Flutter Entertainment PLC (FLUT) 2021 Q2 法說會逐字稿

Good morning, everyone, and thank you for joining our H1 2021 interim results presentation. Jonathan Hill, our CFO, will be joining me shortly. Today, we're going to cover our strong performance in the first half and also going to do a deep dive into our U.S. business.

Given the rapid change in the U.S. market, we thought it was a good opportunity to review our strong fangle performance and also look at the trajectory from here. Given the focus on the U.S. market, we're going to have a slightly longer presentation than usual, but I hope you find the materials informative and useful. Later this morning, we'll be hosting an analyst call, which I hope you'll be able to join.

Slide 3 shows the agenda for our presentation. I'm going to start with a brief H1 summary before Jonathan takes you through the financial performance of the group. I'll then add my usual operational update on the 3 divisions that make up what we call group ex-U.S. before walking through a deep dive on our U.S. business.

Starting on Slide 4, you'll see the first half has exceeded our expectations, both in terms of financial performance and operational progress. We continue to integrate our businesses while delivering strong results. And I'd like to take this opportunity to thank our team of nearly 15,000 colleagues for their hard work and commitment as we expand our business globally. We are seeing excellent collaboration across the group by teams that in many cases are yet to even meet in person. Hopefully, that day is coming soon. We continue to support the well-being of our colleagues during the pandemic, and I'm proud that we have done so without the various government support schemes and offer.

In H1, we have continued to leverage our scale and diversification to deliver a very strong set of earnings. In the U.K. and Ireland, Connor and his team have made good integration progress, including putting in place centers of excellence that are driving new product innovation and facilitating cross-brand insight sharing. I've been delighted to see how well our brand teams are working together, though I must admit I still detected divided camps in the Euro 2020 final, which ultimately was a great result for the bookmakers. In our international division, Dan has done a great job of stabilizing our player base, while putting in place the investments needed to improve the quality and quantum of our future earnings.

In Australia, Sportsbet's performance has been nothing short of outstanding, with Brian and the team now increasingly confident that we have experienced a permanent step change in the scale of our business there. And in the U.S., we have maintained our leadership position and have extended our scale advantage over our competitors. Under our Interim CEO, Amy Hau, the team has remained laser-focused on driving the business forward. It is remarkable to think that the U.S. has become our second biggest division in revenue terms in the last quarter with Team USA already starting to chase down the U.K. and Ireland. This time last year, they were our smallest division, really is amazing.

Before going further, just a couple of brief updates on Kentucky and Fox as I know you'll be keen to get an update on both. With respect to the Kentucky case, we will shortly file our application to the U.S. Supreme Court, requesting the court to review the Kentucky Supreme Court ruling. We hope to learn during Q4 whether the court will agree to do so. And in the case of our legal dispute with FOX, an independent arbitrator has now been appointed. The hearing in that arbitration is likely to be in the first quarter 2022. And as we have said previously, we won't be commenting on the process publicly until it is complete.

Slide 5 shows how the scale and quality of the group continues to improve. We've used H1 2019 as a basis for comparison, as it represents what our player base stood like before the structural changes brought about by the events of the last 18 months. In the first half alone, over 7.6 million average monthly players or AMPs, engaged with our brands with good 2-year growth across all regions. While growth in the U.S. will likely capture the headlines, it's important to look at the growth in both Australia and the U.K. and Ireland, given that they were perceived to be more mature online markets. The geographic mix of our revenues is also evolving and improving in terms of quality and sustainability. Australia and U.S. combined now account for 40% of group revenues, with the U.K. and Ireland share revenues declining despite good ongoing growth there.

Looking at the chart on the right-hand side, our geographic split feels more balanced now than it did 2 years ago. Our unregulated market exposure has also declined over the last 2 years. 90% of our revenues now come from regulated markets, and no single unregulated market accounts to more than 1.5% of total group revenues at this point. As we grow further, we expect our regulated market exposure to continue to rise.

On Slide 6, I want to provide an update on how we are taking the lead when it comes to safer gambling initiatives. We continue to improve our safer gambling framework, developing initiatives that are tailored to market-specific requirements. During the period, we've increased our investments in dedicated teams, resource and technology in each division. Improvements in our algorithms have enabled us to step up the number and quality of individual interactions we're having with customers. Campaigns to raise awareness of our safer gambling tools are resulting in a noticeable uptick in the application deposit limits and other tool usage. In the U.K., we continue to refine our affordability triple step framework, which we believe strikes the right balance between protecting those that are vulnerable without infringing on the freedoms and the vast majority of our customers. We have shared our thoughts on this approach with the government as part of our response to the gambling at review, and we anticipate the next phase of rollout will go live before the end of the year.

Wherever we operate, we're committed to introducing measures which instill the highest standards of customer protections, working with regulators, the wide industry or even unilaterally as we've done recently in Ireland with a self-imposed credit card ban, amongst other measures. On the right-hand side of the slide, you'll see how the changes we are making are translating to a more sustainable recreational customer base, with 40% growth in AMPs translating into revenue growth of 30%. Revenue per customer is, therefore, trending down, demonstrating we believe the more recreational focus of our brands. We know we can never do too much in this area and look forward to making more progress in the second half.

And with that, I'll hand you over to Jonathan to take you through the financials.

Thanks, Peter, and good morning, everyone. Before I go through the numbers, I'd like to reiterate my thanks to all of the team for their fantastic efforts in delivering this really strong set of first half numbers.

Starting on Slide 8. On a pro forma basis, we delivered revenue growth of 30% in the first half. This compares with growth of 28% in full year 2020. So it's great to see that we have been able to maintain strong momentum across the business. As we flagged this time last year, COVID had very contrasting implications for our sports and gaming businesses, and this is evident in the relative growth of both this year. Sports revenues were 57% higher, helped by a return to more normal sporting calendars and also good player volume growth. Meanwhile, we maintained our group gaming revenues at the high levels achieved last year, thanks to good growth in both the U.S. and in the U.K. and Ireland, offsetting lower poker revenues.

Adjusted EBITDA was GBP 597 million, 12% lower than 2020, reflecting a number of key factors that I will cover in coming slides. Adjusted EPS declined by 40%, reflecting both lower profits and our increased share count following the buyout of the FanDuel minority investors at the end of 2020. The group continues to turn profit into cash at a high rate. And we ended the half year with net debt of GBP 2.7 billion and leverage of 2.3x.

I won't spend too long on Slide 9, which outlines the statutory income statement for the period. Clearly, the high year-on-year growth reflects the benefit of a full 6-month contribution from the Stars Group in this year's numbers. versus less than 2 months in H1 2020. The one item I would call out on this slide is separately disclosed items of GBP 321 million, which were primarily non-cash items, including amortization of acquired intangibles of GBP 276 million, mainly from the TSG merger.

On Slide 10, we show our AMPs by division since Q1 2020. We strongly believe that player volumes are a key indicator of the underlying health of our business. As they show high effectively, we are delivering against our customer acquisition and customer retention strategies. At an overall group level, AMPs grew 40% year-on-year to over GBP 7.6 million. Year-on-year numbers are the most relevant comparator given the seasonality of our different divisions. As an example, Australia has a huge spring racing carnival primarily in Q4.

Turning to divisions. In the U.K. and Ireland, numbers increased 44% in H1 to GBP 3.3 million, benefiting from record Cheltenham and Grand National activity as well as the rearranged European football championships. Player retention in Australia has exceeded even our most optimistic expectations, with Sportsbet AMPs up by 52%. In international, which faced the most challenging COVID-related comparators, our performance has been encouraging with player numbers down just 3% in H1. Our investment in player retention and some ongoing tailwinds from localized lockdowns and parts of Europe drove this performance. In the U.S., we grew our player base by 166%. We executed well throughout the half, but particularly at the Super Bowl, where only Tom Brady delivered a better performance.

On Slide 11, you can see how this customer growth led to group revenue growth of 30%. This revenue growth was delivered even with sports net revenue margins 160 basis points lower. While sports results were better than expected, they were 80 basis points less beneficial than in H1 2020. And we also made additional investment in player generosity, reducing our net revenue margin further. In the U.K. and Ireland, our online business grew revenues by 37%. This growth was partially offset by ongoing shop closures in our retail estate. And as a result, overall growth was 30%. In Australia, net revenue margins were broadly flat year-on-year, with revenue growth matching staking growth at 27%. The lower revenues per customer reflected the sports mix, particularly in Q2, with only racing on for the vast majority of Q2 2020.

In international, as anticipated, the significant uplift in customer activity in Q2 last year led to revenue declining year-on-year. When combined with the impact of regulatory headwinds and our investment in additional generosity, this decline was 11%. This revenue reduction, however, was around GBP 50 million less than we originally anticipated. In the U.S., our revenue performance reflects our continued market leadership position, which Peter will go through in detail later, aligned to excellent sports results, especially in Q2.

Slide 12 really demonstrates the benefits of having a diversified portfolio of products. As you can see from the trend lines, which represent year-on-year growth for our products, COVID had a profound impact last year. You can see the 66% gaming increase in Q2 2020, driven by the poker uplift during the first COVID lockdowns and then the gaming reduction in Q2 2021. Conversely, sports fell 7% in Q2 2020, but rebounded by 75% in Q2 '21. For me, it highlights how robust our revenue mix now is.

Slide 13 provides a year-on-year EBITDA bridge for H1 for Group ex-U.S. A couple of key points to drive here. Firstly, the excellent performance and operating leverage achieved in both UK&I online and Australia. These businesses delivered a combined EBITDA increase of GBP 208 million year-on-year, with our Australian EBITDA margin, in particular, benefiting from merger synergies.

Secondly, the year-on-year decline in international, while significant, was less than originally anticipated, and I'll touch on this in more detail on the next slide. Total synergies for the half were GBP 52 million, with the majority of these savings accruing in Australia and in corporate. Overall, we are running marginally ahead of schedule in terms of the timing of cost savings though our overall synergy guidance remains unchanged at GBP 170 million. Overall, adjusted EBITDA declined by just 2% despite the Q2 poker spike last year, ongoing retail closures and adverse regulatory changes.

Now turning to Slide 14. Given the number of moving parts within international, we have provided a year-on-year bridge. Having rebased for FX movements, we adjust 2020 for 3 key items previously flagged: the first was the significant uplift in poker and casino activity within PokerStars in H1 last year, which we said gave us a one-off revenue benefit of around GBP 205 million equating to EBITDA of around GBP 155 million; the second was the enhanced compliance standards that we put in place in TSG, which we flagged would cost us GBP 65 million with half landing in H1; finally, there is the impact of new German regulations, which will cost us GBP 50 million in a full year and GBP 25 million in H1.

We then have the investment we are making in the business to drive growth. We invested an incremental GBP 88 million year-on-year in H1 to drive our growth in new markets and to reinvigorate the PokerStars brand. Total OpEx in H1 was similar to H2 2020 levels. Peter will talk to you more about how these investments are improving the quality of our business, but you can already see from the chart how it is starting to drive growth. Underlying EBITDA growth in the half was GBP 105 million, of which we think around GBP 38 million was an ongoing benefit of lockdowns. Assuming no further lockdowns, we don't anticipate a repeat of this tailwind in H2.

Slide 15 provides an EBITDA and EBITDA margin summary. Flutter generated EBITDA of GBP 597 million in H1 or GBP 684 million, excluding the U.S. business. Even with our retail business shut for much of the half, the adverse regulatory changes in international and the long-term investments we are making, we generated an EBITDA margin of 28% in Group ex-U.S. In the U.S., our investment losses increased from GBP 19 million to GBP 87 million. While we delivered a 159% increase in revenues to GBP 652 million, we invested heavily in sales and marketing, spending nearly GBP 300 million in H1 to drive exceptional levels of customer acquisition, which Peter will talk later.

We also invested in operating cost to support this scaling business. Ultimately, we believe our group scale is giving us a competitive advantage in the structural shift to online in our sector. This is evident in H1 alone, where we increased our sales and marketing investment globally by around GBP 300 million year-on-year to GBP 728 million. When you couple this investment with our product quality and brand reach, we believe it positions us very well for future growth.

On Slide 16, you will see we generated adjusted free cash flow of GBP 358 million in the half. This is lower than the prior year, primarily due to our higher U.S. investment losses and after the significant working capital benefit in H1 2020. However, comparing operating profit to pretax adjusted free cash flow, we converted profits into cash at 95%. I won't go through all of the individual line items here, but will just call out a few of the bigger ones.

Firstly, we incurred a higher corporate tax charge given the change in the geographic mix of our earnings. Secondly, we paid GBP 71 million the State of Kentucky to settle a bond relating to the Kentucky Supreme Court ruling from December. Thirdly, our Employee Benefit Trust acquired GBP 89 million worth of shares relating to FanDuel incentive schemes that were put in place at the time of the original acquisition and designed to incentivize value creation. We anticipate a further cash cost in H2 at GBP 92 million relating to these schemes. And fourthly, we acquired a majority stake in Junglee Games, the #2 online rummy business in India for an initial GBP 51 million. As a result, we finished the half with net debt of GBP 2.7 billion.

Slide 17 provides an up-to-date overview of our debt position following our recent refinancing. In July, we repriced our debt and upsized our TLB dollar loan to repay our most expensive debt, the $1 billion of 7% senior unsecured notes. This will result in annual interest savings of around GBP 50 million. These savings are incremental to the cost synergies of GBP 170 million previously guided. Leverage at 2.3x remains broadly unchanged following the refinancing. We continue to be committed to our medium-term target of 1 to 2x leverage, and the Board will review the group's dividend policy once leverage is within the targeted range.

Slide 18 provides a summary of the earnings guidance we have shared with the market this morning. Given our strong first half and good underlying momentum, we now anticipate that group ex-U.S. will generate EBITDA of between GBP 1.27 billion and GBP 1.37 billion in 2021. In the U.S. we now expect to generate net revenue of between $1.8 billion and $2 billion this year and to incur an investment phase loss of between GBP 225 million and GBP 275 million. This level of loss reflects our continued scaling of the U.S. business in H2, particularly through the all-important start to the NFL season. It also reflects our expectation of 2 new states opening during H2. Continuing our disciplined investment is the right thing to do for the U.S. business.

Overall, I'm delighted with the performance of the group in the first half, with plenty to look forward to in H2. And with that, I'll hand back to Peter.

Thanks, Jonathan. I will now provide an update on key developments across the group ex-U.S. Starting on Slide 20 with U.K. and Ireland. One of the goals upon completion of our merger was to manage integration to protect momentum within the underlying business. The numbers Jonathan has talked you through demonstrate clearly that we have managed to maintain momentum, and be rest assured, though, that we're making good progress on integration at the same time. Our U.K. and Ireland teams kicked off work in H1 to create sensitive excellence across key areas within the business, namely people, customers, pricing and risk management and technology. These centers are supporting our brands who shared insights, technology and tools so we can provide our customers with best-in-class products.

In bringing together 3 powerful brands, we've developed a club and country ethos that will see our brand focused teams supported by central support functions. Ultimately, what we are hoping to create is a structure that allows our brands to work together to become something greater than the sum of their parts. Our complement to compete strategy will allow us to reach the entire market and appeal to all types of customer segments through a truly differentiated product proposition. We are pleased with progress to date. Our 3 brands attracted more than 3.3 million AMPs in the first half, up 44% year-on-year.

Slide 21 highlights the benefits we're already seeing from cross-brand collaboration. Using insights and Sky Bet's Request a Bet product, both Paddy Power and Betfair improved their equivalent Bet Builder products during H1 developing more intuitive customer journeys. These products are hugely popular recreational customers and have a positive impact on our structural margins, too. Combined with targeted marketing campaigns, the volume on these types of bets doubled in the first half compared to last year. Meanwhile, SBG has been improving its in-play experience. bringing more markets traded in-house and using insights from the PPB team.

On the gaming side, the team has delivered an excellent performance with AMPs up 59% compared to 2019. Leveraging the success that the prize machine free-to-play game had on customer engagement for Sky Vegas, Paddy Power and Betfair launched Paddy's Wonder Wheel and Betfair's Prize Pinball at the end of 2020 and these products are continuing to drive customer momentum. 77% of Paddy Power daily punters played the wonder wheel during the month of Cheltenham, demonstrating the high levels of engagement of products such as this can generate.

We also enhanced our content during H1, with SBG now having access to PokerStars in-house and third-party content with Paddy Power due to get the same access later this year. You can see the live casino, spin-and-win roulette games in the graphic. This has developed in collaboration with a third-party provider exclusively for PokerStars. As part of the Flutter Group, Paddy Power and Sky Vegas can now also offer this game with the content already distributed to Sky Vegas customers via the group gaming network. What excites me about the progress in U.K. and Ireland is that we're only getting started with lots more collaboration to come.

Turning to International on Slide 22. We've spoken over the last year about how we are transforming this business to put it on a more stable footing and to correct historical underinvestment. I'm really proud of the work the team have done. The left-hand side summarizes what we're doing and how we're doing it. First, we want to revitalize the PokerStars brand and stabilize the player base. Secondly, we want to scale our casino business through a combination of cross-sell and direct acquisition. And thirdly, we wanted to develop a world-class sports betting offering for our international customers, leveraging the capabilities of the wider group. Achieving these goals will take time and investment, particularly with our increased focus on regulated markets. And while regulated markets tend to come with higher taxes and regulatory costs, they are generally more sustainable in nature.

Achieving our goals will also require us to supplement our capabilities and geographic presence through bolt-on M&A. Year-to-date, we have acquired 2 businesses that we believe will help accelerate our international growth. The first is Junglee, the #2 online rummy business in India with the newly launched DFS business that we believe has very exciting prospects in what is an increasingly attractive and high-growth market. The second is Singular, a sports betting technology platform that already supports our business which we believe will provide us with greater tech flexibility to enter certain new markets at pace. The right-hand side of this slide just details some of the specific campaigns we've invested in during H1 to achieve our 3 primary goals.

On Slide 23, you can see how the changes and investments we have made are starting to have an impact across the 3 product verticals. In poker, we've seen a positive response to our Neymar "I'm in" campaign, and our Sunday Millions 15-year anniversary event was the second largest tournament we've ever hosted. Another way we've been driving engagement is through our brand ambassadors and collaborators. Team PokerStars Pro, our group of ambassadors who represent the brand and live stream their action online on platforms like Twitch, a great way of delivering content and reaching poker enthusiasts. PokerStars had the leading market share viewers streaming live poker content during the half with an estimated 66% of all poker viewers tuning in to watch PokerStars sponsored content on the world's leading streaming platform.

In casino, we've seen a fivefold increase in direct casino AMPs at PokerStars versus 2019, helping casino become the largest product vertical for the international division. This has been complemented by ongoing very strong performance as a Adjarabet, the #1 operator in Georgia and Armenia. In sports, we're seeing improved cross-sell within PokerStars. Albeit it will only really be when we give PokerStars customers access to Flutter's global betting platform, that a more noticeable difference will likely be made. Overall, I'm pleased with the progress being made in international. We know that transforming this business will not happen overnight. But with assets at our disposal, I am confident we can get there.

Next, I want to touch on the performance of Sportsbet our market-leading Australian business because the progress Brian and the team have made there continues to be astounding. The charts on Slide 24 speak for themselves. The investment we've made to engage with traditional retail customers that migrated to our platform during 2020 has led to retention rates exceeding our expectations. We've continued to innovate in the business with the launch of our social betting product, [bet with mates], going down exceptionally well. And we've achieved all this while building on the benefits of the rapid integration of our 2 businesses and the delivery of material cost savings. This, in turn, has helped us to deliver a 192% increase in EBITDA on where we were just 2 years ago.

Perhaps most importantly, we have growing confidence that this step change in scale is permanent in nature. Our team in Sportsbet has learned a great amount about the changing shape of its customer base over the last 18 months, and we are very keen for them to share those learnings with investors. With that in mind, we're going to host a dedicated sports bet a virtual Investor Day on September 22nd and I hope you can join us for that event.

And so to the U.S., where I believe our performance truly stands out. The U.S. online market continues to evolve rapidly. With each passing quarter, we learn more about our customers and what the longer-term shape of the market will be. And I want to spend some time this morning laying out why we believe the U.S. business FanDuel is winning in America.

Slide 26 provides an overview of what I'm going to cover. I'm going to start by taking you through an update of the customer economics we're seeing after 3 years operating sports betting in the U.S. I'll explain what these customer economics mean for the trajectory of our business and what we believe is a credible path to profitability in 2023. We've really got our U.S. flywheel motoring at this point. And I'll outline why we feel this is absolutely critical to maintaining our leadership position long term. Ultimately, our success is based on the structural advantages we enjoy in the areas of product quality and the strength and reach of the FanDuel brand. I'll spend some time providing examples of what makes us different from our competitors in these areas. And finally, I'll touch on our recent market share performance before providing a glimpse of what is coming next. As you'll see, we believe the shape of the regulatory pipeline in America is really going to play to our strengths over the next 18 to 24 months.

Turning to Slide 27. On the left-hand side, we show how our presence has grown over the last 3 years. We are now live in 10 states following the addition of Michigan and Virginia earlier this year. And in total, we have now acquired over 2.2 million online sportsbook and gaming customers to date. 75% of this space has been acquired in the last 12 months alone.

On the right-hand side, you can see our cost per customer acquisition. Over the last 3 years, our CPA has been $291, which we believe is amongst the lowest in the industry, and almost certainly the lowest achieved at this kind of scale of acquisition. This is a gross CPA number and is calculated by taking all of our marketing spend on sportsbook and casino and dividing it by the number of customers acquired, there's no adjustments for retention spend. While having a low CPA is advantageous, what is more important is the gap between CPA and the expected lifetime value of the customers being acquired. We believe this gap has widened for FanDuel over the last 3 years as LTVs have grown due to a combination of better-than-expected player retention, structural improvements in our win margins and improved cross-sell rates to gaming as we have simplified customer journeys.

On Slide 28, we provide more detail on the return side of the equation. You'll see from the chart on the left that the customers who've been with FanDuel at least 12 months. The average return on investments in their first year of activity has been 1.2x the cost to acquire them. Or to put it in another way, these customers have contributed positively by the end of their first year. The chart shows how we make an initial upfront marketing investments in the acquisition quarter and that this investment is then returned to us by the end of the fourth full quarter post acquisition.

Equally encouraging is what we're seeing from the cohort of customers that have been with us for more than years. Revenue from these customers has been 11% higher in year 2 than in year 1. While these 2-year customers still represent only a very small fraction of the player base, given the nascent nature of our business, we are nonetheless encouraged by the trend. It's one we're familiar with. In Sky Bet and Sportsbet, our earlier customer cohorts have gone on to provide consistent revenue streams for multiple years and we'll be hopeful we'll see the same thing happen here over time.

So what does this mean for current and future profitability? The bridge on Slide 29 shows how the positive contribution from our existing customer base is funding the acquisition of new customers. Starting at the top, you can see that in the 12 months to June 2020, FanDuel generated $169 million in contribution, which as a reminder, is gross profit less marketing investment. The positive contribution came from approximately 550,000 sportsbook and casino customers acquired to that point, together with a positive contribution from our DFS and TVG businesses. The next 2 blue bars show how the contribution from these existing customers has grown over the last 12 months.

Our DFS and TVG businesses generated an additional $33 million in contribution. But it is the growth in our sportsbook and gaming contribution that really stands out. We generated an additional $200 million in these customer cohorts in the last 12 months, meaning that in total, FanDuel's existing business contribution grew by 138% to $402 million. This has helped to fund a huge level of customer acquisition in 6 new states as well as in existing states.

In total, FanDuel invested a net $311 million in the last 12 months, acquiring 1.7 million new customers. That figure includes both the cost to acquire these customers but also the contribution they generate post acquisition. While that resulted in a $78 million reduction in contribution year-on-year, we emerged from the last 12 months with the sportsbook and gaming customer base that is 3x bigger than that with which we started. There are 2 important consequences of this. Firstly, for others looking to compete against us who don't have the benefit of the significant positive contribution that our existing base is providing to us, the ongoing funding requirement is going to be very big, indeed. And secondly, over the next 12 months, should the spending patterns of our newly acquired customers prove to be similar to our earlier cohorts than the negative contribution of the last 12 months will turn positive.

On Slide 30, you can see that we're approaching a tipping point where the positive contribution from our existing U.S. player base will exceed the investment cost of new player acquisition. If you look at where we are today, 75% of our total sports and gaming customer base has been acquired in the last 12 months. Furthermore, we think the next 18 months are going to remain very big for customer acquisition. We expect to have to 9 new states to regulate in that time, including New York. But by the end of 2022, we believe that we will have reached the tipping point where the size of our existing customer base has reached sufficient scale that the positive contribution from existing customers should more than offset the cost of new player acquisition and our other operating costs.

And if that is the case, then now would imply that Flutter U.S. will generate positive EBITDA in 2023. And this would take the profitability in cash generation of the overall Flutter Group to a different level. Of course, this projection assumes that none of Florida, Texas or California go live before 2024. Should that occur, new player acquisition would increase substantially, potentially delaying the point of profitability. But that will be a nice problem to have because it would mean our addressable market will be even bigger again.

For us, profitability and when we achieve it remains an output. Our goal must be to accelerate our flywheel because doing so will ultimately dictate the quantum of our returns in the U.S. long term. We summarize the progress we've made in getting our U.S. flywheel turning on Slide 31. It shows how the superior scale of our business is compounding our advantage over time. As we generate more and more revenue, we can further invest in our customer proposition, which in turn should result in better customer retention and growth, leading to a virtuous circle of high revenue growth again.

I won't go through all the metrics provided on this slide, but I'm particularly pleased with the product and tech improvements we've made over the last 12 months, some of which are running through in the coming slides. We believe strongly that the long-term winners in this market will be determined by the quality of their product. Having a proprietary platform allows us to have greater control of our product road map, ensuring we can get the most important product features to customers quickly.

The FanDuel app has consistently been ranked #1 in third-party research, scoring highly in important areas such as speed and ease of use. The phased state-by-state rollout of our group betting platform between November last year and July this year is now complete and has increased the speed, scalability and reliability of our betting platform. This is particularly important for peak load periods on NFL Sundays. Our DFS accounts on wallet was adapted to support Sportsbook and gaming in 2020, which allows for a smoother cross-sell journey for existing customers. In states where this proprietary technology has been available since launch, we've seen doubled the cross-sell rates from DFS to Sportsbook in the first 8 weeks post state launch.

The final pillar of our sports product is pricing and risk management. I will touch on this area in more detail in the coming slides as it has been a key driver of FanDuel's outperformance to date. Our U.S. platforms are supported by the Flutter teams of nearly 5,000 experienced engineers. We believe this is a very powerful differentiator for us, where we develop new products in another part of the group. We can use open source technology to ensure that each of our brands cannot only integrate them onto their platforms, but also contribute to further improving them over time.

Slide 33 provides a great example of how we have leveraged this open source technology approach to help us establish our lead in the market. Our trademark, Same Game Parlay product, has really taken off in the U.S. Our Same Game Parlay bet is one where a customer makes a bet on a selection of correlated events within an individual game. A key element of the success of this product is ensuring that it is fully integrated into the user experience and that it provides a highly intuitive customer journey.

To achieve this, FanDuel is able to leverage the work of the other tech teams in Flutter to offer Same Game Parlay on all major U.S. sports. This helps us to achieve an unrivaled speed to market in turn providing us with a significant competitive advantage in H1. For example, over 50% of our player base placed the Same Game Parlay bet during the NFL season. But longer term, it is not simply about just having any Same Game Parlay products offer customers that will matter. It is about the quality of our product and having the proprietary models that bring a depth to our offering within these sports, the third-party providers struggle to develop. We believe the quality and popularity of our Same Game Parlay product substantially improves our customer retention rates and enables us to generate structurally higher win margins than our competitors.

Slide 34 shows how our pricing and risk management capabilities are giving us an edge in the U.S. Flutter has been investing in proprietary pricing models for approximately 20 years. Our deep heritage and sports betting allows us to offer a greater range of betting options for customers and also ensures a seamless betting experience. We offer twice the number of betting markets on NBA pregame and 7x the number for in-game versus the average of our closest peers. Our greater range of markets also contributes to higher-margin parlay bets forming a greater proportion of our total handle.

We're continuously improving the live betting experience by removing friction from placing in-play bets and increasing the availability of markets for customers, as you can see from the chart on the right. This 91% uptime is significantly ahead of our competitors. However, it also shows that we have further scope to improve our product as we reiterate our models and build our local expertise within the FanDuel team. We also have scale to bring more of our pricing in house. 75% of our handle is currently priced in house, and we expect this to rise to 95% by the end of 2023. As such, while we have a lead today, we are not standing still, and we will continue to invest to ensure we stay ahead of our competitors.

Slide 35 shows why these capabilities are so important. The combination of proprietary pricing models, experienced traders and access to the information provided by the global Flutter betting data pool is providing us with a clear advantage in determining the true probability on the outcome of an event. This pricing accuracy gives us confidence to take bigger bets and customers, but also makes us comfortable offering the lowest prices in the market, as you can see from the chart on the left. Our over-round is lower than our nearest competitors, meaning that our customers get better value at FanDuel. The middle chart shows what happens when we combine this pricing accuracy with our higher mix of parlay bets. Our underlying trading margins, which exclude the impact of customer promotions and acquisition activity through margin has steadily increased over the last 3 years.

This means that despite having very competitive for customers, we generated 140 basis points more in margin over the last 12 months. That equates to 20% more hold for any given level of handle versus the average of our competitors. When you apply this advantage to the expected staking TAM for the U.S. market, that equates to a very, very big number. So that's our product advantage and something we're working hard to sustain.

Now I want to talk about FanDuel's second key competitive advantage, namely the reach and power of our brand. As you can see from the chart on Slide 36, we have leading awareness amongst brands in the sector. Our customers also spend a greater amount of time engaging on our sites and apps in many other sports-led platforms. This high level of engagement is due to the multifaceted content that we provide to customers across sportsbook, gaming, DFS and in-house produced editorial content.

On Slide 37, you can see that the FanDuel DFS database has continued to grow from 7 million registered players in 2017 to 13 million today, benefiting from the halo effect of our sportsbook investment. We have always firmly believed in the high relevance of this player database as sports betters. And to date, it has provided us with over 40% of our sportsbook customers. The benefit of this database is compounded when we look at how customers first come to FanDuel. Based on our customer research, we estimate that 1/3 of our customers placed their first back with us following a friend referral or a viral marketing campaign such as the very popular "Spread the Love" promotions.

While growing our sports betting customer base is critical, we also recognize the immense importance of growing our gaming first base, too. As such, we've been making product improvements that allow us to appeal more broadly to direct casino customers. This includes adding more gaming titles and upgrading our app with the help of our colleagues at our development hub in Cluj. These improvements are starting to pay off, as illustrated by the chart on the right-hand side. We know, however, there is more we can do to grow our gaming share, and we've put in place a clear road map of further improvements over the coming 12 months. We will support those product improvements with greater investments in our direct casino customer acquisition and have already started to do so. Customer acquisition investments in the first half of 2021 was 26% higher than our spend in the entirety of 2020.

We continue to invest thoughtfully in the FanDuel brand. The brand has benefited from over $1.2 billion in investment to date. At a national level, we've recently signed a premium partnership deal with NFL, and we maintained strong partnerships with the other leagues, too. Our national integrations are enabling us to reach a large sports audience, in particular, our successful partnership with NBA on TNT. Locally in sports states, we have a 39% share of voice with the regional sports networks, which are the platforms consumers use to watch their local sports content. We combine this TV presence with preferential access to the #1 sports radio company, Odyssey, and partnerships with at least 1 major league team in each of our sportsbook states.

The significant scale of the FanDuel business with $300 million invested in H1 alone, means we can combine these headline partnerships with a huge range of other traditional and digital marketing assets. Our highly sophisticated data analytics platforms allow us to accurately assess the returns that each asset provides while we maintain the flexibility to double down on successful campaigns.

On Slide 39, you can see what all of this has resulted in. In Q2, we delivered an online sports betting market share of 45%. You can see from the slide that we lead in the 6 largest U.S. states open, which between them account for over 80% of the online sports GGR generated in the U.S. market today. Our pricing and risk management advantages have had a meaningful impact, helping us to generate a gross win margin 300 basis points higher than the rest of the market in the last quarter alone. We believe that approximately half this gap is down to our structural product advantages. The remaining 150 basis points come from favorable sports results, which are compounded by our advantageous product mix.

Slide 40 shows that we had a 51% share in the early states of New Jersey, Pennsylvania and Indiana in Q2, making us bigger than the rest of our competitors combined. It is easy to forget the fact that in each of these states, we weren't actually the first to market. Thankfully, we've been able to build leadership positions over time, quickly in the case of New Jersey and Pennsylvania, though it took us a little longer to get there in Indiana.

Slide 41 shows our progress in more recent large state launches. Our August 2020 launch in Illinois was 3 weeks behind some of our competitors, but we've managed to quickly generate the leading share there. Illinois reverted to remote signup framework in April, but we look forward to the lifting of this restriction next year when we hope to be able to materially grow our player base further. In Michigan, we launched with most of our competitors in January and have invested very heavily in customer acquisition to gain a foothold in that market. This impacted our margin to the point where we generated very little GGR in those first few months. Thankfully, as our margin has started to normalize, you can see how our share has grown. We pursued a similar acquisition strategy in Virginia, where we had the added benefit of being first to market by 2 days.

While the U.S. online market is and will remain very competitive, our U.S. team remains focused and is executing brilliantly in state after state. So how do we top this? First, we must make sure that our team remains hungry, retaining the challenger mindset that has got it to where it is today. Second, we'll continue to keep the flywheel turning by continuing to invest in our customer propositions to extend our existing advantages and to create new ones. Thirdly, we see a particular opportunity to improve our position in gaming and we'll pursue that opportunity relentlessly, leveraging the wider group's platforms and expertise in this area. Fourthly, we will continue to proactively engage with legislators in states yet to regulate sports betting and gaming to encourage them to follow the lead of the states that have already done so.

In the meantime, the pipeline of new state openings has never looked better. We expect 9 new states to launch over the next 18 months, with 8 of these being sportsbook only to begin with. Only one is expected to regulate for both sports and gaming initially. We believe that this positions FanDuel exceptionally well to build on its success to date.

In summary, we are really pleased with the performance of the group in H1. My entire team and I remain completely focused on execution of our strategy and building the embedded value of our businesses around the world. I have never been more excited than I am today about the prospects for the group.

Thank you very much for listening. I look forward to joining you shortly to answer any questions you may have.