Summer Series: Appen (ASX: APX)

Bryce: [00:01:14] Welcome to another episode of Equity Mates, a podcast where we help you learn to invest in 45 minutes or less. We break down the world of investing from beginning to dividend so that you can hopefully make some returns. My name is Bryce and as always, I'm joined by my equity buddy Ren. How's it going, bro? [00:01:29][14.9]

Alec: [00:01:30] I'm very good. Bryce. How are you? [00:01:31][1.1]

Bryce: [00:01:31] Good. Pumped to continue on with our summer series Shallow dive into some stocks that have been suggested by our Equity Mates community. So are really enjoying this series also really enjoying a bit of time off, to be honest. So it's been a busy year. So Ren today a listener has recommended. I think this came in a couple of times and it's one of the darlings stocks of the ASX 2019, and that is APEN or [00:01:56][24.9]

Alec: [00:01:56] apon or apon [00:01:57][0.5]

Bryce: [00:01:58] onshore. Yeah, I'm saying apen. [00:01:59][1.6]

Alec: [00:02:00] All right, I'll say Athen, OK, [00:02:01][1.4]

Bryce: [00:02:02] it's a tech stock. It's one of the wax stocks. So we have wise tech Afterpay happen happened and zero. Yeah. Yeah. So has had a very good year in terms of share price growth. [00:02:14][12.0]

Alec: [00:02:14] It has, it has [00:02:15][0.8]

Bryce: [00:02:16] so apen in a very important space at the moment. And that's data, all things data. We know how important data is to businesses at the moment or everyone really. It's a huge industry and apen play in this space by providing, I guess, machine learning and artificial intelligence products and helping improve the quality of data for businesses so that they can then feed that into their systems. [00:02:38][22.9]

Alec: [00:02:39] Yeah, what they do is they get images, audio, video, all that stuff, compile it together and then they give it to the Googles and the Amazons and the Facebook's of this world to train the AI and machine learning algorithms on these massive data sets that apon created. Yeah, yeah. [00:02:59][20.2]

Bryce: [00:03:00] Incredibly laborious work. Yeah, yeah. [00:03:02][2.2]

Alec: [00:03:02] It's seen as a tech stock and it gets valued as a tech stock probably. Fair enough. But it is a manual job. [00:03:10][7.7]

Bryce: [00:03:11] A very manual. [00:03:11][0.5]

Alec: [00:03:12] Yeah. And, and you can't get algorithms to do it because you need it to be done by humans and then train the algorithms, the algorithms screw it up then you can't train the algorithms. Correct. [00:03:22][10.1]

Bryce: [00:03:22] Yeah. So let's address that. Firstly, it started researching this and I had no idea really I knew what they did but didn't know that they actually crowdsource their. Yeah. Their employee base to do that. So they have over a million remote workers around the world helping them with their services. And to your point, Ren, that might be labelling photo after photo that this is a trade, this is a trade. Yeah, sure. [00:03:46][23.9]

Alec: [00:03:46] Yeah. You know, this is off topic. But you know how when you had to use to prove that you weren't a robot and you had to put that string of random numbers and letters? Yes. That was actually crowd sourcing data set to train I. Yes. And now you know how you have to click the photos that, you know, have a street sign or a crosswalk or whatever. That's the same thing. Essentially, they're using it one as a security mechanism, but they're using us to crowdsource. Yeah. What happened here to train? [00:04:15][28.9]

Bryce: [00:04:16] I yeah. I mean, there's many ways in which you can do it. And I happen to have just gone with the well, let's crowdsource a million people around the world. So their whole aim is to, I guess, improve the relevance of of this search data so that, as you said, Ren Google, when we when we as a consumer search, we're going to get more relevant search results coming through and they use real people to sit there and plug in this. Now, my question is, who says that the person's right? [00:04:40][24.2]

Alec: [00:04:40] Uh, I think I think they then have people that check it. Oh, they've got another. [00:04:46][5.7]

Bryce: [00:04:47] Yeah, yeah, yeah. [00:04:47][0.7]

Alec: [00:04:48] Quality control. But there are a million contract is a spread across 130 countries and one hundred and eighty languages. So even just the quality control place would be a nightmare. [00:04:58][10.8]

Bryce: [00:04:59] Crazy. And so there's a guy called Don't Have a surname apology's. But Kerry who oversees the recruitment of the crowds. So he's he's a crowd recruit. I like how it is that actually [00:05:11][12.1]

Alec: [00:05:12] these days it would just be like ads on Freelancer five, like all those sites. And it's just like want to make money from home, want to, you know, sit in your pyjamas and get paid a good wage, like just the label for for us. [00:05:28][16.1]

Bryce: [00:05:28] Yeah, but you've got to be, I guess, a particular person to enjoy that. Do you do you watch [00:05:33][4.8]

Alec: [00:05:33] YouTube all day, watch our videos instead and get paid for it. [00:05:37][3.3]

Bryce: [00:05:37] Yeah, true. So a really interesting business model. I guess it lends itself to be out of scale because there's more than a million people out in the world. They probably become a tipping point, I guess, but certainly in the right space. The reason that I say that is because even in my my line of work where we're exposed to artificial intelligence in the growth, but we are also in a position where we know our data is not up to scratch in terms of being able to feed into these artificial intelligence systems with relevance. And I can understand how important it is to have data that is relevant, clean and. Ready for use in these sort of technologies? [00:06:16][39.4]

Alec: [00:06:17] Yeah, so they service eight of the top 10 technology companies in the world. I don't think they tell us which eight. But I mean, we could probably have a crack Google, Facebook, Amazon, Alibaba, Tencent, Oracle, Urangan, [00:06:32][14.8]

Bryce: [00:06:32] Oracle, maybe it's [00:06:33][1.1]

Alec: [00:06:34] IBM like they doing a bunch of stuff with I like. Yeah, you could just assume that it's most of the big snap. [00:06:40][6.1]

Bryce: [00:06:44] Evan Bayh, maybe [00:06:44][0.8]

Alec: [00:06:46] Evan Bayh IPO next year. Yes. [00:06:47][1.6]

Bryce: [00:06:49] Anyway, that's off topic. So yes, super, super interesting company and one that is Australian based, which is exciting. It's in the information technology sector on the ASX. [00:06:59][10.8]

Alec: [00:07:00] It is. It is. And it has had stunning growth over the last few years. The last three years, it's grown its net profit after tax at about 60 percent per year, which is pretty good. Yeah. Has had another earnings upgrade in November. I think it's had three earnings upgrades of light. And as we keep hearing from our expert investors, stocks that continually upgrade their earnings, continually have positive news for the market, tend to continue to do so. Absolutely. It is a company that is well situated in a strong area of the market. And let's not kid ourselves. The market is definitely pricing some of that in. But we'll get to that in a sec. If you're going to start a business training, I in you know, 2020 is not a bad one to do. [00:07:46][46.4]

Bryce: [00:07:49] But they might if they're relying on crowdsource people to actually do the work, what's stopping someone else coming in and doing this? What's the differentiating [00:07:58][8.6]

Alec: [00:07:59] factor? What's their MO? Yeah, the fact that they have the millions of contractors and established datasets, like, I guess if you and I, we're going to try and start this from scratch, there would be a lot of work that goes into it before we even can start trying to sell into Amazon. [00:08:17][17.9]

Bryce: [00:08:17] I can tell you what a trade is. [00:08:18][1.1]

Alec: [00:08:19] Can you tell us what it is in 180 different languages? That's the [00:08:22][2.8]

Bryce: [00:08:22] point, because [00:08:23][0.6]

Alec: [00:08:24] what I would be worried about if I was Athen is similar to what I was talking about before, about how I don't know the company that did it, but the company that used this security mechanism of us having to put in those random strings of numbers and letters to train that eye, to recognise numbers and letters, I would be worried that someone else comes up with something like that that, you know, uses an existing part of the Internet that we all interact with and then finds a way to use that to train I. And what that is, I don't really know. [00:08:57][32.7]

Bryce: [00:08:57] Well, the company that run that robot thing that you're talking about is still innovating ways to do it. There's an interesting Planet Money episode on it. I can't remember what the actual piece of technology is called, but they're still innovating around ways in which they can continue. Yeah, it's interesting. [00:09:13][15.3]

Alec: [00:09:13] I mean, like, you know, if Apple or like everyone puts their passcode in their phone hundreds of times a day rather than using a passcode, will get them to do something to try. And I don't know [00:09:25][12.0]

Bryce: [00:09:26] what happens when you don't [00:09:27][0.8]

Alec: [00:09:27] I to work. Yeah. And then how do you know it's actually the owner of the phone putting the cloud in it? Yeah. So, like, that's not a great idea. But you know where I'm going. You find a common point and then you use that. Yes. [00:09:40][13.0]

Bryce: [00:09:41] Yeah. The reason I ask about its moat is because they spend one hundred and seventy five million dollars acquiring figure eight, which was another crowdsourced annotation data annotation company. And my thinking is, are they just on a on a rampage now or are they just out there to acquire anything that sort of starts up in this area just to be an absolute monopoly? Or because I'm sure from what I imagine, it's something that is going to become more and more important as time goes on. [00:10:10][29.1]

Alec: [00:10:10] Yeah, I see what you're saying. I think there is a benefit to scale. And, you know, with the existing contracts with big companies like, let's say, Google or like where you want to to your point about identifying traits, we want to be our AI to be able to identify every single tree in existence, in every single language. If you and I had a Start-Up that was then tasked with getting the datasets ready to train, the AI or APON was a task, was that same thing like they would just because of their scale and because of their established network, they would turn that around a lot quicker than us. Yeah, and so it's not an insurmountable moat, but it is a nice little head start to have. [00:10:50][39.5]

Bryce: [00:10:50] Maybe then the business that we need to start is purely nasch and just to try try and nail it. [00:10:57][7.1]

Alec: [00:10:57] Yeah, yeah, yeah. [00:10:58][0.7]

Bryce: [00:10:59] We've established that they're in a great industry, great area and are in high demand. And to your point, Ren, we could probably move more towards the financial side. Now they've had phenomenal growth. Phenomenal. [00:11:09][10.4]

Alec: [00:11:09] Yes. So topline their revenue over the last few years, calendar year. Fifteen, eighty two million to one hundred and ten million to one hundred and sixty six million to three hundred and sixty four million, which has been a nice trajectory. As I said before, with the net earnings growing at about sixty percent a year from eight million in calendar twenty fifteen to forty one point seven million in calendar year twenty eighteen, really strong. If we talk about valuing this company, [00:11:37][27.5]

Bryce: [00:11:38] let's do it. Two point seven dollars billion market cap. Quick question, though, with that sort of market cap, I still classify this as a growth stock. Where do you kind of see it? [00:11:46][8.7]

Alec: [00:11:47] One hundred percent. The market definitely values that as a growth stock. Yeah. So if we start with the relative valuation, just looking at its price to earnings. Yeah. Fifty six. Yes. Which is high. [00:11:57][10.3]

Bryce: [00:11:58] Yes. I'd pay for it at the moment. Now considering that the IT industry in Australia not as significant as it is over in the US, so the sector pay is twenty two point eighty seven. Okay. So significantly more expensive than a lot of the competitors in the space. But you need to also consider what are its competitors deemed to be. And they're not the growth stocks that we're sort of seeing in the wax, that's for sure. Yeah. [00:12:24][26.5]

Alec: [00:12:25] So on a relative basis it's expensive. Yes. But I guess if you compared it to some of its US peers, it would be less so like fifty percent out of out of line. Yes. There is a question about some of the reasons that you can give software based companies such a high pay is some of the unique points of software. It's highly scalable at a low incremental cost, I would argue that Athans cost structure is a little bit different and so maybe doesn't justify the same pay going forward unless you can justify some of the growth numbers. [00:13:03][38.1]

Bryce: [00:13:04] It's also worth pointing out that companies like this that are very much in early growth stage, strong growth stage pay is not really used as a price, as a fair value indicator. It's also tough to be talking about value when you're talking about growth stocks like this, because you have to make some pretty decent assumptions. So let's give it a crack Ren. [00:13:27][23.5]

Alec: [00:13:28] So after relative, we will look at discount cash flow. And if people want a refresher on these different valuation metrics, we did a valuation episode mid-December before we started this summer series. But if we have a crack at it, so I think this might be a good one to unpack some of the assumptions and discount cash flow, because, yes, you can play with the numbers and you can get very different such things [00:13:51][23.0]

Bryce: [00:13:51] as much as you [00:13:52][0.5]

Alec: [00:13:52] want. But let's start. The last three years, the earnings have grown at 60 percent. Realistically, 10 years of 60 percent growth is not going to happen. Like, it's just it'd be unbelievable. Yeah, yeah. But let's say let's let's use five years. Let's say it can continue that growth rate for five years, which is probably still quite ambitious. But let's start there. So the earnings per share number without any non-recurring items is 41 cents. If we say the 10 percent discount rate, which is the rate we're always using and we say five years of growth at 60 percent, which is just so ambitious, and then inflation after that, I get twenty six dollars and 16 cents. [00:14:33][40.6]

Bryce: [00:14:33] OK, so twenty six point six cents. It's currently trading at twenty two bucks. So slightly undervalued in its current number. [00:14:41][7.5]

Alec: [00:14:41] Twenty two ninety five. [00:14:42][0.6]

Bryce: [00:14:42] Just want to circle back and say that all these numbers that we're talking about, particularly earnings per share you can find on Yahoo Finance or Google or anything like that. So very easy to do this calculation yourself. And we're using a website called Guru Focussed Dotcom to get our discount cash flow calculator now Ren. So you've gone 60 percent growth over the next five years. If we were to stick with the 10 year period that we've been using throughout these episodes, we come to a valuation current fair value of ninety seven dollars. No, no, sorry. That's with a 50 percent. If I throw in 60 percent of one hundred and seventy seven point fifty four, [00:15:23][40.7]

Alec: [00:15:23] that's the beauty of compounding. [00:15:24][0.7]

Bryce: [00:15:25] So this just goes to show the difference that your assumptions can make in a discount cash flow model. All of a sudden we've gone from a value of twenty six point or whatever it was, to a value of one hundred and seventy seven dollars purely by changing the assumption that it's going to grow at that rate for a longer period of time. So Ren my second question then would be rather than change the growth rate from a 10 year period to a five year period, why not change the growth rate itself to be more realistic over a longer period? [00:15:55][30.0]

Alec: [00:15:56] Yeah, I mean, we can definitely do that if we have a crack at that and let's say twenty five percent. Yeah. So let's say 10 years at twenty five percent. Everything else stays the same. You get a fair value of nineteen dollars and twenty nine cents, which is below the share price currently. But it's sort of in the, in the range. Yeah. Because there's so many assumptions that you can sort of throw in and you can really we can make it align with the twenty two point ninety five. It's trading at now a way to just strip away the assumptions and just think about what the market is telling you is based on the discount rate that you're willing to give it. And stuff like that is what's the required growth rate based on the current share price. So that's the reverse DCF. And if we leave everything the same, so we're looking at after 10 years, we go to inflation, we have a discount rate of 10 percent. The market is saying that the company needs to grow its earnings per share at twenty seven point five nine percent over the next ten years to justify that cost. And look, that's probably not outrageous given [00:16:58][62.7]

Bryce: [00:16:59] the unrest in [00:17:00][0.6]

Alec: [00:17:00] the last few years, although sustaining that kind of growth rate over 10 years. You know, when you're double your size, triple your size becomes harder. So that's one valuation method. Do you have a Morningstar one for this? [00:17:12][12.8]

Bryce: [00:17:13] I do have a Morningstar one for this. And they've come in at a fair value of twenty four point two cents. [00:17:19][6.2]

Alec: [00:17:20] Morningstar I have been bullish. They've been above the share price for every company that we've looked at so far. [00:17:25][5.4]

Bryce: [00:17:25] I think there's a couple that we'll dig into in this series that they don't. But yeah, so far, whoever's running the numbers, the guy who's feeling good, he is. [00:17:35][9.4]

Alec: [00:17:35] He is. [00:17:35][0.1]

Bryce: [00:17:36] So yeah. I mean, they reckon that it's trading. Well, we know that it's trading within a range that they think is fair value. [00:17:41][5.1]

Alec: [00:17:41] Yeah. So then the other way that we have spoken about is this Roger Montgomery method where you look at. The book value per share, and then you look at how the company earns a return on equity based on that, and then you value the company based on those metrics under that model. Athans current book value per share is three point seventy nine and its return on equity as an average over the last four years is 32 percent. So the last thing you need to do is look at how much of its earnings it paid out in dividends and how much it retained. We can say that last year it paid out 19 percent of its earnings as dividends. So you plug that in and you look in Roger Montgomery book for the tables that he's got there. And you can see that the fair value that it spits out is twenty four dollars and thirty four cents, which means the company right now is a little bit cheaper than expected to give you a comparison if you plugged in all the same numbers, but you dropped the return on equity from 32 percent that it's averaging over the last four years to twenty five percent. All the other numbers stay the same. You get a fair value of seventeen point seventy seven cents. So the value drops about seven bucks. And that's because over time, if it's earning a lower return on equity, then every year it's going to be compounding at a lower rate. [00:18:59][77.8]

Bryce: [00:19:00] Makes sense. So in terms of outlook Ren, they gave us a full year guidance in November that came through. Well, it was an improvement. I guess they're expecting year on year growth to be between 34 and 38 percent. So to your point, I think we've worked out the need. Twenty seven percent growth over 10 years or whatever it was [00:19:18][18.5]

Alec: [00:19:19] on track until you won, you [00:19:20][1.6]

Bryce: [00:19:21] won. And obviously that's being driven by their content relevance service, which is one of the biggest parts of their business. So outlook is pretty positive. I'm very keen to keep an eye on this one. I love all the wax companies, especially as their home grown Aussie tech success stories. And this one, I'm sort of kicking myself that I didn't pick up on it and do this sort of research two years ago or so, but always opportunities [00:19:43][22.1]

Alec: [00:19:44] you just got to on the next one. [00:19:45][0.9]

Bryce: [00:19:45] Exactly. So where do you think Ren anything else from you before we talk about circle of confidence, know well, where does this fall in your circle of confidence? [00:19:51][6.0]

Alec: [00:19:52] I think obviously we don't understand. Well, I don't understand the intricacies of I but I do understand the intricacies of just labelling data. And so I guess the question then becomes, what's the value of that? Like, what the business actually does is quite simple to understand. So I think, yeah, it's within the circle of competence. There's obviously some questions around how does that translate into creating value for those other companies that are selling into and stuff like that. But it's not like you need to understand the intricacies of IBM machine learning to understand how the bought the value of the data that Apple is creating. [00:20:27][34.7]

Bryce: [00:20:27] Yeah, yeah. No, I agree. I think in terms of understanding what their business model is pretty straightforward. And to be honest, it's something that's pretty interesting to me as well. So I'm sure you've spent a bit more time digging into it. We'd get to understand it fairly well, so we'll leave it there. Always good chatting stocks Ren will continue next episode. Sounds good. [00:20:27][0.0]