Rarely do I still own, let alone write up, stocks that have doubled in a couple of months, so Gan Plc (LSE: GAN) is very much an exception for me. But what would you pay for a software services company growing revenues >100%; where >50% of those revenues are already recurring in nature; where the end-markets the company services are growing near triple digits and are many years from saturation; and where said company’s customers not only have a dominant market share at the moment, but that market lead is actually increasing rapidly? What if I further told you that along with the gargantuan top line growth, this company is already profitable, and likely to become highly cash-generative in the next couple of years, as the investments generating today and tomorrow’s growth have already been made?
A company like this – rapidly growing, with mostly recurring revenue; a key provider to the market leaders in an industry likely to grow secularly for many years; and already profitable and cash generative – would likely garner a high multiple of revenues, if not an astronomical multiple of today’s earnings. Indeed, many ‘SaaS’ businesses that provide just the high revenue growth – if not profits – are rewarded with heady (7-10x) revenue multiples.
GAN PLC (London: GAN) – a B2B provider of software and services to the real-money gaming industry – is just such a situation that, despite checking most all the boxes above, and indeed rallying aggressively this year, remains – to my mind – pretty cheap (if not as cheap as it was). Consider, very briefly, the following:
– GAN will grow revenues >100% this year, and has guided to ongoing high growth for many years;
– the nature of GAN’s service proposition – it develops and maintains the back-end software for casino and sportsbook clients, earning both a licensing fee on delivery but more importantly a revenue share of subsequent business on that platform – means the growing revenue stream is extremely valuable (incremental revenue dropping down onto the same platform at very high margins);
– GAN’s clients collectively have ~50% and 65% market share in the respective online sports betting markets in New Jersey and Pennsylvania – both dominant positions in the two nascent markets that have opened up thus far;
– Sports betting legalization in the US has only occurred in <30% of the US adult population (and only in NJ and PA has the business grown to scalability). It stands to reason that the other states simply follow suit since – as we shall see – NJ and PA sports betting is quickly becoming a gold mine for state coffers;
– GAN is levered not just to sports betting adoption but to the growth of online sports betting – already near 90% of total wagering in NJ and PA in very little time – as well as the cross-sell opportunity available when sports betting clients gamble online using the same app (as GAN’s clients invariably offer). This added revenue stream is often higher margin (given greater profitability for the clients in online casino that online sports betting).
I could go on. But despite the plethora of near-, mid- and long-term catalysts, GAN shares are still available today at – to my mind – the reasonable price of 23x EV/FCF. Looking out to FY21 – just 18mos away on a forward basis – and I think the shares are trading at just 3.5x EV/sales, 9x EV/EBITDA, and a verifiably value multiple of 11x EV/FCF – and this despite a number of conservative assumptions regarding the pace of current growth (that we will discuss).
Why is GAN so cheap? Well, today the market cap – despite the rally of late – is barely 160mm GBP. Furthermore, it is listed on the London junior exchange (the AIM), despite the fact that essentially all the business going forward is US based. Thus, it is (for now) a microcap on a forgotten exchange, divorced from many of its most logical owners (US investors). This, however, will change, likely in 1H next year, as the company is well on the way to relisting in the US. That change alone – bringing with it the attendant eyeballs of US based investors more familiar with the end market dynamics – is probably good for another 30% upside in the stock.
Longer-term, what is the right price for a business like this, growing like this? While a lot depends on how quickly the industry continues to scale, and if or how GAN is able to replicate their position in other states that open up, I feel comfortable contending this shouldn’t trade much lower than 5-6x FY1 revenues. Even just on my ’21 numbers, that implies 40-60% upside from today’s price, and earnings multiples in the 13-16x EV/EBITDA range – hardly aggressive for a company likely to grow the topline at 25%+ for many years.
To get comfortable with the GAN story, you need to believe three things:
1) The rapid growth in US sports betting will continue apace in NJ and PA;
2) GAN’s key client, FanDuel, will be the primary beneficiary of this ongoing growth, maintaining its current market lead; and
3) GAN will be able to maintain its economic share within the FanDuel ecosystem.
There are many other angles to the GAN story (upside from new states coming online, like Michigan, which could be up and running by March Madness; or the cross-sell opportunity in online casino). But such is the explosive growth in the sports betting part of the business, I don’t really think you need to underwrite much more than the above to do well here. The key takeaway is, I believe at the current price you are buying the NJ + PA sports betting opportunity alone – irrespective of other states that may legalize, cross-selling into casino, and any additional non-recurring revenue lines – at <20x recurring free cash flow out one year, for a business growing >100%. That’s pretty juicy…let’s see how.
The growth in US sports betting: this is what a gold rush looks like
Americans, clearly, have been waiting a long time to legally bet on sports (despite this being legal in many other developed market jurisdictions worldwide for many years). I guess it should be little surprise that the two legalized markets currently open for business – New Jersey and Pennsylvania – have been growing like weeds, but still, perhaps the pace of adoption has been eye-opening.
Here is the growth in sports ‘handle’ (total bets wagered) in New Jersey, on a monthly basis, over the last couple of years – both physical and online:
Sports betting has been legal in NJ for barely two years and already – if December numbers keep up the current 70-80% annual growth rates – the NJ market will be basically as large as the Nevada sports betting market (a state where sports betting has been legal for two generations). It is important to note also that the vast majority – almost 90% – of wagers are being placed online. Since GAN provides the software infrastructure to allow this, clearly this is a key part of the GAN investment story.
Pennsylvania really only started allowing sports bets in earnest this summer, but the story there is similar – in fact, adoption in PA has been even more rapid that in NJ:
The faster adoption curve in PA is likely driven by the leading operators – principally FanDuel, GAN’s key client – applying learnings from their two years of experience in the neighboring NJ market. In any event, it seems quite likely that PA simply grows to its adjusted per-capita size in line with wherever NJ ends up, over time (the two states’ gambling behavior should be very similar, after all). It is hard to argue that other states – such as Michigan, which just legalized – won’t see a similar rapid adoption pattern if or when they come to the party.
Where, then, could per capita sports gambling end up? Well, in NJ at the moment, the likely FY19 handle implies around $670 in wagers per adult resident; the PA number – lower as it is given the more recent launch – is still much lower at $140 per adult resident. So clearly the PA number has plenty of catching up to do, but I think the NJ number too – and other states that soon follow suit – are likely to grow strongly for many, many years.
For context, in the UK – the most mature sports betting market that could be considered comparable to the US – 2017/18 saw around $860 per adult resident gambled on sports, so clearly a good deal higher than NJ is today. But consider also that in the UK, the vast majority of bets still take place physically in betting shops, whereas the US is already almost entirely an internet-enabled market. This likely has consequences for betting volumes, since the ability to drive volume is exponentially easier over the internet than in traditional retail. Secondly, the largest share of bets by sport in the UK are still largely horse-racing; in the US betting is clearly dominated by the three major sports (football, basketball, and baseball) – all of which are 3-4 hour endeavors, allowing for a plethora of in-game betting permutations. It stands to reason, then, that over time the per-capita betting quantum in a fully-developed US market could be much higher than in the UK, for a combination of these reasons. Of course, with NJ betting volumes still ~30% below the levels of the UK today (yet still growing 70%+ per annum), it seems likely even the more mature NJ market will continue to grow aggressively for years to come.
Modeling the opportunity in NJ and PA
Let’s look at what could happen in the US, just in the two states where online betting has progressed to the point of meaningful scale (West Virginia is de minimis and Indiana just began operations a couple months ago). As mentioned earlier, I believe these two states alone basically underwrite the investment proposition on their own, and everything else GAN brings – other states covering the other 70% of the US population; cross selling opportunities into casino, which carry fatter take rates; the other revenue streams in the business – are thrown in basically as free optionality.
See below for how I have sized the market out to FY21E in NJ and PA:
Clearly I have made a number of assumptions to get to the conclusion that we are paying around 20x FCF next yeear and 18x in 2021. Here are some of the major ones:
– Sports handle: this has been growing astronomically in both states, but I conservatively assume annual growth retreats massively (to 25%/15% in NJ, and 259%/29% in PA – PA needing to catch up as it legalized more recently) once spend per adult capita reaches UK levels (even though, as discussed, I think the American market can be much juicier on a per capita basis than the UK);
– GGR: this is a function of both ‘hold’ (the amount of wagered $$ that the operators keep, which fluctuates with luck somewhat) as well as in-state taxation rules. I have tried to be conservative on both measures;
– GAN client market share: again, I have assumed GAN’s clients – principally FanDuel – actually lose 10-15pts of share over the next 2yrs, EVEN THOUGH the opposite has been occurring in recent months (GAN clients extending their lead in both states);
– GAN take rates: this is a bit more of an estimate. The company discloses take rates for different services and games varies quite a bit in the 5-25% range (and also by customer). I have assumed very much near the lower end of the range (6%) as I believe sports is less profitable than other games, and the agreement with FanDuel is for 5yrs and thus probably a lower share % given the assumed absolute growth in the market;
– profitability and cash flow: GAN management suggests incremental profitability is 60-70c in every dollar from here. This makes sense if you think about the business model – the platform has been built, the tech infrastructure exists, there are simply more gamblers and more transactions per gambler now – but even so I assume the lower end of this guidance. I also assume 80% FCF/EBITDA conversion (despite the likely much higher conversion as GAN can pay suppliers slower than they get paid, carries no inventory, and has no debt so there is no interest cost; GAN showed 120% EBITDA/FCF conversion in 1H’19).
Even given all of the above, I think the market is according barely a 20x FCF multiple to the two extant states at the moment (assuming FY1 numbers are steady-state recurring, itself too conservative). That seems fairly crazy to me given that GAN is clearly partnered with the market leader (FanDuel); that other states will definitely open up and GAN is, if not the front runner to provide back-end services to FanDuel in those states then at least in with a very good shot of doing so; and that of course the above completely ignores the non-sports opportunity (real money gaming) or the other business lines at the company (simulated gaming, and patent licensing). Getting the existing, high-growth recurring revenue business at a reasonable multiple, and all the other optionality for free? I like me those odds…
The event setup adds intrigue
Aside from valuation and the mid-term growth prospects, there is a technical attraction to this trade too. GAN is listed in London on the most junior exchange now (the AIM), where it trades barely $1mm a day. As such, almost no US institutions can even look at the company, nor trade it. Management realizes the limitations of such a position as the US market takes off, and is in the process of relisting in the US – something that should be accomplished in the next 3-4months, I believe (they have already appoint FBR Riley as advisors to captain the process). Such an outcome – aligning the company with its most natural investor base – the US, where the vast majority of the business actually resides – can only but help increase exposure and, given the attractiveness of the setup, the stock’s valuation. And since it is still such a small company, only a handful of funds need to take an interest to really move a stock like this much closer to fair value…
What’s it worth, then?
I haven’t gone into detail in this post regarding all the other business lines at GAN, nor provided a more holistic top-down model for the out-years (assuming the pace of legalization continues). It is not that I think that work is unnecessary, or not valuable; its just that at current prices I feel I am buying what is near-term, measurable, and highly valuable at very reasonable prices such that I can let the mid-term growth story play out. For the sake of completeness, however, I think this business can eminently trade at closer to 5-6x FY21E revenues – which, including all the other business lines, I have at around 40mm GBP using similarly conservative assumptions as above – and implies a further 40-60% upside from here. Of course it should be noted I am not strictly valuing the business on a revenue multiple, as I expect GAN to be highly profitable and cash-generative through this period given their business model. But the recent Diamond Acquisition/Draft Kings M&A transaction highlighted the relevance of revenue multiples for high-growth online gaming businesses (indeed they paid >4x FY1E revs for a much slower-growing, lower-quality, platform provider, SBTech, as part of that transaction, helping benchmark my price target) so for a short-hand approach, I believe this is the right postcode for the valuation. Of course it is entirely possible that FanDuel – now part of the enlarged PaddyPower Betfair group, Flutter – decides to in-source their platform provider via a takeover before the market has a chance to grow too rapidly. Of course I believe that could only be consummated much closer to fair value…that is closer to 300 GBP/share than where we are today (185 GBP).
One final word
This is a microcap and small stocks have larger risks than big stocks, clearly. The main one here is customer concentration: the relationship with FanDuel is clearly massively beneficial but it could already represent >30% of GAN revenues and clearly that number will just go up. As such there is always attendant risk around disintermediation and substitution. I am somewhat comfortable that is not a pressing risk for now – not least because the NJ/PA contract runs until 2024, and because FanDuel proactively re-upped with GAN when they entered Indiana recently – but it is definitely something to keep a close eye on.
Disclosure: long GAN
22 thoughts on “GAN Plc: explosive mid-term growth + attractive technical setup still not priced in”
Pingback: Links of The Week: January 4, 2020 - Value Never Sleeps
Do you have a view on how Flutter’s merger with The Stars Group affects your thesis? Is it a positive, bringing Fox Bet into the fold? Or is it a negative, as Fox Bet uses another provider or owns more of their own back-end?
no real strong view either way at this stage. it may be another 6mos before the Stars Group merger is approved.
Regarding free cash flow: how do you explain that in H1 2019, excluding the 3.0m of one-off licensing fee, “Operating cash flow before movement in working capital and taxation” grew only by GBP0.7m vs. H1 2018 despite the strong revenue growth? Why is the high operating leverage not materializing here?
hi there. you need to look at the YoY comps to understand it. in 1H’18, revs were 4.6mm vs 1H’19 11.3mm, so +6.7mm YoY; in 1H’18 FCF was -2.6mm vs 1H’19 +3.7mm so a swing of +6.3mm, basically all of incremental revenues. Even stripping out the 3mm of licensing fees from the incremental revs and cash flow still shows very high incremental cash conversion. This is because capex dropped appreciably (they finished building out their multi-year investment into the product, including the Sofia, Bulgaria offices); and because they see a structural working capital benefit as they scale (they pay their suppliers slower than they get paid).
So, i’m fairly comfortable in forseeing ongoing high FCF conversion rates for the business (though clearly not as high as 1H’19 given the licensing component).
I see your point but in order to put working capital movements out of the equation, let’s just look at the Clean EBITDA then.
H1 2018: -0.3m
H1 2019: +3.0m including the one-off license fee, or 0.0m excluding it…
What am I missing here? I assume that there is very little costs attached to the one-off fee. Are you adjusting for the “significant expansion of engineering resources commenced in H2 2018”?
two points: 1) i don’t want to put working capital movements out of the equation since, as a service/platform business, this is a structural cash creator as revenue grows; and 2) yes you are right, i am implicitly saying that a much larger portion of core EBIT will now drop through to cash flow given the business was investing previously and is seeing an inflection in revenue growth (that is not really present in the recent historical figures). if you believe the business model is valid and has reached the point of scaling, this is achievable (as mgmt has guided) but i do agree they need to prove it out in the coming quarters/year.
Jeremy – Really well written. Curious on 2 assumptions –
1. For PA, you are using GGR of $69mm. However, if you go to this website and add up the months, the GGR (calculated as gross revenues LESS taxes) is ~$53.8mm https://gamingcontrolboard.pa.gov/?p=304
2. Same comment for NJ – this shows total revenue at $299mm, but that is before taxes so GGR is probably 15% below that figure vs. your $312 (https://www.nj.gov/oag/ge/docs/Financials/PressRel2019/December2019.pdf)
hi there! thanks for the questions:
1) i added in my assumption for Dec numbers, since when i wrote the article this was just 11mos.
2) it depends on what take rate you are assuming. I assume 6% take rate from GGR, which is functionally equivalent to 7.5-8% of NGR (depending on taxes/bonuses, your point). My understanding is GAN gets a 5-6% take rate from GGR as I define it, as this number is closest to what GAN defines as ‘Gross Operator Revenue’ (and their historical take rate there in 1H was around 7.5%
Do you know why Dermot Smurfit has sold 1m shares over the last months, or any context around those sales? I can’t imagine he really needed that cash given his wealthy family. Given business prospects are now supposed to be strongest than ever, it is not the strongest vote of confidence.
hi there – i don’t know why. the only consolation is its a tiny portion of his/his family’s total stake so if he was really bearish on the business he would sell much more. i believe he may have had some tax obligations or wanted to purchase a house. recall he is the nephew of the rich Smurfit guy, so he may not have direct access to his family’s money.
Jeremy- in the super bear scenario where US major league sports don’t start back up this Fall, would GAN still be FCF positive?
yes. in the meantime they are making lot of $$ on their online gaming business since a lot of people who can’t go to physical casinos and/or bet sports are gambling online instead. Online gambling is actually much higher margin (better take rate) than sports betting…so while I do think numbers of course take a hit, I think Gan comes through this much better than most (including the market) think
Hey Jeremy, thanks for the writeup.. was interesting to listen to your podcasts lately.
Audited Net Revenue as per their 2018 annual report is around 10.5mm, but their public announcement for Nasdaq listing from 25th March says “Net Revenue of $30.0m (2018: $14.0m), an increase of 114% on 2018”.. that’s suggesting 35% more revenue for FY 2018. Even other numbers like net loss/cash/assets in the latest announcement do not match the 2018 FY report. happy to be corrected if I’m wrong here. But if right, then such egregious errors don’t shed good light.
take care and stay safe in the crazy times.
hi there – thanks for reading/listening. i think you are confusing GBP with USD. the latest numbers (in the trading update) were in USD (because the co is redomiciling to Bermuda and will soon list on Nasdaq). previously disclosures were in GBP. that probably explains most/all the discrepancy
aah that explains it all..thanks mate
Pingback: What I’m buying now – part 2 | Raper Capital
Congrats on that idea. Well done & thanks!
Hey, do you think the new deal with Penn would replace the lost opportunity from the FanDuel sports betting termination? Does this termination change your view on the company’s prospects?
my views of GAN have changed substantially since this was written – given the price appreciation along with many other things. you can read my latest thoughts here:
I understand the valuation point but the valuation would still work if the opportunity is the same. I am more interested in understanding if the Penn deal (considering the Barstool involvement) could offset the impact of the fanduel cancellation or does this termination severely limit the future growth potential of Gan period? I can’t access your post but considering it is from May then I doubt it covers this point.