moved for a few days with a toddler, so trying to get everything unboxed and everything ready to record.

[00:04:11] So I think I've done a decent job, but it's possible there's a little bit more echo than usual. It should get fixed in the next couple of weeks. Now, like I mentioned, going back to AI-related stocks really... I mean, I don't know if they've all... They didn't all crash, but there is some big ones that had a significant dive yesterday. We're recording that on Tuesday, January 28th.

[00:04:39] On the 27th, you saw NVIDIA, the most prominent one was down 17%, lost a record, I think. $600 billion worth of market cap, which I believe is an all-time record. Maybe not inflation adjusted, but on the pure number basis, it is. Yeah, it was a pretty ugly day for... Well, kind of US-based, but there was also a couple Canadian companies hit. Capital Power was one of them.

[00:05:06] They're a utility company and they fell like 16%. Well, I don't know what they closed at, but at one point I looked and they were down 16%. And I think they were supposed to be a pretty big beneficiary of that big data center expansion we have here in Alberta. I don't know if that was the only main tailwind moving them over the last while. But I mean, obviously with this news, there could be a lot less demand for data center and data center expansion overall.

[00:05:35] So they took quite a hit and Hammond Power would be another one. And so they've just absolutely skyrocketed over the last year or two because they provide a lot of transformers and kind of electricity components for a lot of these data centers too. So I think that was down 12% or 14%. So it wasn't just US. I mean, obviously these are much smaller companies. Way smaller. Hammond might even still be a small cap company, but it wasn't just the US that took a hit.

[00:06:03] There were some Canadian companies as well. Yeah, exactly. And I'll go over what happened. I've listened to several podcasts with people that are well in the know on this subject because I'm not an AI expert. I know you're not. Me neither. And I've read a lot of articles, some tech specific articles as well. So I'll try to make it as simple as possible, at least to my understanding.

[00:06:26] And of course, if you're someone that's deep into the space, it's possible that you'll think this is over simple and we don't want to make a whole episode on it. We do want to touch on a few companies that report it as well. So I'll start off here with DeepSeek, which is a Chinese large language model, LLM. I'll be mentioning this quite a bit, released its new AI model DeepSeek R1 on January 20th.

[00:06:50] So you might be wondering why the market actually tanked yesterday, a week later, because there's a lot of skepticism that comes from anything China. Let's be real about it.

[00:07:32] What's the case? We're going to be talking about a lot of crypto, right? We're going to be talking about a lot of companies. For the company that comes from a crypto domain, it's a lot of crypto domain, an open AI. So it really allowed a lot of people in the industry, including in those companies, to look at the model, see how it was. And then as more testing was done, as more expert looked at it, people came to realize that this was legit. This was not a false claim.

[00:07:57] Now, the chatbot interface is free to use while the API access is a fraction of the cost of the other LLM models. By fraction of the cost, I mean less than 5% of the cost for the API versus a chat GPT, for example. I think it's around like 3.5% of the cost, which is absolutely wild. But apparently that is the introductory price. It's likely going to be about 25% of the cost once regular pricing comes into effect.

[00:08:25] But still, you're slashing the cost by 75%. That's massive. And DeepSeek claims that it only spends $6 million to develop the model, which has a lot of people skeptical. Granted, I think this is probably not... Yeah, there's no way. Yeah. I don't think it's probably correct. But even if you say, let's just say it costs 100 times that, $600 million, it's still much cheaper than the other LLMs out there.

[00:08:54] Because those have required investments in the billions of dollars and not 1 billion, like billions of dollars. Now, I'm not like an AI pro or anything, but my understanding is that they've essentially focused on alternatives to using massive amounts of computing power because of the chip restrictions imposed on China by the US. We've talked about those quite a bit over the years or over the last year, year and a half, which most of the restrictions have come into effect.

[00:09:24] And to keep it simple, they focus on new training methods, more efficient algorithms, and software optimization over brute force, which is what was being done in the US with big tech. Essentially, they focus on being much more efficient.

[00:10:06] So, if you're looking at the cost, what LLM are you going to use? Yeah, exactly. I mean, it's just so weird because who knows the accuracy of the data? I mean, I was watching some videos this morning because, again, I know nothing about this. Most of my knowledge came over the last 24 hours just watching YouTube videos and stuff. And I mean, from the videos that I've seen, they said it was more accurate. They said it was slightly more accurate than the GPT model.

[00:10:34] So, I mean, especially when costs are this low, like who knows? Like if you think about it, I think Meta has $60 billion planned out next year in capital expenditures, which are like mostly AI focused. So, you're talking like a monumental amount of money being spent. And then this thing comes out that effectively it says it can do it for substantially less cost.

[00:10:57] But like you said, even if they're under exaggerating this, as you said, 100x would still only be $600 million, which is not even close to the amount of money that all these big tech firms have poured out over the last, what would it be? Probably 15, 18 months on infrastructure. And yeah, it's crazy. Yeah, I think the accuracy is really in the computing itself. So, it's a little lead accurate.

[00:11:24] I think they, without, because that's a bit beyond what I, you know, my expertise. But they use, I think, 8-bit versus 32-bit for a lot of the computing, which will be slightly less accurate. But the overall product may be as accurate, if not a bit more. It really goes to the efficiency and the power being used because they're using that 8-bit over 32-bit. That's what I've read.

[00:11:50] Again, I'm not an expert and I'm trying to simplify it as much as I can. And my knowledge is limited on this. But they also claim that they've only used 2,000 NVIDIA H800 GPUs. Those are the GPUs that are a bit less powerful than the highest-end one. They're actually made and tailored for China because they're made to be able to sell them to China in face of U.S. restrictions.

[00:12:17] So, I think it's important for people to remind themselves of that. And the GPUs are a little less powerful. But from what I've read as well, the other LLMs I've used in excess of 10,000 GPUs. So, even if you think they've used more than 2,000, if it's 5,000, 6,000, 7,000, 8,000, it's still a lot less. And they're using the less powerful GPUs as well compared to the other one.

[00:12:41] So, it's fair to say that it's definitely, I think, I think it's probably going to be a game changer. What kind of impact will it have? I'm not quite sure. But it's so efficient that apparently you can run it well on personal computers and smartphones. And one thing that I think a lot of people like is you can see it reason as well. There's the LLM will actually show you. I haven't used it yet myself. So, I'm looking forward to trying it out. But those are the main things I learned.

[00:13:11] And for me, what it kind of reminds me is you're old enough to remember that too. Remember like in the early 2000s when CPUs were like a single core? Yeah. And then at some point they got the dual, the multi-core. So, I think it was around 2005. And then I think it was Intel. Then AMD came out with CPUs that had multi-cores. And nowadays, I mean, you're seeing like 10, 15 cores if not more.

[00:13:37] But the reasoning back then was that if you had one core, you had to increase the frequency or essentially it was taking more and more power to generate more performance. And then they came in and added those additional cores and they were able to reduce the amount of power being used yet achieve better performance. And that kind of reminds me that a little bit, although it is different. I get it in this situation.

[00:14:04] But it reminds me of that where technology got better and the power usage actually got down. Yeah. I mean, I don't really know too much about that. I would have been like 10 years old at that time. I probably wasn't keeping up on that. But yeah, I mean, there's some rumors that, you know, they're using more powerful GPUs, but they can't really talk about it because of the trade restrictions. Like they would kind of be like... They're circumventing it. Yeah.

[00:14:33] Buying it for third-party countries and stuff like that. So, they're kind of like, you know... Or smuggling them. They can't really openly talk about having access to them because it would be, you know, because of those trade restrictions. Because effectively, you know, the US has restricted them. Which, I mean, in turn might have bit them in the butt a bit. The US restrict them because obviously, I mean, they've found a way to innovate using lower end technology if that truly is all they're using.

[00:15:01] And have come out with a, again, you know, whether or not it's a more efficient product is, I mean, it's kind of just rumors at this point in time. I mean, I tried to sign up for it. I was going to play around with it, but it wouldn't take my email for some reason. I think it was just maybe overloaded. I think they closed off new registration yesterday because of the hack. Yeah. They got hacked, like, not too long after it came out. So, I couldn't get into it because I was going to play around with it. Because I do use GPT a lot.

[00:15:31] And I mean, if this thing's better, I have no problem going over and using this. Especially because it doesn't cost anything. GPT, I think we pay like 30 US dollars a month or something for GPT. Yeah. Yeah, we have it. I have it too. Yeah. Yeah. So, I mean. I mean, you have a free version, but it's never the best model. Yeah. So, if you're using it, the free version, if you want the most up-to-date model, it is costing something.

[00:15:56] What's interesting in all of that is, regardless of how many GPUs they use and how powerful they are, the reality is people validated code and can say that, yes, it's much more efficient. Oh, okay. So, I think that's the reality is people, because it is open source, from my understanding is people are like, yeah, they actually found a way to make it much more efficient with that 8-bit versus 32-bit that I was saying.

[00:16:22] And I think that's where there could be some impact from NVIDIA. So, like we mentioned, dropped 17%. Clearly, investors were spooked with the news. It will likely impact demand for the most advanced GPUs, with most of them being currently produced by NVIDIA. So, I think this is a likely. And when you see people with hot takes, I mean, you have to be careful a little bit just because it's pretty new. We don't have all the information.

[00:16:52] Who knows how things are going to play out? The reality, though, and we've chatted about that for NVIDIA, is that it's trading very expensively. And I'm just showing here a chart for joint TCI viewers and subscribers. So, you can see here the chart that I have. It's the price-to-earnings ratio on a trailing basis. Whatever. You can use forward basis if you want. That's backward looking. That's fine.

[00:17:20] But now, the forward basis is probably even trickier to use because then who knows what the forward estimates will look like. You don't know what the impact is going to be. Exactly. So, I think the trailing is probably the better one to use at this point. But again, these are all to give people an idea. And it's trading at 47p and 34p. I actually have the 34p. But the 34p is based on previous estimates. So, we'll have to see whether that changes or not.

[00:17:49] And if you look at it from price to sales, it's still trading at around 25p, 26p, which is extremely expensive. And people might say, okay, 47p, 30p is not that bad. Yeah, if the company is growing super quickly and has massive growth rates, sure. And granted, NVIDIA could very well keep growing pretty well. But even if the growth rate decelerates a little bit, that's where you could see some big swings

[00:18:17] in the stock price for NVIDIA because these expectations were so high. And that's the issue I see with people that are saying this is an overreaction is they never mention valuation. Sure, it might be an overreaction in the overall business of NVIDIA. I'll give them that. Maybe it is. I'm not a clairvoyant. I don't know what will happen in the future.

[00:18:42] But the reality is when you have a stock company that is priced to perfection, it does not take much for it to have a pretty massive drawdown. So, that's the one thing I will push back for people saying that. Now, if we look at NVIDIA and we just look at how concentrated its revenues are, it's pretty staggering. So, for joint TCI viewers, you'll see the chart here.

[00:19:11] So, you have the data center revenue, which is pretty much all related to AI. And then you have total revenue. And over the years, it's been a bigger and bigger portion of NVIDIA's revenue. So, right now, you're looking at the last 12 months. The data center revenue is 87% of all revenue. Now, compare that to 40% in 2022 when AI was just starting. You had companies that were already investing in it.

[00:19:40] Even though ChadGPT came out in the fall of 2022, you still had investment. But again, it was 40% of revenue back then. So, there's not much room margin for error or margin of safety involved here, especially when you're so reliant on that one segment. And the problem that you start seeing here is NVIDIA's margins. So, if you're looking at their margins, their margins over the last 10 years have never been this high.

[00:20:08] So, you're looking in terms of gross margins at 76% and 63% for operating margins. And that's where it gets really tricky, in my opinion, for NVIDIA is because who knows what will happen in the future. You can make a case that, yes, demand for NVIDIA will still be pretty strong.

[00:20:29] But again, if this new technology is allowing companies to invest in AI but using less of these GPUs, for example, or alternative GPUs that are produced by AMD, for example, that may be 90% as good as the NVIDIA one. But for a fraction of the price, you may start seeing some demand shifting over to competitors because now they can say, well, you know what? Yeah, you have the best GPUs.

[00:20:58] But for my money, for the value, for the use that I'm going to do, it's much better that I go with the cheaper alternative from your competitor. And that's another risk that I think you'll see from an NVIDIA. And if that starts happening, then their margins will start compressing. I think the one thing I will say here that I'm almost certain is that you will see that their margins have peaked. Those margins, a gross margin of 76 and 63.

[00:21:26] If you're an NVIDIA shareholder, enjoy it while it lasts because I don't think you'll see these margins last for the next couple of years. I think you'll see that coming down. That will impact profitability. And again, even if you say, you know what? Over the long term, this will just accelerate adoption into AI. I know Satya Nadella posted something. The Javon's paradox strikes again. This was his tweet.

[00:21:54] As AI gets more efficient and accessible, we will see its use skyrocket turning into a commodity we just can't get enough of, which could be a tailwind for NVIDIA. But at the same time, when you think about all the implications, you can make a case that yes, demand will stay relatively strong from NVIDIA, but there's going to be margin compression happening. The demand will not be as strong as a lot of people projected.

[00:22:22] And it's going to result in probably some lower returns than people anticipated over the next few years and longer term as well. Yeah. I mean, that was the main theory that I seen where it would kind of be bullish for NVIDIA. They pretty much stayed just like you said that, you know, as costs get lower to develop, the technology gets adopted more.

[00:22:46] And, you know, maybe where it was unprofitable now turns into profitable as the technology gets cheaper. So it gets, you know, wide scale use by, you know, more companies, things like that, more applications. But I mean, that, like you said, that data center revenue, it's compounded at like a 78% pace over the last, since 2013. So, I mean, if you see a slowdown in that, ultimately when you have, so before the crash, I think NVIDIA was trading at like 65 times trailing free cash flows.

[00:23:15] And I mean, as soon as you get any sort of uncertainty in the future or even a potential, no matter how small you think the chance of a potential disruption would be, I mean, any company that price is going to tank. And I mean, we saw it tank, you know, 20 plus percent just off the worries, because like you said, it was, it was priced to perfection. There's no doubt. 65 times free cash flow is a very expensive valuation.

[00:23:44] When we look at a company like Meta, it's trading at 30x, or I think it was like 32x. So like half the price of a company like, like NVIDIA. And it's still expensive. Yeah. And that's, yeah, like 32x free cash flow is fairly expensive. So, I mean, it's not surprising to see like any disruption to, you know, the overall outlook on the business. I mean, it's going to hit the stock price hard, no matter how unjustified you think it is. I mean, there's going to be a lot of panic there for sure.

[00:24:12] Yeah, because people were pricing it or were willing to pay a really high multiple for NVIDIA because they were projecting in the future thinking, okay, AI is superpower hungry. We need the best companies will need to buy the best GPUs. NVIDIA is the best that there is right now. It'll probably be the best for the years to come as well. So they'll be able to keep their margins. They'll keep generating cash, gobs of cash over and over.

[00:24:41] So it justifies the valuation. A lot of the arguments that I just said, now you can start poking a lot of holes in these arguments, which is the reason why the stock took such a big hit. And I wouldn't be surprised if it continues to face a lot of downward pressure over the next few weeks, months, years. I wouldn't be surprised. I think NVIDIA is still going to be a really good company.

[00:25:07] But again, if the valuation is not right, you can buy a really good company and you can be looking at subpar returns or even negative returns because you paid too much for the company.

[00:25:48] Yeah. I mean, they had a dominant market share. What would that be? That would have been like probably just financial crisis, post-financial crisis. And then they kind of fell off. I mean, they... But I mean, I think this situation is a little bit different, but I just...

[00:26:07] I think that's kind of a comparable situation to, you know, a player who back then probably, you know, a lot of people thought, you know, the moat was untouchable until it wasn't. And I mean, in a situation like this, you know, a lot of people thought that this company was just going to continue to rip moving forward, which, I mean, prior to yesterday when this just came out of nowhere, you know, you could argue that it could. But now you have, you know, something like this comes in. It's more efficient. It's like half the cost.

[00:26:37] Maybe companies don't need to spend, well, Meta, $60 billion in capital expenditures every year to develop these programs, which ultimately, like, who were they spending that CAPEX on? It's companies like this. Yeah, it'll be interesting listening to those conference calls coming this week, right, from Big Tech. So I think all of Big Tech is reporting this week, and especially Google, Tesla, Microsoft, Meta.

[00:27:02] It'll be really interesting to hear what the CEOs have to say in terms of DeepSea, because you can guarantee every single one of them will be asked about it. Satya Nadella was on in Davos, I believe, and saying this was the real deal, too. So people are really taking notice, and you may have some shareholders asking, like,

[00:27:25] okay, so you spend all this money for no reason, or you won't see the return on investment that you thought you were going to do. Because keep in mind, these companies, they weren't spending all that money out of the goodness of their hearts. They were spending all that money because they wanted to build such a big mouth around it that they would remain dominant, integrate that with their various services.

[00:27:49] And now they've spent billions and billions of dollars that they might not see a lot of return on investment on. So it'll be interesting if shareholders start asking some question. And I don't know, it'll be, I'll definitely listen to most of them just to see what they have to say. I don't have too much more to say. Anything else you want to add before we move on to earnings here? Nope, that's it. Okay, so we'll move on to earnings.

[00:28:18] The first one, I guess I'll let you go ahead with Good Food, because I feel like I've talked most of the first segment, so I need a little break here. Yeah, so Good Food, it's a company I typically follow. And we kind of go over it a lot on here because it reports like kind of off the path of many Canadian companies. So it always gets brought up because it's an early reporter. But I also think it's kind of a pretty good insight into, you know, discretionary spending for Canadians.

[00:28:46] I would definitely call this a luxury type item. And I mean, their results certainly paint a picture how soft the Canadian consumer is. And I say this now because we're comparing results to December of 2023, which really would have no pandemic-related tailwinds left. Like, we're now just comparing year-over-year results in relatively the same environment. And the thing is, it's still struggling so badly.

[00:29:15] I mean, the company reported revenue of $34.6 million. That's down 14% from the first quarter of 2024. So this was their first quarter fiscal 2025 years. So it's down 14% from the first quarter of 2024. Gross margins came in at 39.6%, which is a 0.2% improvement. With food inflation coming down, it looks like the company is starting to report some stability in gross margins,

[00:29:44] at least because it took a pretty big hit in this regard when, well, what was food inflation? It was 7% or 8%, I think, for a while there. So I could imagine that it took a big hit on that front. I mean, operating margins are razor thin. They're only 0.7%. The company is having a very hard time pushing out any sort of profitability in this environment. Back in November, they made an acquisition of Genuine Tea in order to try and expand the business.

[00:30:12] But I mean, their leverage ratio is now 2.64x. So this would be debt to, I would imagine, adjusted EBITDA. It used to be so one year ago, it would have been 2.34x. So I mean, I would argue that an acquisition for a company struggling to even achieve profitability isn't the best. However, I can't really say. I don't know much about Genuine Tea. Maybe it does well for them. But I just don't think, you know, spending in this type of environment,

[00:30:40] especially when they're struggling all that bad, is all that good. But they've tried to shift their business so many times. Like, I think during COVID, when they kind of knew the meal boxes would slow down, they tried to get on, like, meal delivery, like grocery delivery. They were building warehouses in Ontario and everything to try all that. It didn't work out that well. Obviously, they're trying now to, you know, further diversify into, you know, a tea company, maybe for delivery, maybe to be included in the boxes.

[00:31:09] I'm not exactly sure. Pre-pandemic, the company had interest expenses of around $1 million annually. So it now has annual interest expenses of $5.6 million. And that's despite revenue being lower than it was back in 2019. So revenue is down. Interest expenses have almost 6X'd. And it's just, it's a pretty rough time for good food. I mean, the company's guidance doesn't really bring any sort of confidence either.

[00:31:35] So they suggest that the meal kit penetration for Canadian households will be about 4.2% in 2029. But the thing is, it's only 3.5% right now. So, I mean, that's a compound, like, that's a compound growth rate of around 4% a year. So, I mean, that's not really, like, encouraging guidance. Based on what? Yeah. Yeah. I don't know. Like, it's just like, oh, that's really not that high.

[00:32:02] I feel like they just took a number out of thin air and they're like, oh, you know, this is achievable. If you're going to take a number out of thin air, I guess you should probably try to make it more impressive, though. Like, as soon as I've seen that, I'm like, 3.5% to 4.2%. That's really not that much. But the investor presentation is pretty funny, too. So they have, like, their net income and free cash flow charts.

[00:32:26] And, like, on the top of them, it says they highlight consistent growth and profitability and meaningful free cash flow generation. But the charts are all over the map. Like, net income up, down, up, down a little more, down, up. It's all the same with free cash flow. So, I mean, there's nothing consistent about those charts, yet they kind of label them as, you know, consistent growth. The only thing that I would say has been consistent is just the decline in active customers.

[00:32:56] This is a company I actually own this during the pandemic. I bought it relatively early and booked some profits when it soared up. So I didn't end up doing all that bad. But I sold off the final chunk of my position. I think it was in between $4.50 or $5. So it's down, what is it now? I think it's, like, $0.50. $0.48. Yeah, so it's down, like, 90% since then. Yeah. And they did have a ton of momentum during COVID originally.

[00:33:25] And I felt growth would certainly slow, but not this badly. I thought maybe, you know, the food box type situation during the pandemic, it might get more people to the business that would stick around. But I think when food inflation started ramping up, profitability just completely got crushed. And then I think interest rates as well really hurt the company. Because, I mean, people just don't have the money to be spending on food boxes right now.

[00:33:49] And I know there was a big issue with the company as well where people were just abusing discounts. Oh, yeah. I mean, I personally, myself, did that. Yeah. I mean, it was, you know, there's so much churn at that business, I would imagine, that they just always have to have those offers. And, I mean, I can't imagine they make any money on those discounts. And, ultimately, people are just going to exploit them and use them lots.

[00:34:16] And I think they were just really lax on, like, who would qualify for one. So, I mean, you could just indefinitely use those discounts and effectively get... I mean, the cost that I was paying during the pandemic, it would be borderline free food. Because it was like $45 and you'd get four meals from it. But, yeah. Yeah, I think now they are making it so you can only get the offer once in a while. So, I think you have to wait a few months until you get the offer again.

[00:34:45] But we've used it on and off. We basically will use it when we're eligible for the offer. Usually, it's like 50% off for the first week, 30% for the second week, 20% and then 10%. Usually, we'll use it for the first, like, three weeks and then we just stop it. We're like, okay, got some good value out of it. And we'll restart it later. It's crazy good value on the initial discounts. Like, you can't get food that cheap.

[00:35:13] Like, the convenience to a certain point is worth it. I mean, especially when you get, like, four meals delivered to your door for, like, $45. But then once it gets up to regular price and it's like $130, $140, people just cancel. So, I mean, who knows? Maybe, you know, interest rates come down, you know, things improve a bit. But, I mean, I think there's a pretty big cost of living crisis here in Canada. I just can't see food boxes being, you know, on the top of the priority for many people.

[00:35:42] And it's kind of showing in the results. Yeah. And it's convenient to some extent. I mean, you still have to do the meal and prepare it, right? There's stuff that's cut, but not everything is cut. So, yes, it does save you time. It also saves you the time of not having to think of what to make for the meal. But, again, you still have to choose what you want. If not, you'll get random stuff that you might not like. So, there is – it is convenient.

[00:36:08] But it's also not, like – it's not a meal that's already prepared that you just put in the oven. Yeah. You still have to do some stuff. I mean, I don't see an easy path to, you know, getting things back on track for this company. I just can't see it returning to popularity. Yeah. No, that's a good way to end it. Now, we'll move on. I think we'll just talk about this last company. I think it'll be an episode.

[00:36:37] We're getting on the 40-minute mark. So, we weren't sure exactly how much time we'd spend on the whole DeepSeek thing. So, it was good to have a couple extra companies to talk about. The next one on the slate here is American Express Q4 2020 earnings. Now, revenues, net of interest expense were up 8% to $17.2 billion. Net income was up 12% to $2.2 billion. Earnings per share was up 16% to $3.04.

[00:37:06] And these are all the results for the quarter. Provisions for credit losses were $1.29 billion, which was 5% less than last year. Which is a good trend to see for American Express. And people may be wondering. We've talked about it again. But I know we get new listeners at this time of the year. American Express is different and similar in some ways than Visa and MasterCard. So, it's a hybrid between a Visa, MasterCard and a bank. That's the best way to see it.

[00:37:35] So, it has its own network. So, the American Express network. So, that's why you can go to like a Scotiabank, for example. Or BMO or whichever bank you want to choose in Canada. I think most of them will offer some Amex cards that are issued by them. So, the cards can be issued by banks, third parties that will use the American Express network. But where Amex is different from a Visa or MasterCard is they also issue their own card.

[00:38:03] So, you can get an American Express, not with another bank. It's American Express and the bank is actually American Express. And they actually, that's where they differentiate it. Because they are a bank. So, they're a chartered bank. So, that's the biggest difference between the Visa and MasterCard. Where Visa and MasterCard, they just have the network and they take a percentage of the fees. Where American Express does issue their own cards. And they get revenues in other ways.

[00:38:31] Interest revenues and things like that. Expenses were up 9%. Which is a bit of a concern. And it was a bit more than revenue. So, definitely something that if it's a company that you're interested in, you'll want to keep an eye on. The CET1 ratio remained high. So, that's a ratio that's used for bank at 13.2%. But it was down a little bit compared to last year. Compared to last year, they had 4% more total active cards.

[00:38:59] With 50% of those being issued directly by Amex. So, that's what I was talking earlier. So, they're directly issued by Amex. And then 43% are issued by third-party banks. The total Amex network had 7% more volume in the quarter versus the previous year. Write-off rates were pretty much unchanged compared to last year. So, that is something to... That's definitely a good thing. Because we've seen write-off rates for banks and credit card companies. Subprime letters.

[00:39:29] We've seen those rates definitely go up in the last year. So, the fact that they're pretty stable depending... And there's a bunch of different write-off rates that they use. But they're pretty much identical compared to last year. And they had a few interesting things on the call that I wanted to just mention before we wrap up over here. And I'm just pulling up a chart at the same time. They mentioned that small business sentiment improved during the quarter.

[00:39:57] Which resulted in higher spend from these small and medium-sized businesses. So, SMEs. They're focusing on adding merchants around the globe. And they added more than 1 million new merchants during the year. Their focus is international growth, millennials, and SMEs. They have 25% of all US fee-based premium cards. Which is pretty impressive. Because clearly Visa and MasterCard are much larger. But the fact that they have 25%.

[00:40:25] And you can see here I'm showing the average fee card. Which is a pretty good portion of the revenue that they get. And it's just steadily... It's a line. It's up and to the right. Yeah, this is... It's impressive to me. Because so many cards are coming out now that are offering a ton of features. But no fees. I mean, American Express must have some pretty impressive rewards. To be able to generate a lot of fees from their cards.

[00:40:55] I've looked at them. I've had one in the past. I haven't had it for several years now. But they do have some of the best rewards. At least I've looked at them. I've seen in the past. The problem with Amex. And I think they're trying to address that based on the call. Is you need a backup credit card if you like to pay by credit card. So, you need a backup MasterCard or Visa. It's not as big of a network. Like I know not a lot of places even take Amex. It's a little bit more restricted. No, more than Discover. But yeah.

[00:41:26] Yeah. Yeah. Not a lot of places take them. And I found when I had them. And I think there was more places that take them now. But still back then. And I would always get into this awkward situation where I'd be at the restaurant ready to pay. I'm like, oh, do you take Amex? Question mark. You know, in my face. And that was the annoying thing is you can always pay with it. But they are also putting a focus on refreshing their offerings. And they refreshed over 30 products last year.

[00:41:55] And are looking to do the same thing about the same amount of product refreshed this year. They are celebrating their 175th year. I don't know if you knew that. No. So, they said it on call. So, they got founded in 1850. Which I was, I had no idea. But they pivoted to the credit card business in the 1960s. And back when they were founded, they were actually a freight forwarding company. Oh, that's weird.

[00:42:25] Yeah. Very weird. But the CEO actually talked about that on the call. American Express. Yeah. Yeah. Exactly. That's where it came from. So, I was just, you know, you listen to the call on 50% increased speed, 1.5x. And I had to rewind a few times just to listen to that part. Which I thought was interesting. And for 2025, in terms of guidance, they expect revenue growth between 8% and 10%. And earnings per share increase of 12% to 16%.

[00:42:55] So, definitely really good results. I don't think there's much to say that's bad for American Express. I think every time I talk about them, I wonder why I don't own them. I'm not a big fan of banks. But the fact that it gives me a little bit of a hybrid between a card network and a bank, I do like that. It's always trading at a pretty high valuation. Not crazy, but still on the higher end.

[00:43:24] So, I think that's what's been preventing me from pulling the trigger. But every time we do them on the podcast, I do wonder, like, why don't I own this stuff? Yeah, like, they trade. They're cheaper than the payment networks. Like, Visa and MasterCard typically trade higher. But I think that's kind of because they have similar growth rates. But obviously not as much risk because... And probably higher margins as well. Just because they're not... You know, they don't have that financing side.

[00:43:54] Whereas American Express does. But, yeah, like, I've always just found the limitation in terms of usage. Maybe it's gotten better over the last while. I'm sure it has. I mean, I know they had the Costco card for a bit. But then I think they lost that. Or maybe they still do in the U.S. I don't know. But MasterCard has it now. Like, CIBC has it.

[00:44:18] But I remember thinking, like, American Express, Costco would have been the first thing I thought of, like, for a while. Just because they had that card. But that's one of the main reasons I've never really thought about owning them. Just because of the... I wouldn't say limited network. They probably still have a ton of places you can use the card. But it's definitely more limited than something like Visa and MasterCard. Plus, it just takes on that extra financing side of things. Yeah. And so I just Google in the U.S.

[00:44:46] It's Visa by Citigroup. Oh, yeah. Yeah. So they lost it both here. They lost that one too. Yeah. Yeah. Which I think they've... It's fair to say they've recovered from. But I think it was a big blow to them when Costco ended up going with, I guess, Visa and MasterCard here in Canada. But it's a really good company. It's definitely trading more expensive than banks. But like you said, less than the Visa and MasterCard.

[00:45:15] And the one thing it probably has going for it is that it will likely not face as much scrutiny from a competitive standpoint as a Visa and MasterCard. Yeah. I know it's a new administration.

[00:45:41] I know it's a new administration now with Trump in the U.S.

[00:46:24] Visa, but maybe a MasterCard as well. Visa is 81%. Yeah. So, I mean, there's a big difference there. There's no operating margins listed there. But, I mean, Visa has 67% operating margins. So, I mean, it's just... There's just less... Fewer things going on, I guess, with something like Visa, which is probably why it tends to trade at a premium. Yeah. Yeah. No, I think that's fair. But given the regulatory risk associated with Visa and MasterCard, I think I would be more comfortable, I think, going with American Express.

[00:46:54] I mean, I own Visa and MasterCard in small portion. So, maybe it is as simple for me as doing a basket approach and adding MasterCard to the mix and keeping my positions in Visa and MasterCard kind of unchanged. Something I, you know, I wasn't thinking about, but now that we looked at them and I'm pretty impressed with the results and knowing that it's a hybrid, too, between a bank and a Visa and MasterCard, a lot to like. So, I think that's about it for today.

[00:47:23] Dan, anything else you wanted to chat before we sign off? Nope. That's it. I'm sure we'll have more to talk about next week when this DeepSeek... I'm sure there's going to be new stuff every single day on this. Well, aren't you away next week, I thought? Oh, yeah? Yeah. Is that next week? Yeah. Yeah. Maybe you'll bring your podcast equipment. Yeah, maybe you'll bring your podcast equipment on vacation and going to that golf tournament. Yeah. With a few beers in you, we'll unleash Dan Kent. Yeah.

[00:47:54] Yeah, I doubt it, but I'll be busy golfing and enjoying hopefully warm weather, although it's plus 10 here today. So, I mean, it's probably... Oh, wow. Okay. Yeah. Very nice. Like minus 20 here. So, yeah, I'm definitely jealous there. But I think we'll wrap it up here before we start rambling too much. Thanks again, everyone, for listening. It was a fun episode to do. Definitely came out a bit of left field.

[00:48:20] Wouldn't have expected that we would do such a big segment on AI if you asked me last week. But again, this is news and earnings. A lot of development happened. It was really interesting doing the research. I learned a lot. Looking forward to try DeepSeek as well. Once the registration is back open, maybe it is now. I'm not quite sure. But definitely something I want to check out because I think it could be a game changer. We'll have to see. But it sounds like it might be.

[00:48:48] Yeah, I'm definitely going to check it out as soon as I can get in there. But yeah, thanks for listening, everybody. The Canadian Investor Podcast should not be construed as investment or financial advice. The hosts and guests featured may own securities or assets discussed on this podcast. Always do your own due diligence or consult with a financial professional before making any financial or investment decisions. The Canadian Investor Podcast is a production of the United States.