Good morning, ladies and gentlemen. Welcome to this FDJ's 2020 annual results presentation. I am today with Pascal Chaffard, Executive Vice President, Finance, Performance, and Strategy. We will present the results together. I hope that you are fine in those times of health crisis. I am happy to be able to talk to you, although, of course, we would have preferred to have a physical event. Since it was not possible, we are very happy to welcome you in this beautiful studio where we broadcast Lotto and Euromillions, our famous draw games. Let me turn now to some general comments about this very special year that we've all been through. Of course, this year comes after one year where we presented our first results after our IPO in February 2020.
I remember then when I described our results one year ago, I said that, of course, 2019 was a historical year for FDJ. 2020 is not easy to qualify. It is a difficult time for everybody. The health crisis is not over. It is not easy to grasp all its consequences on our environment, our economies, and societies. Nevertheless, I think that FDJ in this year, if I had to summarize this year in two words for us, I would definitely say resilience and commitment. Resilience, because the group has demonstrated its responsiveness, in particular by implementing a major savings plan at the beginning of the crisis. Therefore, we were able to protect its profitability and results. Also because we seized the opportunities by having a strong online performance based on the investments that we did over the last years.
Commitment also, because this year we actually put our purpose, our raison d'être in French, in our by-laws in our June General Assembly. We demonstrated definitely evidence of this commitment from the onset of the pandemic with the solidarity actions that we took both to our network, to our point of sales, and also to the community in France. Before commenting further on our results and activity, I want to start with our extra-financial commitment. As I just said, we included in our by-laws our purpose during our General Assembly in last June. It was, of course, an important milestone in this very special year. It was for us, as you know, a natural step that we announced before.
It is completely in line with the history of the company that, as you know, has been created for the benefit of war veterans, which are still our shareholders today. It's also what has made the company unique, I think, for 85 years. It is clearly, for me, at the heart of the sustainability of its business model. We also took, of course, to develop these six commitments to implement this on our products, on our clients, on responsible gaming, and on our retail network, not to quote all of them. We also established a stakeholders' committee to deploy these commitments with high-level and diverse personalities. Thirteen people, experts in their different fields, chaired by Rose-Marie Van Lerberghe, who is a former managing director of AP-HP, a hospital in Paris, and former chairwoman of Korian, a company specialized in senior care.
This committee already met twice by video conference and had vivid and constructive exchanges and suggested ideas on how to progress further for the implementation of those commitments. We also, of course, as every year, and even more in this year, affirmed our CSR commitments and we measured them. We have been rated for several years now by Vigéo. Vigéo, which is an extra-financial rating agency, has been rating us at A1+ , which FDJ at the top 5% of all companies rated by Vigéo. We are also very proud of our diversity credential with a 100% grade on the French Gender Parity Index introduced by Deloitte in 2018. We are, of course, very dedicated to responsible gaming with, for instance, the commitment of dedicating 10% of our TV advertising budget to a responsible gaming campaign. We work further on our environment commitments.
For instance, 100% of our tickets are on paper issued from sustainably managed forests. We also did measure this year, as every year, our contribution to the economy and society by international experts. We found out that we contributed over EUR 5 billion to French GDP, 0.2%, which is roughly the same as last year. In terms of employment, we are responsible for the creation or protection of more than 50,000 jobs in France. Of course, our contribution was particularly significant for some stakeholders, such as the state, the state budget, with more than EUR 3.6 billion that was served to the state budget in 2020. Of course, very important economic contribution to our network of point of sales, with almost EUR 800 million that were paid to our retailers, an equivalent of more than 20,000 jobs that are preserved or safeguarded or created by our activity.
I'm not going to talk about other stakeholders, such as, for instance, the French sports sector, where we maintain our commitments, or the French national heritage that we have continued to support over this year. I want to focus for a few moments on the specific mobilization that we had during this year to react to the health crisis. This has been very important on two fronts. One is, of course, solidarity with all those who needed urgent support, medical personnel or researchers, and vulnerable people because of the crisis. We gave to them EUR 2.7 million over the year into half. We also were very active in terms of solidarity with our network of retailers, specifically tobacconists and news agencies. Most of them were authorized to open. However, they have been under strong pressure during this year. We maintained an active dialogue with them.
We helped them to comply with health measures. We provided them with protective masks when it was difficult to find them at the beginning. We managed their financial relationship with us. We offered them support and coaching all over the year. We are and we will remain at their side to accompany them in the medium term. We have the project to set a specific investment vehicle with other partners where we probably will put a level close to EUR 15 million for them to get some loans in the medium term. Now let me turn to the activity as such in 2020. I want really to stress that we were able to preserve our performance despite a very significant impact of the health crisis on the activity of the company.
We have also seen during this year, and we'll come back to that, a lot of lessons. At the end of the day, our strategic priority is strengthened by the context, particularly, of course, our digital strategy. I think that the company has shown during this year its capacity to rebound when restrictions were lowered, particularly during the summer when the gradual recovery was confirmed throughout the second half of the year, which takes up 3% over this second half despite the demanding 2019 comparison basis. This growth has allowed for a decline, but a decline that has been, at the end of the year, limited to -7% on our stakes after, you remember, a decline of 18% during the first half of the year due to the first lockdown and the halt to international sports competition.
Decline of revenue was roughly at the same level as stakes, - 60%. What was absolutely decisive was the capacity of the company to put in place a cost reduction plan of EUR 80 million right at the beginning of the crisis to preserve its EBITDA and results. Therefore, we were able not only to preserve our EBITDA, but to come at the end of the year with a net income of EUR 214 million, quite comparable, actually, even a little over the adjusted 2019 net income of EUR 202 million that you should compare it with. Pascal will explain that. Therefore, we will be in a position this year to ask our shareholders to approve a dividend of EUR 0.90 per share at the annual general meeting that will happen in June, representing a total dividend of nearly EUR 172 million.
Therefore, of course, delivering the promise that we have made to our shareholders to distribute 80% of our consolidated net income, as we stated during the IPO. I want now to go a little bit in details on the impact of the health crisis of the company. First, on our general activity, you can see that we had very different semesters. It's not a year, it's really two different periods for the company. Each one knew a severe drop in activity because, as you know, during the first lockdown, even though most of point of sales were authorized to remain open, traffic was low, bars were closed, and some of our retailers, in fact, did close completely or partially their point of sales during this period.
We also decided to stop Amigo because we judged that Amigo, as a game that requires you to stay in the point of sale, was not compatible with the context. Of course, most sporting events were, as you know, also canceled or postponed. This, of course, is the main factor for this first half. In consequence, we adapted our commercial and marketing strategy. We postponed most of our game launches or relaunches planned for the second quarter, and we reduced our advertising budget. However, H2 showed the reactivity of our business and the capacity of the company to adapt to new conditions because, as you know, we had a second confinement, although in different conditions, but less drastic. H2 was definitely back to growth in both businesses, mostly in sports with 20% growth, to a lesser extent lottery, and I'm going to come back to that.
In lottery, we had a very different situation between the different categories of games. In draw games, you have the impact of the decline of Amigo for the reasons I've just mentioned, not only in the first half, but also in the second half, where Amigo is open, but with less attraction and less traffic on it. However, our big draw games, Lotto and Euromillions, recorded a +60% increase in stakes. This performance reflects the success of the relaunches of Lotto in November 2019 and of Euromillions in February 2020 with new gaming options and the increase in the Euromillions price cap to a potential level of EUR 250 million. It shows also the resistance of those games because they are really part of our customer habits. Regarding instant games, scratch cards, the story is a little bit more complex. Decrease in stakes was 6% for the whole year.
Of course, first half was very low with minus 11%, and we almost came back to stability in H2. The second half was not able to compensate for the first half despite the launch of new games, particularly in the fourth quarter, including a phygital game. The most interesting feature of this period was definitely the strong increase that we developed on online lottery stakes, which increased by more than 60% over the year and showed definitely the relevance of the group's growth strategy based on investment in digital and development of a more omnichannel gaming experience. We'll come back to that. It's a solid growth, the growth that we had on the online lottery because it's also a growth of new players with 60% growth of new players, 1 million more basically.
I think an interesting feature is the fact that our online players do also play in our point of sale for half of them. Therefore, we think it is really part of the sustainability of this omnichannel business where our players can actually play in both channels. We think it is good for them and good for the company. On sports betting, the story is different. As you see on this graph, more like a V curve. As I have explained, we had a very drastic drop of sports betting in the first half, minus 40% in terms of stakes. The rebound, the very nice rebound that we got in the second half with 20%, did not totally compensate, but it limited the decline to minus 10%. I will not come back to the reason for that.
I think you have followed what happened during the first half with basically no sports events and therefore very limited sports betting. Also, the bars have been specially hit by the period. They've been closed during the first confinement, and they've been, as you know, closed since the second confinement and not been reopened. Bars are, of course, one of our important networks for sports betting. What we saw, however, is a very dynamic market from mid-May, lots of sporting events in the second half, and a very good opportunity to develop the attractiveness of our offer, which we did both in the point of sales with new Lotto Foot, for instance, and increasing the number of possible daily bets for our clients, and also on our online activity with a new website, new streaming features, cash-out feature implemented.
Also in the point of sales, continued dematerialization of our bets with more than 80% of our stakes that are actually prepared on mobile in the point of sale. I want to turn now to our distribution network with, I think, two remarkable features for this year. One is actually the good resistance of our network. I remember at the beginning of the year, we were all very worried about how this network could resist the crisis. At the end of the year, it is still accounting for 90% of FDJ's total stakes, EUR 14.4 billion. After, of course, a difficult first half, it has bounced back in the second half. Therefore, the decline for the whole year was limited to 10% of stakes for our network. Again, if you exclude Amigo, it has been growing positively in the second half by 2%.
We are particularly proud to have protected the size of our network, its capillarity, and to say that actually today we are still, roughly speaking, at 30,000 point of sale all over the territory. We think it's part of the robustness of the company, and this is definitely linked to the support that we have provided to our retailers. Second feature, of course, is the strong momentum on online stakes. All over the company, it's been plus 40%, more than EUR 1.5 billion of stakes, 10% of the stakes of the company. The momentum has been driven, as we saw, by very strong growth in online lottery stakes, over 60%. Online lottery stakes have doubled in two years, reaching more than EUR 1.1 billion.
Roughly speaking, I think the company has gained two years in terms of development of our digital business, and I know that it's very positive news for everybody. Again, we believe that this online business is part of our omnichannel strategy and model. If we look at the digitalized stakes, which is the KPI that we gave us in our strategy plan five years ago, they actually represent over 24% of the stakes of the company. Over the 20% that was set for the end of 2020, and that we have beaten. Last point for the activity during this year, during this year, we were actually able to continue to develop our adjacent activities.
That is to say, the new activities that are today not material in our figures, but that will contribute to the resilience of our business model in the future by providing, over the years, more extra revenues and EBITDA. We were actually able to develop our international B2B business with the gain of several contracts in Europe and our first contract in North America in Ontario for sports betting. In payment and services, we were actually able to deploy the public service bills and tax payment facility that we gained from the French Treasury in 2019. This service was launched in February 2020, and it was actually rolled during the year. We have now more than 9,000 point of sales that do offer this service to pay not only taxes, but also bills in those point of sales.
We have seen EUR 30 million of payments coming through this channel. We believe it is a very good start, although it is evidently small, and it gives us good confidence, as we see, to continue to grow this part of our business. I will now hand over to Pascal. He will cover our 2020 financial results in more detail.
Thank you as well, Stéphane, and good morning, ladies and gentlemen. Before diving into the figures of 2020, it is worth mentioning that the 2019 figures Stéphane has presented and that I will present during all this call have been adjusted for the new regulatory fiscal regime in force as of January 1, 2020, but also for the exceptionally long lottery cycles of 2019, for the acquisition of Sporting Group on a full-year basis, and for one-off costs related to the privatization of FDJ in 2019.
Those restatements already presented during our previous financial communication make 2019 figures comparable, fully comparable to 2020 figures. This particular year that Stéphane has just mentioned resulted in an unusual drop in stakes for FDJ, unusual because, as a reminder, FDJ had only been seeing two annual drops in stakes over the last 25 to 30 years. The decline in stakes led to a similar decline in revenue, 6.3% to EUR 1.9 billion. EBITDA was EUR 427 million, up 1.1% from EUR 422 million, which was the figure of 2019. The cash flow from operating activities increased by 30% to EUR 386 million, thus amounting to 91% of EBITDA. Before going into more details on these financial results, allow me to come back for a few moments to the company's financial model. From the stakes, you know this model, I think, will be very quick.
From the stakes, we deduct player winnings of EUR 10.9 billion in 2020, i.e., 68% of the stakes, to get to our gross gaming revenue, the GGR. In other words, the players' net spending. GGR of EUR 5.1 billion, down 6%. It's then split between the state via public levies for EUR 3.2 billion and FDJ, hence our net gaming revenue, NGR, of EUR 1.9 billion. Group revenue is calculated by adding NGR and revenue from other activities, B2B international payment and services, as mentioned by Stéphane a few minutes ago, of EUR 41 million in 2020. When operating and administrative expenses, excluding depreciation and amortization, are deducted from this revenue, EUR 1.5 billion, we get to an EBITDA of EUR 427 million in 2020. We will come back to this in more detail, but you probably remember that point of sales commission represents more than half of our cost base.
These commissions are a percentage of the stakes recorded in point of sales. Overall, about 60% of our costs, including point of sale commissions, are variable with stakes, and therefore, 40% are fixed. I won't get back over the fluctuation in stakes commented by Stéphane. Our player payout ratio is stable year on year at 68%, but this masks significant differences by quarter. Those quarterly differences have already been commented during our previous financial presentations. Thus, revenue fell in the same proportion as stakes to EUR 1.9 billion. By activity in lottery, the decline in revenue was totally in line with the decline in stakes, 6% each. In sports betting, revenue declined less than bets, 8.5% versus 10% on bets, due to a lower player payout in 2020 than in 2019, - 0.4 percentage points to 75.9%, particularly in H1 and in the last few weeks of the year.
The revenue from adjacent activities remained stable. As indicated earlier, EBITDA increased by 1.1% to EUR 427 million, hence a margin of 22.2%, up 160 basis points compared with 2019. This performance mainly reflects the implementation of the cost reduction plan of more than EUR 80 million versus our initial budget, and/or EUR 40 million versus 2019, more than half of which related to advertising expenses. Let's see now how the main cost items have varied. POS commissions are correlated to POS stakes. The decline of more than 7% in other cost of sales mainly reflects lower trade-related promotional activity, particularly during the lockdowns. Marketing and communication costs relate to, firstly, advertising costs, down 16% over 2020. This decline was particularly strong in the second and third quarter before returning in the fourth quarter to a level similar to 2019 to support a busy marketing schedule.
Secondly, the cost of developing our offering, cost of designing and operating games and services, both in marketing and IT. As we have indicated on several occasions, these costs continued to rise, both in the first half and on an annual basis. The increase of 5% in 2020 illustrates our commitment to strengthen the foundation of our future growth, in particular on digital. The decline in GNA expenses illustrates the positive impact of our cost reduction plan also. In the first half of the year, the decline in EBITDA was greater than the decline in sales, respectively minus 16.4 versus minus 14.7, leading to a slight EBITDA margin erosion of 40 basis points to 20.5%. In the second half of the year, stable revenue and the completion of the cost reduction plan led to a 12% increase in EBITDA, hence a margin improvement of 250 basis points to 23.6%.
This slide bridges 2019 EBITDA to 2020 EBITDA by showing the impact of each major item already commented in the previous slide. The increase on online state had a lucrative impact on the group's gross margin. The low player payout on sports betting in 2020, especially at the end of the year, had also a very positive impact on the EBITDA margin. Finally, the cost reduction plan allowed for EBITDA to be maintained with also a positive effect on the EBITDA margin. It is very important to note that the main expenses items that declined in 2020 should return to a normative level in 2021, for example, sports betting advertising with the Euro Football Championship. Moreover, this event could have an upward impact on the sports betting player payout, which was especially low in 2020. We will detail the various business unit margin on the following slide.
Just regarding our acceleration business unit or adjacent activities, the decline in contribution margin is mainly due to the decline in sporting group business volume due to the pandemic in H1. Holding costs have been reduced by EUR 15 million thanks to the cost reduction plan. If we now look at lottery, lottery revenue was down nearly 6%, and the contribution margin was relatively stable at EUR 502 million, minus 1%. Costs were reduced by 8%. POS commissions are correlated to stakes recorded at POS. Other cost of sales are down by nearly 7%, notably due to reduced promotional activity as mentioned before. The limited minus 1% drop in marketing and communication costs reflects the continued development of our game offering, partially offset by the reduction in advertising cost, plus 9% on game offering development and minus 12% on advertisement cost.
The contribution margin rate improved by 150 basis points to 33.6%, thanks in particular to the favorable mix effect with online representing nearly 9% of total lottery stakes compared with over 5% in 2019. As we already said, the contribution margin on the online lottery is between 1.5 and 2 times higher than offline lottery. Sports betting now, the revenue was down 8.5%, while the contribution margin improved by nearly 3.8% to EUR 103 million. Costs were reduced by 12% with POS commission correlated also to the POS stakes. Other cost of sales fell almost 18% following the cancellation of a large number of competitions. Marketing and communication costs fell by nearly 11% following the very sharp decline in advertising costs in the second quarter, partially offset by an increase in the second half of the year to support a very busy sporting events calendar.
The contribution margin represents 27.7% of the revenue, up by more than 300 basis points compared with 2019. Overall, the sports betting PPO came to 75.9%, as I said before. Its 40 basis point decline from an already low level in 2019 was amplified by unexpected sports results towards the end of the year. Note that the PPO level is not replicable in 2021. On this next slide, as I explained, EBITDA includes BU and LU contribution margin net of holding expenses. Now we will have a look at the bridge from EBITDA to net income. Depreciation and amortization expenses amounted to EUR 102 million, up EUR 6 million compared with 2019, with a full-year amortization of the 25-year exclusive operating rights entrusted by the PACTE Law in France, what we call the equalization payment, compared with only seven months in 2019.
Non-recurring items represented a net expense of EUR 32 million, mainly including asset impairment on H1, nothing new on H2. Net financial income amounted to EUR 5 million compared with nearly EUR 21 million in 2019. This change is explained by financial market trends and also by the increase of the cost of debt. The debt increased as a result of a EUR 380 million syndicated loan to finance equalization payments. After taking into account the next tax charge of EUR 85 million, consolidated net profit amounted to EUR 214 million at the end of December 2020. This net result should not be compared to the 2019 reported net profit of EUR 133 million, but to the EUR 202 million net profit adjusted of this year. It is a 5.9% gross on the net income. In 2020, the change in working capital resulted in an inflow of EUR 39 million.
It has been adjusted for various positive but non-recurring calendar impacts. CapEx totaled EUR 80 million compared with EUR 67 million in 2019. The increase in IT investments more than offsets the crisis-induced decline in investment in POS equipment. The increase in other investment reflects the development of our logistic tools and our commitment for the Paris 2024 Olympic Games. The free cash flow amounted to EUR 386 million, showing an excellent EBITDA to cash conversion of 91%. One of the indicators representing the level of the net cash inherited by the group is the net cash surplus. At the end of December 2020, the net cash surplus amounted to EUR 577 million compared with EUR 298 million at the end of June 2020 and EUR 80 million at the end of December 2019.
The increase is mainly due to the EBITDA contribution and the change in working capital, as explained previously. Throughout 2020, we have, on several occasions, communicated on an available cash position of more than EUR 800 million. It reached more than EUR 1 billion at the end of the year versus EUR 853 million at the end of June 2020. The bridge from cash and cash equivalent of EUR 673 million as per the balance sheet to the available cash position of just over EUR 1 billion, as I just commented, is the following. You add EUR 472 million of term deposit, and you deduct EUR 86 million of Euromillions funds not available. When then the bridge from available cash of EUR 1 billion to the net cash surplus of EUR 574 million is the following.
You add the EUR 246 million of other financial investment, and you deduct the gross financial debt of EUR 571 million. You also deduct the remaining funds to be retroceded to the French state for EUR 156 million. Thank you for your attention, and I will now hand over to Stéphane to conclude this presentation.
Thank you, Pascal. I want to conclude briefly this presentation by saying a few words on our strategy and turning to 2021 perspectives. On our strategy, I think this year allows us to express confidence in the future. The group has proved to be solid. Our business model has shown its resilience, and we are deploying a strategy that enables us to combine medium-term growth with good profitability. This is what we actually, I think, showed in those results.
It gives, again, credits, I think, to this strategy, particularly the strategy that was developed over the last years in terms of digital investments and, of course, in sports betting, but particularly in the lottery, which was the less digitalized of our activity, has proved to be very efficient and very convenient to give us the possibility to accelerate our online development as we did during this year. 2020 is a year that has reinforced our conviction, our strategic convictions for 2020-2025 strategy, built around three business priorities. One is, of course, to continue to develop omnichannelity in lottery, continue to invest both in our point of sales network and on our digital offer as we did and as we continuously do.
Second is, of course, to continue to accelerate our growth in sports betting, both in our point of sales where we have been able to enlarge our offer and to show how this is generating growth, but, of course, also in the online sports betting market, which we expect to be quite, to continue to be quite dynamic this year and the years over. Also, as we stated, by making our business model more diversified in the future and therefore even more resilient with emerging activities such as international B2B and payments and services. To foster these three business objectives, we will need to be definitely even more client-centric and to establish a deeper value-creating customer relationship thanks to progressive customer identification, deeper knowledge of our customers and their ways of playing and interacting with us, and also giving us new tools for responsible gaming excellence.
Now when I turn to 2021, that has, of course, already started, I think it is fair to say that uncertainties persist today and that it is a little bit too early to expect a fully normalized situation, although, of course, we keep hope that it will come during this year. I think we see some positive signs, but we have still to wait a little bit to see those signs materializing. We have started the year, however, on the basis of the lessons we learned last year and on the basis of what we have learned in terms of adapting the activity of the company to new extraordinary conditions. We believe that in 2021, we are in a position to continue to develop our activity, to continue to develop our portfolio of games and services sustained by our innovation policy and marketing policy and advertising.
Lottery should return to a usual calendar of events and activities with more than 10 Lotto and Euromillions special events during the year. Instant games also should return to a more normalized planning of events. We have started in January by launching Objectif Maison, which is a scratch-card three-year annuity game, and it is starting well. We will have a number of new games or relaunch of games planned during this year with, of course, some usual events such as Mission Patrimoine and also a third phygital game, both offline and online. Regarding online games, we are, of course, going to continue to enrich the range of our offer with two or three launches per month, including some exclusive games for online.
Sports betting, of course, will be fueled this year by big events that I think we all hope, particularly the Euro 2021 Football Championship and the Tokyo Olympic and Paralympic Games. In anticipation for that, we are going to fuel our own growth by continuing strengthening the richness and attractiveness of the offer, both in point of sales and online. In our other activities, for instance, payment and services, we will continue to develop. For instance, in payment and services, we expect to continue to roll out the French treasury collection of payment service. We are going to extend this solution for payments, for bill payments from major service providers, energy, rents, etc., with the aim of reaching a significant number of point of sales offering this service by the end of 2021.
I think you will understand that given the current uncertainties and the fast-changing nature of the health situation and the conditions of our business, we will not today restate the precise new guidance. We hope we will be in a position to do it and to communicate on our 2021 outlook as soon as possible in the coming months when I think the situation will become a little more predictable in terms of our context. In the meantime, again, I think we have the confidence that the company is able to manage its activity as we did last year and hopefully so in better and better conditions. Thank you for listening to us, and we are now ready to answer your questions.
Ladies and gentlemen, if you wish to ask a question, please press star one on your telephone keypad.
The first question comes from the line of Jaafar Mestari from Exane BNP Paribas.
Hi, good morning, everyone. I've got three questions, if that's okay. Firstly, on the online mix, which was 10% on average for 2020, during the year, are you able to say what was the highest online mix you achieved at any point in time? Did the online mix come down after the end of the second lockdown, or is the shift now becoming more permanent? Secondly, you're flagging a number of exceptional or unusual mix impacts in H2 and in Q4 that may or may not repeat next year. Last year, you had a number for that. You were saying, "We think there's EUR 12 million of EBITDA from long cycles." Are you able to quantify the euro impact in the same way for 2020?
How much extra EBITDA would you say was from long cycles, from low player payouts, from above-trend digital mix during the lockdowns, for example? Lastly, just on your shareholder base, in May, if I'm correct, retail shareholders who have held for 18 months will get an extra share. Have you been able to gather any intelligence on the makeup of your 18% of retail shareholders? Do you think they're mostly invested since IPO, or has it churned a lot already, for example?
Thank you very much. I will start by answering question one and question three, and we'll let Pascal answer question two. On question one, which is online mix, I think the highest growth of online was definitely during the first confinement. At some point, we had 100% growth of our online lottery digital stakes. After that, I think that does answer your question.
I think it has more or less stabilized at a level of 40%, which gives this 60% over the year of growth of our digital mix. I think what is interesting, again, we are not in a normal period yet, so it's difficult to observe in reality what should happen when we come back to normal, if normal "does exist." What we have seen in the second half is that, yes, a lot of our players have come back to our point of sale networks, but the online activity has kept quite a dynamic trend, higher than the one that we had before. Before, we were growing at a level over, I would say, over 20% per year. Now I think that we can expect to grow at a level, I think, closer between 20-30%, but closer to 30%, sorry, in a normal period.
Again, this is still recent and not in a context which is completely normalized. On our shareholder base, good question. Yes, it's true that next May, the individual shareholders that have bought their shares and kept their shares since the IPO will get one extra share per 10 shares. On the indication that we have, we believe that the majority of individual shareholders that have subscribed during the IPO have stayed with us. We expect around 450. We think we have 450,000 individual shareholders today, and probably that most of them were shareholders at the time of the IPO. We do not have the precise answer yet, but roughly speaking, this should be the type of numbers you might expect in terms of people being able to get this extra share.
After that, as you have understood, we will have our General Assembly in June, and the dividend will be paid on the 23rd of June. I think all this will probably incentivize our shareholders to stay with us during this year. Pascal, on the margin.
Yes, on the margin and the different effects. What we can say firstly on the euro impact, the euro usually is with a higher PPO than other events. We have seen that in 2016. We have seen that also for the World Cup in 2018 because it is more obvious to bet for the players and also because there is a competitive impact of every operator who tries to offer the best odds on this event. It is usually with a higher PPO. That is why we do not think that the PPO will be as low as it was in 2020.
Second thing, Euro 2021 will be an important event to recruit new players who will invest a lot in marketing, in advertisement expenses. It will have a return in investment, but later. It is a little bit it will be negative on our EBITDA margin. To maybe give you some more precise idea on the PPO impact, in December, the global PPO of sports betting lowered by one percentage point during December. This is an impact of a little bit less than EUR 20 million of EBITDA. This is quite a huge impact that you should not take into account as it was in 2020 in your expectation for 2021. Maybe a few words about the long cycles. Yes, in 2019, we had exceptional long cycles with an exceptional amount, and it has been an impact of EUR 12 million on the EBITDA.
In 2020, the impact of long cycles was slightly positive, but not as much as in 2019. That's why we did not restate something for long cycles. On the first three quarters of the year, we were low long cycle than average. In the fourth quarter, it was better than average with this EUR 200 million jackpot on Euromillions. Overall, on the year, it was slightly above average, but not at a very important level. I think I've answered your questions.
Thank you. That's very helpful. Things like the digital mix, which was above trend as well, is that something you'd quantify?
Yes. The digital mix will have two impacts in 2021. First, we will continue to have a growth higher on digital than on point of sales. It is positive for the margin. This is right for lottery.
For the sports betting, you have to remind that the margin that we have on the sports betting online is not higher than the margin on sports betting offline. It's not the same than on the lottery side. On the point, the mix between sports betting and lottery in 2021 will be more favorable to sports betting than lottery. In 2020, sports betting dropped higher than lottery. In 2021, we will have a higher growth on sports betting than in lottery. The mix will be more favorable to sports betting. If you see our figures, the margin that we have in sports betting is important, but less important than on lottery. This mix effect will be negative in 2021.
Thank you. That's very helpful.
The next question comes from the line of Simon Davies from Deutsche Bank.
Good morning. A few questions from me.
Can I start off with cash resources? Can you talk a bit about the M&A pipeline and whether you're seeing much in the way of potential deals? Given the significant amount of net cash on the balance sheet, at what stage might you consider returns of cash to shareholders?
Okay. Yes, we have been able to protect our cash also. That was not obvious at the beginning of the year, but I think we are quite satisfied with that. We have not modified our view and our policy on potential M&A. We have not modified our priorities. Our priorities are still one, to consolidate our sports betting activity if we have opportunities. Those opportunities have not materialized at this point.
Probably we were quite concentrated on the management of the company also, so maybe it was not at the top of our priority specifically at the beginning of this year. Also, contrary to what some people expected, the crisis did not have a negative impact on the value and price of sports betting players, which have been able, as we did also in the second half, to regain part of what they lost at the beginning of the year through very dynamic activity and also who had other activities online where they were more able to continue to grow than sports. It is still at the top of priority, but I would say no materialization at this point.
Other priorities to a lesser extent or smaller size does include everything that could boost our growth in, I would say, adjacent activities that would accelerate our growth in those activities that we expect to be more contributive to growth in EBITDA into the future. At this point, we are not going to modify our expectations. We still think that strategic M&A is one of the key assets for the company to continue to develop further. No project to give it back to shareholders at this point, apart from the dividend that I think is quite the positive news for this year. Thank you.
Great. My second question was on online. Can you say what you think has happened to your market share in online sports betting during 2020?
Do you see scope to materially increase that without broadening out your offering, i.e., moving into areas such as online poker?
For our market share, we think it was roughly stable from what we say. It was quite a fight for us this year, precisely for a reason, I think, related to your second question, which is that contrary to other sports betting online players, we did not have poker or horse racing to, I would say, keep some activity during the first half of the year and cross-sell those players to sports betting. Clearly, there we have, I think, a disadvantage. I think we were quite satisfied to be able to roughly maintain our market share as we expect. Therefore, we are definitely thinking that to have a sustained sports betting activity in the future, we need to diversify it.
That is part of the reasons that we might do some M&A or other ways to diversify this activity.
Great. Lastly, the regulator, the ANJ, recently expressed some concerns about marketing practices in the gambling space in France. What is your response to that? Do you expect any change in marketing strategies in the short to medium term?
You have to have in mind that the new regulator, the ANJ, in fact, did for the first time ever examine the promotional strategies of all the actors in the gaming and gambling market. It was really the first time that there was this examination and therefore communication. For us, there was approval of our strategy. There was no negative decision as such.
I think it is quite normal that the regulators, and specifically the new regulators, with these new responsibilities on advertising and promotion plans, should express some warning about the players. I think that's quite, again, normal and expected. We don't expect at this point any change in regulation, but we expect, I would say, natural dialogue with the regulator on the basis of what they will see on the market. I think this, of course, is not only directed at us, but also at all the other players that I will not mention here because it's not none of my business. Thank you.
Thank you very much.
The next question comes from the line of Sabrina Blanc from Société Générale.
Good morning. Sabrina Blanc speaking. I have a few questions, if I may.
Firstly, can you come back and see what you said about the EBITDA margin in 2020? Because, for example, you said that taking into account the player payout in December had something like 1% impact at EBITDA level. If we exclude this level, can you provide any guidance or your confidence about the margin for 2021? My second question is regarding the ongoing curfew that we have today in France. Have you noticed any differences compared to the previous one due to the fact that we are stuck at home at 6:00 P.M.?
On the margin, Pascal, you want to come back?
Yes. On the margin, the three major impacts we had, positive impacts we had on the margin on 2020 is, one, this PPO level very low that you mentioned, and I explained a little bit before.
We think that we don't know what will be the PPO of 2021, to be clear, but we cannot expect it to be as low as it was in 2020. As I explained with the Euro Championship, with very high competition on the sports betting area, we cannot expect such a low PPO. Second impact, just to come back to that, I said that it was a little bit less than EUR 20 million. It is quite exactly the gap between our guidance and our EBITDA, our final EBITDA, if you see it. Second impact was the mix very favorable to online lottery that we mentioned before. This will still be the case in 2021 because we expect online lottery to continue to grow. You can expect the margin of the lottery to continue to be higher or even to grow if the conditions are there in 2021.
We do not know what will be really 2021. The third thing was our cost reduction plan of, I remember, EUR 80 million compared to our budget and EUR 40 million compared to 2019. We will reinvest a major part of that because half of this cost reduction plan was marketing expenses. We will be higher in 2021 on marketing expenses than 2020, obviously, but higher also than 2019 and higher than was the budget of 2020 initial. We will not have this positive impact. If we come back to 2021, we will not provide any guidance for 2021, but I can give you some more color about the year, how can it be this year.
If we don't have a new lockdown, the first half of the year should be quite similar to the H2 of 2020, with some maybe slightly lower because we will have a longer period of closure of bars. You know that the closure of bars has an impact, has an impact especially on Amigo. We will not have in H1 the particularly high activity that we had at the end of the year. This is not, we will not have also the catch-up that we had on H2 2020 on sports betting because we have more events on sports betting on H2 2020 to catch up the events canceled on H1. We can expect first half of the year slightly lower, but comparable to H2 2020.
If the conditions and the health crisis is coming to an end by the middle of the year, we could expect a second part of the year with a growth back to our normal growth. Again, it's just to give some color, it's clearly not a guidance. We don't know what will be 2021. On the margin on 2021, I think you should not take the margin of 2020 as a good benchmark for 2021. It could be lower for the reasons I've already explained, but it should be higher than the margin of 2019. We are in the process of improving our margin on the medium term, and we will try to improve it comparing to 2019. The 2020 figures are not a good comparison, if I may say that.
On your question on the impact of the curfew, not just that I will, it's, we have not measured it in a long period. I think it's still to be assessed over the next month if it was to continue. We don't see major impact of the curfew. Probably the most significant impact could be on sports betting in our point of sales because it is definitely an activity that people tend to practice more at the end of the day than at the beginning of the day. On lottery, it doesn't seem to be significant. That's what we know at this point. Thank you.
A reminder, if you would like to ask a question or make a contribution, please press star one. The next question comes from the line of Philippe Sarreau from Pictet.
Hello? Hello? Hello. I have two questions.
The first one is related to the sales commission. I was just making a simple calculation on sales commission for 2019 and 2020 coming from your press release. In 2019, you paid EUR 859 million of commission, which is about 41.9%, which dropped to EUR 772 million in 2020, which is about 40.2%. You explained well that this level of commission was coming from the development of online sales versus the POS. That part of that should be sustained in 2021, which there's an echo. How much of this 170 basis points of margin could you keep in a normalized environment? You will have a kind of restart of your normalized POS. The second one, if I understand well, the previous question you mentioned about H1 and H2. If I take H1 EBITDA, H2 EBITDA you mentioned, the EUR 253 million.
You said H1 2021 should be comparable to this minus a certain number. Let's take 20-30 million out. The second half should grow normally over this level. We get to an EBITDA for 2021 of EUR 48,500 million, if I'm not mistaken, in that respect. Could you maybe enter those two elements?
Maybe I will start with some general comments and let Pascal answer. I'm not sure we've got because the line was not that good. I'm not sure we had all your figures well. On the level of commissions that we pay to our retailers, I think you have, in terms of factor, you have, of course, the global level of activity. Maybe we'll have to give you, but I don't have the figures in front of me. You also have an important feature, which is the mix of activity.
Because as you might remember, or as I will remember you, we have, in fact, modified the structure of the commissions paid to the retailers by increasing the commission paid on some lottery games, for instance, particularly the big draw games. We have actually lowered the commission on sports betting. We have this scale of commission. You have to combine the general activity with, in fact, the mix of our activity. Therefore, it's not easy to make the comparison. That was just general comments to your point before saying what could happen in 2021. In 2021, you have clearly, if the year was as we expect, you will have, and particularly in the first half with the Football Euro, you're going to have an important activity in sports betting in the point of sales if they are open to get it.
This is just to nourish your calculation. Maybe I will let Pascal comment on your expectation for EBITDA for first half. I'm not sure I got completely your numbers, but Pascal, you might.
Like you, I did not completely get the numbers, but I will try to comment it globally. What I said when I said that H1 2021 could be slightly lower but comparable to H2 2020, I was referring to activity. You have to remember that part of our cost saving plan has been taken into account on H2. Clearly, you should not take the EBITDA of the second half and try to make it a benchmark for the first half of 2021. To be clear, on activity, H1 2021 could be slightly lower than H2 2020. I comment the EBITDA on the global year of 2021 and on 2021 global year.
I don't comment on H1 particularly. You should not take the 22.2% as a good benchmark, but deduct from this the effect of PPO, the effect of the cost saving plan, and the effects of the mix of product that we mentioned before.
Okay. Could you hear me?
Yes.
Yes. Just a quick calculation. If you will keep the level of commission of 2020 into a normalized year of about, let's say, EUR 2.1 billion, it will be EUR 100 million benefit to your margin.
Yeah. To be clear, on the commission of the point of sales, it has dropped from 10% in 2020 because the point of sales activity has dropped from 10%. We hope it's not a guidance. We hope that in 2021, we will come back with a growth on the point of sales activity.
We will definitely have higher figures on the POS commission, and we will be very happy with that because, as Stéphane said, we like the digital activity, but we also like our point of sale activity.
If your online sales growth is the same higher than the point of sale, the physical point of sale, you will keep the same margin commission because it will grow at the same pace. The structure will not be different.
Yeah. Just maybe to clarify a little bit, last year, our online sales doubled during the first lockdown, but they were plus 40% on Q3 and globally on the full year, plus 60%. You cannot expect our digital activity to have the same figures in 2021 because the situation of the point of sale will not be the same.
With a situation of point of sale coming back to growth, we won't have this kind of growth on digital. It's impossible. We have climbed a step in 2020. We will still have important growth in the next years, but not at the level that we have in 2020.
Yes, but the point is, if they will grow faster than the restart of the physical point of sale, your rate of commission will continue to fall. Just to stabilize, just need to grow, the online sales just need to grow at the same pace as your physical network, and you will keep the same margin percentage mathematically. If online sales just grow by the same level of your physical, the 40.2% commission will stay at 40.2% because it's the kind of the breakdown of your sales which basically will not change versus the whole year.
What I'm trying to get is how much of this benefit that you got into 2020 is going to become actually structural because you increase the share of your online sales in your overall turnover. As long as it does not go back down, the split and the share of your digital commission will stay the same. The commission rate will not go back up.
Okay. Again, I do not think we are going to get to the older calculation right now. We will come back to you later on. Again, I think we are ourselves trying to grasp a very unstable situation, I think, on those issues. I think what Pascal also has pointed out is that in the online stakes, you have lottery, which is relative in terms of EBITDA, but you have also sports betting online, which does not have the same characteristic.
That also, I think, goes into your calculation to nuance your point about the extra EBITDA fueled by the stabilization of this percentage of digital stakes. We will come back to that later on. I understand it's an interesting point. Sure.
The next question comes from the line of Alex Girard.
Yes. Good morning and congratulations for that performance. I have two questions, if I may. The first one is related to the KPIs that you mentioned on page 36 of your presentation, and particularly on the number of players in lottery. You mentioned 23 million players, down from 24 million players last year. If I remember well, seven years ago, you had 27 million players. Curiously, sports betting also the same trend, I mean, downward trend from 3.1 million players last year to 2.5 million players.
Can you comment a bit on these downward trends in terms of the number of players? That is my first question. What is behind that? Second question on sports betting, it is a key priority for you. Can you qualify more precisely how you intend to increase your market share? What is the level of your market share within the online segment? You also mentioned 80% digitalized stakes within the sports betting BU, but what part of that 80% is truly online? Last part of that question, what should be the normative PPO within online sports betting given that mix in terms of clients online and offline and given their respective PPOs, 75% and 85%, if I recall well? Thank you.
Okay. Lots of questions. We will try to maybe answer each of them briefly because it might take a little time.
On the question of number of players, first, you have to have in mind that this measure of number of players under the definition which is as played at least once in a year at a game at Française des Jeux at FDJ is, of course, a measure which has, even though we have perfected it, a significant margin of error because it's only a poll that, in fact, we do. It's a study that we do once in a year. You have some margin of error, but it does not modify what you say in terms of trends. Frankly speaking, I'm not going to comment the trends over the seven last years because I think our business has changed a lot, and I think it's more interesting to try to explain where we are today and particularly what happened this year.
Clearly, this year was not an easy year for recruitment. When we look more specifically into this base, we see that we have some regular players that, in fact, as you might expect, do account for the majority of the stakes. This basis of regular players is actually quite stable. That is good news. It is not the 23 million, but 23 million. If you play once, you remain because you want to get EUR 200 million jackpot, you are counted among the 23rd. We really do look a lot at the regular players, and the regular players have been quite stable. What has been difficult in 2020, for obvious reasons, is recruitment, and particularly since, of course, our business is mostly offline still, 90% of the stakes.
When people are not allowed or, I would say, are strongly incentivized to stay at home and do not go in the point of sales, the level of recruitment of new players, which might be occasional or which might become regular players, is lower. We have measured this impact in 2020. Yes, we have compensated a little bit online, but actually, as we said, half of our online players do also play offline, so they are not all new customers. It is a very key indicator. The key indicator for us is not only the overall base, which is, again, a measure of clients which does not represent the level of activity, as we say, but more how we do recruit because we have, of course, some churn as any business.
If we want to keep a nice and large player base, it's in terms of recruitment and retention, of course, after recruitment that we have to work. This year was not, obviously, very easy in that regard. For sports betting, as we said, we have continuously invested for organic growth to boost our market share. This has worked quite well during the last two years. It's been more difficult this year for the reason I just, I think, explained before because the nature of our sports betting online activity being only sports and not having any other activity to cross-sell made it more difficult during this year. Again, we have been working on an organic base. The question for us is how we might be able to grasp some opportunities to grow our sports betting online business through M&A. It has not come this year.
It is still at the top of our priorities. We don't disclose our online market share. I think that's a usual answer to a usual question. We are still around what we were last year in the second tier in this market. We're not amongst the three leaders. We would love to be, but we're not there. It has not changed drastically. In the 80% that you refer to that I mentioned in my presentation are actually offline sales because it's really sales on which people prepare their bets on their mobile in the point of sale but are paid to the retailer. All those stakes are accounted in our point of sale sports betting business. Pascal, if you want to have if you have an idea of what is the normative PPO?
I think there is no normative PPO.
Maybe just to summarize it, PPO is a matter of competition and odds. It's a matter of mix of type of bets. It's a matter of expertise of players and of sports event results. What we can say is that on the online market, the competition is high, and the PPO will probably be close to 85% if we can say normative level. On the point of sales, as we have said during the IPO, our thinking is that the trend will be to see a slightly higher PPO year on year because the expertise of players is becoming more important, and because also the mix of type of bets is not very favorable for us. You can expect a growing PPO year on year on the point of sales.
Okay. Thank you very much.
Thank you.
The next question comes from the line of Joanna Jordan from Odoo.
Yes. Good morning. Two questions for me. The first one is regarding holding costs. I think you mentioned EUR 15 million of savings in 2020 for the holding costs. What is the path of those cost savings that will remain permanent in 2021? My second question is regarding the payments and services activity. You mentioned a collection of EUR 36 million of tax payments from the French Treasury in 2020. What is the revenue coming from that, and what should we expect in terms of additional volume coming from the extension to other service providers? I think you mentioned energy and rent. Thank you.
On the second one, on payments and services, I think it's not material today to talk about revenues. That's why we don't mention those figures.
It is even more too early to talk about the prospect that we have in energy and other features because we have not signed yet. I think it is a little bit too early to talk about that. I think we will come back to that probably, I would say, second half of the year when we will have effectively contracted with some of them, and then we will be in a better position to answer to your question. I think today is really the beginning of this activity. That is why it is not a business unit as such. It is a big investment activity, which is, again, taking support in the investment we have already in the point of sales.
It is more a way to develop future extra revenues and EBITDA in the medium term over an investment that is already there than something that is material to our revenue and EBITDA.
To answer to your question on the holding costs, maybe to answer more broadly, we talk about a cost-saving plan of EUR 80 million, referring to our initial budget 2020. It is EUR 40 million referring to 2019. What you could expect is globally to have EUR 15 million left in savings in 2021, half of them in G&A. You will see it in the holding costs, and half of them on cost of sales because 2021 will be also hard to have a normal cost of sale activity, a normal commercial activity on the point of sales.
Thank you. You said EUR 15 million left in savings, right? Yes. Okay. Thank you.
As a reminder, if you would like to ask a question, please press star one. We have no more questions coming from the line.
Okay. Thank you very much for listening to us. Thank you for your questions. We will certainly be in contact with you in the next weeks. Of course, we will have our Q1 results mid-April and General Assembly, as I said, 16th June. We will be in contact and talk to you soon. Thank you very much. Bye-bye.
Thank you. Goodbye.