Accor SA (AC) Earnings Call Transcript & Summary

Hello, and welcome to the Accor 2021 Half Year Results Call. My name is Jazz, and I'll be your coordinator for today's event. [Operator Instructions] I will now hand you over to your host, Sebastien Bazin, Chairman and CEO of Accor, to begin today's call. Thank you.

Thank you very much. Good morning, everyone. Here we are for the H1 semiannual reporting numbers for Accor. Very happy to be with you. Jean-Jacques is next to me, here on my right. So let's try to have a good, positive conversation, and of course, be super factual. What you have on Page 4, which says recovery context, there's 3 things on this page. The first, of course, on the left, it's certainly better than where we stood last year on June '20 where 87% of all our countries were actually closed or partially closed. That number is today 63%. I would have wished that number to be way below 50%. This is not the case as of yet. So we're still in a bit of a muddy water in terms of navigating us through different countries and different rules. However, what you have in the middle of that page is very, very interesting and certainly positive. We haven't shown you here, and we should have, the months of January, February, April -- you do see April here at minus 64%. If you really were looking for the 2 months prior to this, that was 63% to 64% the exact number. So it is noticeable for sure that for the last 3 months in a row, we've been improving RevPAR by exactly 500 basis points and probably likely to be the case and maybe even better for the month of August. So no question, there's a positive momentum in terms of rebound, in terms of recovery for the last 4 months and undoubtedly due to the vaccine rollout and the progressive opening of many borders in many countries we operate in. So there is certainly a sign of light and very positive light in the next few weeks and the next few months ahead. Again, I think 90% correlated to the efficiencies of vaccines being rolled out. However, what you have on the right side of is we do have still numbers of variants and it is certainly nonhomogeneous depending on the countries you're addressing, whether you call it gamma in South America, whether you call it delta in Northern Europe, it is certainly still messy. But clearly, vaccination helps then there's few people going through a hospital so that shouldn't be impairing as it did last year on conduct of activities. What you see is, unfortunately, for us still and given the announcement this morning from England on obviously eliminating quarantine for a lot of European Union members coming to the U.K. with the notable exception of France, which I don't know why that is. But again, every day, we have faced with different rules that we have to adapt and we have to explain to our customers, which is why we're really still depending quite a bit on the domestic markets. The third point here on that page is, no question, the size of market matter quite a bit in terms of performances and the 2 largest markets, U.S. and China, still have quasi borders being closed, but rebounding both of them in different matters, which you see on Page 5. Page 5, we just gave you 2 axis. The vertical one is the pace of recovery. The horizontal one is the weight of domestic market. No surprise, and you see it on the upper right, U.S.A clearly is benefiting from the stronger rebound because 95% of the travel and hospitality sector, as you know, in U.S. is depending on domestic clientele. China is also recovering quite fast as you've seen for the last now 3 months. Australia was and will come back. Clearly, we've been impacted for the last 3 weeks now with closing down the Sydney province, which has not been too helpful. But they had a huge rebound, the monsters. Before, that didn't happen. And it's due to be reopening, I hope, by mid-August, but still a pretty resilient market and certainly being helpful to us on Mantra and other activities. You see in the middle of the page ready for rebound and we've seen it through the summer season, whether it is in the U.K., in France, certainly on the Med shore. Germany is certainly better than we expected a few weeks ago. And Brazil is a big surprise. Vaccination rollout in Brazil for the last 7 weeks only went from 17% to 45% first dose. Very, very fast in terms of actually rolling out vaccination and certainly helpful for the domestic hospitality market. Bad news, no surprise, and we have to learn and we've learned to be patient. We have a delayed rebound for Southeast Asia, certainly Thailand, Vietnam, Cambodia. All those countries depending 90% plus of international travelers from Korea, from China, from Australia, from Europe for that matter and still some very, very difficult markets in South America like Argentina. That's where we are. It is clearly very eclectic. But let's fight for places in which we see the rebound. And certainly, we're benefiting from every rebound in any market. So now Jean-Jacques, to you.

Okay. So thank You, Sebastien. Good morning, good afternoon to everybody. I am very happy to be with you today. And just as a word of introduction, we will do as we did in Q1, i.e., for the sake of clarity, we've continued to provide to you RevPAR variations versus 2019. That makes the comparison much easier and it reflects base effects. As for revenue and EBITDA figures, we will provide in the presentation the variation versus 2020 and 2019 when it is relevant, so when the ratio makes sense. So without further delay, let's move to Page 7, where you've got the highlights of the first 6 months. On the left part, so you see here the challenging macro environment that Sebastien has introduced. RevPAR was down 60% over the semester versus 2019. The last 12 months of net organic system growth was at 1.9%. And all of that did translate into a group revenue to the tune of EUR 824 million, i.e., a decrease of 53% versus H1 '19 or 6% versus H1 '20 on a like-for-like basis. The underlying figure reflects more heterogenous performances across geographies, but the detail later on. On the right part, you see that in that context we did keep a very strong hand on operational control of the business. RESET cost of the plan on target, and we will achieve the EUR 70 million-plus EBITDA that we had guided you towards. And both on EBITDA sensitivity to RevPAR and monthly cash burn guidance that we had given to you back at the early results in February, we are well in line, well comforted by the actual. And we do confirm the numbers for the full year i.e., an EBITDA sensitivity below EUR 18 million and an average monthly cash burn below EUR 40 million. If you move to the next page, we put that slide back to refresh everybody on where we are on RESET and what is RESET. So RESET, you see on the left part the EBITDA effect of it. No change here, and we will be done and we will exit by 2022 with a EUR 200 million effect to the bottom line at the EBITDA level. The nature of the saving is 50% staff, 50% nonstaff. Just to give you a little bit more granularity. One deals with staff, what is the first portion of it. All the non-European departure plans were completed by Q1. The latest and the largest one is France, which will start in July 2021. And the reason for the differences in time is fundamentally the legal process that you've got to go through depending on which trajection you are. So it's easier to do things in Asia Pacific, a little bit more formal when you are in Europe. The plan reflects the reorganization with the management structure, which has now less layers. But it also reflects a significant amount of tasks that we diminished and reviewed downwards. You see here on the table, 1,000-plus tasks have been eliminated out of 7,000. That's all the work that we've done over 6 months to analyze who does what end to end and make sure that we've got as much as possible a lean process. You see other elements here, just to give you color: the decommissioning of 20% of IT application; a migration plan towards the cloud; the reduction of contractors, which are, as you know, more expensive than internal employees; and create over time dependency in terms of competencies. We also have optimized marketing. And there is a significant effort to get one unified management contract databases that simplifies the life of everybody. That's why we all notated all the HMA. So that's some RESET update. Now if we move to the more classical review of the RevPAR evolution, and I am on Page 9. Overall, 60% down, as explained by Sebastien, an improvement which is very marked since April and that the July numbers, as we see it, is confirming. When we start with Europe, which is about 50% of what we do, we clearly see an acceleration over the second part of the quarter, which is just the cancellation or the restriction relaxation and also, very importantly, the higher rate of vaccination. We are now on double vaccination above 50% in a very large part of Europe. So starting with South Europe, minus 63%. The lockdown in April obviously didn't help. But RevPAR in France improved 22 points, so 22%, driven by Provence between April and June. So very significant increase here. July to date, by the way, the effect of the fourth wave is very little. So we don't see any cancellation of any significance, which is a question you can ask yourself. If you move to North Europe, minus 74%. Two different pictures here. The U.K. is very similar to France and you've got 26% improvement driven by the Provence between April and June. And good news here is that, as you know, since mid-July, there is quarantine-free travel for all fully vaccinated British residents, so that globally helps the business. In Germany, the situation is different. They got delayed relaxation of restriction, and very importantly, there is no business event. And a large part of the business in Germany itself [ so-called method ] in Germany. If you move to Asia Pacific, the RevPAR shows marked improvement. We are down to minus 38%. But here, again, a very heterogeneous picture. You've got the good players, Pacific and Greater China. And they are at RevPAR, which are, respectively, minus 19% and minus 18% so significant improvement. Both regions have got about the same pattern, i.e., they are large domestic markets and very well-controlled sanitary situations. To be quoted is the fact that in Pacific, there were cases, I would say, few cases compared to what we see in Europe, we talk about 100 cases. When in France, we talk about 10,000 cases per day. But this has been enough for Australia to decide to basically shut down Sydney and Melbourne, and that will affect the performance for Q3. Moving to Southeast Asia. RevPAR is at minus 70%. And here, vaccination is clearly lagging and 80% of the business in Southeast Asia is coming from international travel. So you've got a direct effect to the business numbers here. When you move to Middle East, Africa and India, RevPAR as far is down 44%. Dubai has been doing good just like it has been doing good for the last quarter with open borders and restaurants. And the Expo 2020 should be a further catalyst in Q4. If you move to Saudi, there, on the other hand, the pilgrimage permits were stopped in June and so that didn't help the plan. What has been communicated is that they will restart the pilgrimage permits, which is what you need in order to be able to go to the holy cities and which is where our large part of the business is stemming. And so they will restart clearly. So again, that's a positive. If you move to America, again, here, a very different picture. You've got the U.S., which, as you all know, has been steadily improving month-after-month And on the other hand, Canada, very -- has been plagued, in fact, by very stringent restriction over H1. The good news again here is that mid-August, the borders between Canada and the U.S. should reopen. And we do a lot of our business, notably in the Rockies, from Americans spending time on -- Canadian, sorry, resorts. Now so much for the RevPAR. Moving to the net system growth. You see here the overall picture, a net system growth at 1.9% over the last 12 months. We opened over Q1 -- over H1, sorry, 15,000 rooms, which is better than last year, 30%, but which is less than H1 '19 by about 20%. So we've recovered on the opening, but we are not where we used to be. Again, there is no real surprise here. The pace of gross opening is somewhat subdued as owners are cautiously monitoring the activity rebound, and this is extremely true in places like Southeast Asia. Overall, we saw the openings being postponed to the tune of 30% over H1, which is a bit more than what we see historically, less again than H1 '20, but more than what we have seen historically and the translation of the overall pandemic situation. Churn remained under control. This will drive to probably be in the low range of the 3% to 4% net system growth guidance that we have given to you back in February. Just a couple more words on development. I mean on Huazhu, no change. They opened [ 5,000 ] rooms in H1 '21, very much consistent with last year's performance. Again, translation of China been doing good. And as far as conversion, we are at above 40% of the openings, which again is very consistent with historical result. There is no spike per se of conversions. We are in line with what we had foreseen before in history as performance. And moving to the next page, which is now dealing with revenue. You've got the revenue by segment, and I am on Page 11. You see here Accor revenue at EUR 824 million. And the variance between the like-for-like and the reported is Mövenpick and currency effects, but nothing here of significance. If you move to HotelServices, the revenue is down 60%, which, again, is fully in line with the 60% RevPAR drop that we had mentioned. If you go down a little bit more, you've got the M&F, which is a sharper drop at 67%, I'll detail that later on, but again here no surprise. And as far as Service to Owner is concerned, it is only down by 56%. And the lesser revenue decline reflects stronger activity in the U.S. and in Asia Pacific compared to the group average. Again, the U.S. and the places where we've got notably reimbursed cost in Asia Pacific are the places which are doing good like Australia. If you move to the last segment, which is Hotel Assets & Other. This segment is, in fact, doing better than HotelServices, minus 38% on a like-for-like basis, and this is thanks to Australia. So what we saw in Q1 did confirm itself over H1, i.e., leisure demand in holiday destination just like the Gold and the Sunshine Coast. So essentially, Australia was a closed country and the people within that territory have been taking vacation in a -- much more than, in fact, we could have anticipated. And as far as new businesses, which is part of Hotel Assets & Other, as you know, the report, as you would expect, better trend in digital services than in travel-related services. Travel-related services are in line with what we see in Australia business. The digital are much better. If you drill down now on the M&F revenue like we typically would do, you see on M&F decreased by 67%, RevPAR by 60%. It's the euro distortion. It's the effect of the incentive in the management contract. What is interesting is when you look at it not versus H1 '19 but versus H1 '20, you see an increase. You see a 19% increase. And that translates the fact that incentive are triggered again -- are triggered back. So from a situation where they dropped significantly in H1 '20, they are coming back up in H1 '21 and notably in Asia Pacific, in Australia and notably in Dubai. And so that's why, in fact, you've got that increase on a like-for-like basis of M&F between H1 '21 and H1 '20. So a very good sign that when things comes back, obviously incentive flow back. Now if you look not at revenue, but if you go into the EBITDA, the EBITDA reflects the operational upgrade that I've been talking about before. I mean you see that for RevPAR, which in all about the same number, minus 60%, the EBITDA improved by about EUR 100 million, so EUR 107 million to be precise. And so that resulted in a good EBITDA sensitivity, which is exactly below EUR 16 million per point of RevPAR. So below the guidance that we have provided to you comforts us on the fact that the guidance is the right one. Now how do we explain that? There is clearly an effect, which is Australia. We had not anticipated that the Australian people would behave like they have been behaving, and so that has been a good surprise and that shows up in the EBITDA sensitivity. The other thing is the incentive. Again, here, there are some incentives and those incentives are notably good in Middle East and Africa and notably good in Australia. And so that's again a positive. And there was in fact, no debt, bad debt, accrued this semester versus last year. So that's what, if you add up, improved the EBITDA in absolute value. That's what, when you add up, improved the sensitivity to RevPAR when you add up. And I'll keep the last one, but the most important one for my last point, which is that's where you find RESET. All the resetting that we've been talking, that's where you find it. So the improvement in EBITDA is RESET. It's all the good things that we've learned to do on the variable cost, the improvements in variable cost, which are nonpermanent but are, nevertheless, an adaptation of our cost base to the revenue base. And on top of that, there were other few things I said in Australia incentives at the call. Moving to HotelServices, I'll talk about the M&F next slide. On Service to Owners, the EBITDA remains negative here as we are not fully flexing. Just like it has been the case for the last quarter, the cost base to the revenue base. If you move to Hotel Assets, I've mentioned it, it's coming from Australia, but also better performance of new business to a lesser extent. I'll detail that in an upcoming slide. I move now to Page 14, there is here the detail of the M&F EBITDA. I won't repeat the reason for why the M&F EBITDA is better. They are the same reason that the one that I just went through. I'd like to highlight also one thing, which is we did get some one-off governmental subsidies over H1. There was a law in Germany by which fixed costs would be reimbursed as a onetime payment. And so despite fundamentally furloughed scheme disappearing or being, in fact, at least reduced, like in Australia, for example, there is no more furlough scheme and you may recall we've significantly benefited from that because Australia business good. And so the furlough is going down. On the other hand, we did have a one-off -- a nice one-off in Germany. So that's one thing I would like to highlight because, again, these one-offs do not necessarily happen again in H2. That's why they are one-offs. If you move to Hotel Asset. On this one, I'll go relatively quick. I mean Hotel Assets & Other is 2/3 Australia. Then it is 20% Europe, the 20% Europe is the new businesses. And then there is a little bit, which is in America and that's the variable leases that we've got in Brazil. So when you conflate that into profit, the EBITDA for Australia was very good for all the reasons I said and the people really willing to get back to life. And so you've got the largest part of the improvement of EBITDA coming from Australia and Mantra. The new businesses were better than last year. And then Brazil basically is protected because these are variable leases, so there is no variance on this. So that's the transition that you can have on Hotel Assets' profitability. If you to Page 16, now we move to the lines below EBITDA to net profit. You can see that we generated a positive net profit of EUR 67 million. Very large variation here. The share of net losses of associates and joint venture, which is why you see joint venture is now fundamentally AccorInvest or share -- or remaining share in AccorInvest of 30%. Again, we went through that, Europe was not good. AccorInvest is Europe, so AccorInvest performance was impacted by the overall performance of Europe. Slightly better than last year, but nevertheless negative. And then last year was worse, it's worth mentioning, because it included sbe. Since that time, you know that we've been working out a deal, creating that asset-light platform on Lifestyle. And the asset-light business of sbe is now part of the platform, 100% owned since the end of 2020. So essentially, that line has been disappearing and you may recall that in those days sbe was an asset JV element and it was part of asset road map, sorry, than to cleanse that situation. So that we did, and you see the effect of it in the numbers. Nonrecurring items mainly relate to Huazhu. So Huazhu is a great story. We went through that, you saw the kind of return that we got, 5x on the first round offer, 8x on the second round offer. So what you've got here is you've got the effect of the additional sale of the stake of 1.5% in February that generated EUR 240 million of cash. And following that sale, we have adjusted the accounting to report it not as a joint venture, but as a financial investment. Financial investments are recorded at fair value. So when we did a recording at fair value, we captured, in fact, the upside on the sale of the February stake to the full amount, sorry, of the remaining stake of 3.3%. So that's the accounting role. So net-net, the capital gain recognition is to the tune of more than EUR 600 million for the 3.3% residual stake recognition. If you move to the cash, you see here that significant, again, improvement on cash burn. Still negative, but half what it was in H1 last year. A couple of highlights here. On the recurring investment because of the situation that you know of and this difficult context, we did put a very strict controlling over H1. And so the figure is at EUR 38 million on recurring investment. It will go up in H2, that number is not what you will see in H2. In H2, you will have a significantly higher number because notably we're going to open some hotels and there will be some key money -- cash outlay. I'll give you one. Just the Fairmont in Los Angeles is a double-digit figure and the first number is not the one. So that's just to tell you how things can move and be chunky from one period to the other one. Nevertheless, I do confirm to you the guidance that we have provided before. We said EUR 150 million to EUR 200 million. I think we will be in the low end of the EUR 150 million bracket. The second element, which is worth mentioning here is working capital. It's very negative last year. You may recall because of the thrust of the COVID crisis, we had a significantly negative working capital effect. And I would say that it is back a normalized level. It's a good level. It has been well managed. We do collect the fees. And it's negative, but if you recall what the business is, every first half of the year, the number is negative. And by the way, it's much more negative than the EUR 40 million that you've got here. So all in all, again here, a very good working capital management. Last but not least, it's paramount that I want to highlight it, the cost of debt reflects just the credit downgrade that we had back 1 year ago from the rating agencies when we moved from BBB- to BB+. And so that's the step-up that occurred at that point in time. So that's on the cash burn. If you move to the last slide that I will cover, which is the status of liquidity. Obviously, the balance sheet position is very important throughout a crisis, just like the one that we are living with COVID. And our liquidity position over H1 is at EUR 3.4 billion. And we provided to you a bridge between December and June. Most of the items, which are here you know of, the 1.9 -- 1.5, sorry, percent stake disposal, you know. The AccorInvest capital increase and -- in that amount of EUR 241 million. There is also the EUR 35 million for the SPAC and the Mama Shelter minority buyout, which is part of the any small or Lifestyle platform creation you know of. The debt repayment did occur in February. Two things that I'd like to complement the comments with. The cash burn that's what we just went through, so the EUR 40 million [ times 6 ]. And the commercial paper, we were able to issue EUR 100 million more commercial paper in H1. I think it translates the credibility of the name of Accor on the market because besides being noninvestment grade, we are still able to have EUR 400 million of commercial paper outstanding on the program of EUR 500 million. And it translates, in fact, the appetite [ for the accounting ]. In the other, you've got a few things like the hybrid coupon and that's where you also have the restructuring cash cost. So with all that detail, we -- I would like to leave the floor back to Sebastien, which will give you some closing remarks.

Thank you so much, Jean-Jacques. So we're back on Page 20 here. Two or 3 things here. On the left part, just reaffirming something it takes almost 25 years ago, which is all the different measures, different actions being undertaken by Accor and its teams all over the planet on trying to be generous, trying to be protective of everybody who is vulnerable. You know we've been acting quite a bit for helping not only to a large diversity, but to make sure inclusion exists, to make sure we make things easier for handicapped people. We've been honoring United Nations on gender-based equality. We've been partnering with Expedia with also the UN on going forward with a sustainable development program, which we've been doing with 2021 for the last 15 years. You also know and we have collectively to be very proud of it of the ALL Heartist initiative, the fund that we put together 14 months ago. The numbers is huge and probably would have worked with the number being below. We put aside EUR 70 million to help people who did not get access to hospital or didn't get access to food or in disarray because of the furlough, hotels being closed. As of last night, 88,000 people of Accor have been asking for help in which we've been granting roughly $350 per individual. That is $27 million that's been distributed over the last 14 years, and it is not over. We will continue to be there for those in distress, who have been helping Accor for the last 50 years to build what has been built. We also have been displaying the ALLSAFE protocol in 96% of the hotels. Extremely important for the leisure and the corporate business to be back, and that is displayed all over the network of Accor. And of course, we have to do more and we had to stretch it even further, commitment being made on carbon footprint by the year of 2030. We have to execute, and we will, on getting rid of all the plastic, unique guest-related item by 2022. And we committed, and we will achieve, of course, gender parity in the different bodies and governments of Accor. So that's -- it's not optional. We have Brune Poirson who actually joined 3 months ago, her, and her team. We're probably going to get back to you. I don't know the date. It could be end of the fall, it could be early 2022. But we will have a dedicated full day when it comes to ESG, CSR and being action driven and explain to you what has been undertaken, how, why, when. The last page, which is Page 21. It's a page you might have seen and recollect, but there were very few of you with us at the time of the general assembly 2.5 months ago. Those are the exact 5 -- the exact same 5 commitment priorities for 2021. There's no reason to change 7 months -- 7 weeks passing, sorry. The first one we have to be ready, and so are we, to benefit from the summer rebound wherever it is. And we have to be super agile. I told you it was great in the month of June in Australia, it's not as good in July, but it's going to be great again in September. We have to be ready when it's going to be reopening in Southeast Asia. We have to help South America to rebound. We have to be in Dubai to welcome. We have to be in Istanbul to Mexico -- or Tulum in Mexico, wherever it exists. I can tell you we have a team on the ground and we have hotels being opened to the extent of 98% -- or 93% of the group. Number two, no question. We have to deliver 100% and plus on RESET cost savings. Those EUR 200 million that Jean-Jacques talked to you about you can actually put that in the bank. It will be there, they will be sustainable and they're going to be helping us to improve margins moving forward. Number three, we have a very powerful Accor Live Limitless loyalty program. It's being reinvented. We've been launching the new branded card with BNP and Visa as of 3 weeks ago. It's being displayed as we speak. We're going to have probably great success about it. We've been signing major partnerships with a lot of people into entertainment, car rental and many other industries. It's -- yes, it is a good program, and we have 68 million members. It is a plan for people to use it. And of course, they do to get access to better pricing. Number four, we have to continue with developing our brands in different markets, and I think Jean-Jacques touched upon it. We talked to you a bit about Lifestyle. Lifestyle was less than 4% of Accor over the last 20 years. It is more than 15% of all the signings over the last 6 months and it is actually 29% of the fees value that we've been signing the last 6 months. So it has really been progressing very well with the 13 brands under Ennismore. And the pipeline is strong. It's not falling at all and 36% of the pipeline is to the luxury, which is far more than the existing network of Accor. And finally, on the five, talent is key. Human capital is all about Accor, which is why we've launched this Accor ALL Heartist. This is why we've been so present in the field vis-a-vis the owners and the government agencies. We are all over in 110 countries of Accor and we are able to be all of it because we have team on the ground. And having a team on the ground, you also have to show that you're there to help and we're going from being really run by being compensating what we do to do exactly the reverse. We need to positively contribute through AccorHotel activities to the local community. So that's really a 180-degree shift on really help all the people around us, be it education, training, employees, handicap, and of course, anything which is local, full procurement. This is who we are, and we have to make it a great advantage because of being a management, operating company with 60 years of existence. That's where we are. Happy now to take questions with each of you on the phone. Thank you for listening for the last, whatever, 35 or 40 minutes. Floor is back to you.

[Operator Instructions] And the first question comes from the line of Jamie Rollo from Morgan Stanley.

I've got 3 questions, please. First, you've given some encouraging data on the RevPAR improvement into July. Are you expecting that momentum to weaken after the summer as sort of leisure travel slows? And are there any sort of early indications of corporate demand you could perhaps share with us, please? Secondly, on the sensitivity guidance of RevPAR to EBITDA, I appreciate this sounds like there are some sort of one-offs, if you like, in the first half. But the full year guidance of EUR 18 million, that implies something like sort of EUR 19.5 million to EUR 20 million sensitivity in the second half of the year. So I'm just wondering why it would be sort of worse than the original guidance? And then, finally, could you talk a bit more about the slowdown in signings you're seeing? It sounds a little bit more cautious sort of the last update. Where are you seeing that? And I sort of note a drop in the Asia Pacific pipeline. So is that most of the area of the slowdown?

Jamie, this is Jean-Jacques. Let me take the sensitivity question because it's a very important question. I'm not saying that the sensitivity in H2 will be EUR 20 million, that's your computation. The thing I'm saying is I'm saying that the sensitivity will below EUR 18 million. So I'm very encouraged by what we saw in H1, the EUR 16 million. And the EUR 16 million is, in fact, a reflection of very good work that has been done on RESET and on the management and the cost. And that's the first goal, and that will continue. And in fact, you will see out of the EUR 70 million of RESET that I was mentioning a larger part of the saving in the second part of the year than in the first part of the year because just to take a practical example here, I mean, the France program coming in action right now will drive savings -- additional savings on that line for the second part of the year. So I'm going your way here at saying the EUR 20 million looks like the wrong computation. On the other hand, what I'd like to just also say is that if you just go [ that half ], furlough schemes are planned to be stopped in most jurisdictions in the second part of the year. Today, in fact, most of the countries have pursued the furlough schemes. So that's a counterpart of what I just said on people cost. You saw that Mantra has been running on all cylinders in H1 versus what we anticipated. But we also see the RevPAR today in Australia being significantly down in July. How long we did last, God knows, we don't talk about a lot of cases, but Australia has been extremely, I would say, stringent on making sure that their country is totally at 0 COVID cases kind of target. So the effect of that is another moving target. Then you've got the incentive. I just want to incentive and explain that it was a good piece of news in H1. Now the incentive, if Australia doesn't go as well as in H1, you won't that incentive kick for Australia. So that's another element, which is moving in the wrong way. On the other hand, volume should be much better in H2 than in H1. I mean we are just talking about the RevPAR improvement. And so the Q3, as we had been disclosing, should be -- disclosing sorry, in the Q2 call, should be, in fact, much stronger in term of numbers at the top line level. And in the end, the RevPAR sensitivity is first affected by volume. I mean one of the issues that we have is we don't have enough business, hence, the fact that we've got what we've got. And so that's going the opposite way. I'm trying to go a little bit into that question with [ Lance ] because it is a very core question, and I know that all of you are asking or thinking about the same thing. And I just want to make sure that I get a comprehensive answer. So bottom line of it is that we have been doing extremely well. All the plan that we've been launching are in line, and in fact, the focus is doing better. The top line is still imprévisible. I mean what I just explained Australia is another one. What Sebastien started with on the U.K. is another one. But we'll be moving in the right direction. So I'm very confident that we will be below EUR 18 million for the full year.

Yes. On the -- 2 things, Jamie. Sebastien here. On the first question, it's -- we have a pretty good visibility for the next 8 weeks to come, certainly when it comes to August and September and the leisure business, which is strong and stronger than we expected. We have very low visibility for corporate MICE activities for the months of October, November, December. There's good signs, which is you have a lot of small groups between 30 to 80 people in regional domestic corporates clearly rebounding, asking for reservation for those months of October, November, December. We still don't know whether we're going to have large, 300, 400 seminars across regional. It might happen and being confirmed in September. It's still a question mark. But I really believe we probably could buffer the loss of those large events to probably gathering hundreds of small groups of 30 to 50 people. But I guess we'll know better by the end of September. When it comes to signing, I have no fear at all that we -- we have 25,000 rooms being signed for the H1. That was less than 20,000 last year. That was 35,000 the year before. You know why that is? It has nothing to do with lack of pace of activities or lack of attractiveness for the brand. It just happened, as you know, that 50% more of all the signings for the last 5 years happened to be in Asia Pacific. And in Asia Pacific, 90% of all our developers could not travel. You cannot move away from Singapore. You cannot enter Cambodia. You cannot leave China. You cannot enter Malaysia. So it's not they cannot go and sign. So the owners are there. The documents are basically being negotiated. But when you sign something for EUR 100 million-plus commitment for 25 years, of course, you need to have a physical attendance to be in front of that owner. And that owner has many questions in terms of the context on where we sit. So it's only a matter of getting that vaccination rolled out, getting frontiers being reopened. And no hesitation, you'll see the pace of signing coming back to the level of prepandemic only in a matter of the next few months ahead. So it's frontiers related.

And so can I just come back, Jean-Jacques, on the EBITDA sensitivity. Sorry for overanalyzing it. I was just going off the guidance where it does say slightly below EUR 18 million. So I guess you're saying we should just remove the word slightly?

Okay. As part of the deal, that will close with you today, I would go for that. Let's do that.

The next question comes from the line of Jaafar Mestari from Exane BNP Paribas.

I just have one question really. I wondered, if possible, to come back on the drivers the EBITDA improvement in Management & Franchise. So M&F revenue -- if I'm looking at it year-on-year as you did, M&F revenue is EUR 24 million better and M&F EBITDA is EUR 55 million better. How much of that is the RESET cost savings, which I think you mentioned, Jean-Jacques, was the most important factor. How much of that is incentive fees? Your comments suggested that in regions like the Middle East, it's already very material, but perhaps not already at play elsewhere. And then how much of that would be lower-quality stuff like lower provisions? Is that material? You've already flagged that in H2. Yes, if you could break down that EUR 55 million, that would be helpful.

Yes. Let's make it rough, okay, without going into the brunt of all the numbers. To make it simple, there is a 1/3, which is the 3 buckets that's what you said. There is a 1/3 of it, which is incentive. There is a 1/3 of it, which is the fact that last year we did take significant bad debt accrual in the account because nobody knew where this crisis could go. And there is 1/3 of it, which is the net of everything else, which is RESET, I would say, more generally. Because again, after that, you can go into much more granularity. You've got the furlough, which last year was stronger than what you've got today. So the last bucket, which is the recent bucket, encompasses variable cost, RESET and other elements just like the follow of some accrual on bonuses and those kind of things. So make it simple, 1/3, 1/3, 1/3.

Okay. So RESET itself is part of 1/3. It's a lot less than 1/3 so far. And then you said in H2 the RESET cost or something...

No. Because -- no, no, because I said that the last bucket was net of many things. And so the 1/3 is correct. And the fact that the net of everything else besides incentives and bad debt is really 1/3. Is that clear because I just don't want to confuse you, Jaafar.

It's clear for Jaafar.

I can -- I just want to make sure because I don't want to confuse everybody. It is a complex matter.

The next question comes from the line of Simon LeChipre from Stifel.

Three questions, please. First of all, on the trading update, you just shared your positive expectation for the summer season. But could you perhaps share with us the sort of RevPAR range for the rest of Q3 compared with the minus 45% of July or at least August, that would be very helpful. And also, could you give us more detail on the pricing dynamic that you see at the moment? And lastly, what should we expect in terms of working capital for H2?

Okay. So I'll take the working capital one. Just like for sensitivity to RevPAR, I'm very happy with what we got. I said that in the comments but -- in answering your question, I'm very happy with what we did on the sensitivity, cash burn numbers. And part of it is coming from working capital. The guidance that we had given to you back in February still holds true, i.e., 0 net working capital change for the full year is a good number for net working capital. So I'm trying here to give you granularity. And the reason for why I can say that is we know that always H2 is a positive number, whereas H1 is a negative number. So the H1 has been a negative number. But a negative number that has been, I would say, better than what you could have anticipated. And so the H2 will be positive and the working capital target of being positive for the full year holds. So that's, I think, the answer to your question on working cap.

Simon, on the other one, I -- we won't give it to you simply because we don't have enough information. The only thing that I alluded to in my first slide, which I hope will be still true for the next months to come, certainly in the next 2 months to come, I told you the last 4 months we had the benefit of the 5% RevPAR improvement, funny enough, almost symmetrical for the last 3 months. I am hopeful that would be the case for August. I don't know for September, but it might very well be the case. But I cannot prolong that 5% because then coming the fall, then you have a better weight of corporate business versus leisure business. So too soon for us to give you a guidance even on the next quarter. The second thing on pricing, it is so, so nonhomogeneous. In places in which you go above 85% to 90% occupancy, as I told you on the Mediterranean shore, you get better pricing that you had in 2019 because you know you're full and you have pent-up demand. In other places, urban cities in Europe, when you reach a 38% occupancy, of course, your price is still way, way below what you could have seen in 2019. So because it is so nonhomogeneous, I can't give you. It's only a matter of where you sit and which timing you act.

Your next question comes from the line of Bilal Aziz from UBS.

Two for me, please. Just a quick point on the incentive fees. Firstly, could you disclose the exact number you did book in the second quarter? And tied to that, are you now starting to see some reappear in Europe so far in the third quarter? Or perhaps what trigger points of occupancy do you expect that to occur in Europe given your portfolio? And then secondly, some of your U.S. peers have started to talk about increased booking windows with the type of booking now extending. Perhaps just what you're seeing there for perhaps next year right now with advanced bookings?

Okay. So on the incentive, I will give you the precise number because if you look at our financial statements, you can probably get it. So I had mentioned historically 45% as a run rate basis of what is the incentive percentage of the total M&F revenue. Last year, you may recall we said that we would be at 8%. So 8% of the total M&F revenue was incentive. And as of H1, what was booked in 2021 was 18%. So that as a percent and just to give you how it is improving. Now the absolute value, so that you don't have to do the computation. EUR 169 million in 2019, EUR 12 million in 2020 and about EUR 29 million in 2021.

On the booking window, I'll give you 2 numbers, and of course, it also varies per week. For Accor at the group level, 50% of bookings have less than 3 days notice. So -- and in Europe, which I've learned yesterday, 1/3 -- Northern Europe, by the way, 1/3 of Northern Europe working at less than 12 hours. And that's -- so you have to understand it. People are waiting for the very last minute to understand the rules of traveling. Of course, that number varies quite a bit. If you're going really out of the way by car, then you can travel in advance and that's okay. So that's what we're facing. So I hope that, I guess, we're going to have a better notice and probably enlarging the 3 days to probably a week, but we're not there yet.

And just as a complement on this one. What matters is booking, but also the cancellation level. I mean you saw a huge increase in booking window and stuff in the States, but you also saw a lot of cancellation following that when rules were changed. And so you also need -- and hence, the comment that I made when I was talking about the fourth wave of COVID in France and what has been happening. You really have to monitor what's happening on cancellation. And today, based on actuals, I can tell you that we don't see increased level of cancellation.

The next question comes from the line of Richard Clarke from Bernstein.

Three, if I may. Just the first one on Australia, obviously that was a bit of a bright spot in the first half. Lots of sort of news of lockdowns in Sydney, other cities reopening. Maybe you could just sort of unpick as to how your performance has sort of trended in the last months and your expectations of what happened in the second half based on the news there. And second question, the 7% of hotels that are still closed, maybe just some details? Where are those? Is your -- and what is the time frame do you believe for those to reopen now? And then third question is your M&F revenues improved by EUR 24 million. Your SMDL revenues declined by EUR 24 million. And just some commentary on why those are moving in opposite directions. Is that just a regional point? Is that the incentive fees coming in? Or is there some discounts or anything else being put through there?

I'll take the last one because it's something that I should have quoted. It's a very good catch, it's a very good question. In fact, there is no renegotiation whatsoever of fees. It's a very important point. Throughout all the crisis, one of the things that has been extremely sound and positive is that we didn't get into any discussion of changing the fee scheme. What's occurring on SMDL is the fact that, today, when people go and book the hotel, because of the COVID crisis, they go and -- directly go visit the hotel. So they go to the counter and basically book the hotel room much more than what we've been seeing that in history. Quantifying of that is that there is about close to 60% of the business, which is done this way when it used to be less than 40% in history. The consequence of that is there is no fees, which are attached to the bookings that you are doing in the hotel directly, no distribution fees because the guy directly going to the counter, if you will. And so that's why the revenue that you've got in SMDL, everything being equal, is down because of the mix of direct versus indirect or seeable channel, I should say, versus nonseeable channel. That's one element. The other one, which, again, is very much linked to the COVID crisis that we've got significant quarantine business. In Southeast Asia, it's probably 30% of the business, which is quarantine. Quarantine meaning that the government forces people who want to enter into the country to go in a hotel and they have got to pay the fees for that themselves. In Singapore, 80% of our business is quarantine business. So the consequence again of the quarantine business is that there is no sales, marketing, distribution, loyalty fees attached to that. So I think that's the answer on SMDL. In term of Australia, in Australia, you've got the 2 very different picture. You've got the coast, the Gold Coast, which is doing extremely well. And on this one, the RevPAR that we see versus last year at the end of June are up 20%. So very significant increase. We're always talking negative number. Here, it's a positive number and it's very significant.

That's versus 2019.

That's versus 2019. Yes, yes.

You said last year.

Oh, sorry. 2019.

20% versus '19, and it's way up 50% versus last year.

Yes. Yes. Sorry, I was -- and on the other hand, the big cities, which are much more depending on international traveler and much more depending on fundamental business, are still reporting a negative number. So they are at minus 25%, 30% versus 2019. And so you've got this very different picture which, by the way, it's no different than what you find in France, what you find in the U.K. in that 2-pronged recovery that we're going to face over Q3. The leisure business, the domestic business is very good. Anything which is international is not there. And so the portion of the business, which is affected by the closing of Sydney and Melbourne, and by the way, it is more than Sydney and Melbourne, which are affected, it has expanded to other like Perth and these kinds of cities in Australia. That's the portion that we are going to monitor over Q3 to see how well they recover. I always want to be on the positive side here by saying that you talk about a very limited number. Okay, so it's not as if you've got a contagion, which is extremely difficult to monitor. But with the vaccination rates in Australia, which is below 15%, they've got no choice than to be so fantastic.

And Richard, just on that because we spend a lot of time with Simon McGrath, who is the CEO of Pacific, he is really positive. Again, when you get out of that confinement for Sydney, there is a lot of wealth a lot of actually money put aside for the domestic clientele of Australia. They cannot go elsewhere other than stay in Australia. So that money will be spent as it was in January, February in the Golden Coast. It is winter, as you know, over there. Summer is going to back, so near the spring, in the months of October, November, December. So he's looking forward for very good activity on Mantra when it comes to November and December.

And then the 7% of hotels that are closed still today, what's the time frame for those to reopen from here?

It's -- a lot of them are in Southeast Asia. A lot of them are in South America, basically where frontiers are being closed.

The next question comes from the line of Leo Carrington from Credit Suisse.

Could I just ask some -- a few questions on RESET, please? Firstly, are you still expecting the same split of the RESET plan between 60% going to SMDL and the remaining elsewhere? And then just assuming -- and apologies if this is semantics, but is there any difference between savings hitting the bottom line and savings to be retained in outer years? Reason I ask is how do you think, if fully retained, how do you think your owners will perceive SMDL profits once this RESET plan is complete? And will they not be sort of asking for it back one way or the next? And then, lastly, just a related point, but what kind of EBITDA dynamics do you expect in H2 for SMDL? Do you expect this will improve as the SMDL revenues improve? Or any color there would be fantastic.

Yes. The 60-40 is still right. There is no change to that one. As I said, the plan has not fundamentally changed by any means. There are as in every plan, situation as you progress. But the key numbers, the 60-40, EUR 200 million, the EUR 70 million, the time end of 2022, all of that is holding. So no change on this one. On SMDL, I mean, before we get back to a positive line on SMDL, there will be some time that still needs to come to pass through, and as I said, we are still very negative. That's a question that I got earlier on SMDL. And so we've got some time here to get back to something, which is positive. And the other thing that I would say, again, to the difference of many of our competitors, we have no obligations whatsoever in term of reporting what is the profitability that we do in SMDL. I know that in the States, it's part of the contractual obligations and part of the contractual discussions. We have that on Fairmont, right, so on what we bought in the U.S. We don't have that in any other places in the world. So I think that's the answer on this one. And on H2, on SMDL, you would think, everything being equal, that as business will start to kick in from September on, you would have much more of the business, which is booked through the typical channel and not directly by people stopping their car and going to the hotel counter or following directly to the hotel because things don't work, I don't know what. So you would think that, that number should improve on SMDL everything being equal. That's what I will answer on this one.

Okay. And just on the -- is there any difference with the RESET plan between savings hitting the bottom line and being retained in outer years? I'm just thinking of your sort of 2024 margin with this question.

No. There is no difference. I mean what we kind of went through in another call is that, everything being equal, the margin that you would do on the M&F business would probably be significantly improved and be north of 70% at the end of the plan because those savings will make us more cost base lower, and hence, much more profitable on the M&F business. I think that's the answer that I gave, I don't know, in February to one of the question I got. So -- and on the rest, again, you will find the EUR 200 million at the bottom line.

The next question comes from the line of Andre Juillard from Deutsche Bank.

Two questions, if I may. If I continue on the RESET plan and the margin, just wanted to ask you if on the EUR 200 million expected of savings, do you think that a part of them will have to be reinvested in some purposes? I'm especially thinking about distribution. First question. Second question is more general about M&A. Do you any see window for potential Mantra disposal? And what do you think, more generally consolidation, about in the industry considering that we are starting to see the exit of this crisis. But all the groups, you included, are looking for further rationalization and consolidation could be a part of the answer.

So the question to your first answer is no. There is no obligation whatsoever to do any investment. I think what will happen and happens today and will continue to happen in the future and has happened in history is we need to continue to invest into the company, I don't know, in system, in brands and stuff like that. So all of that will be pursued obviously. But I'm not counting or there is no allocation per se of those EUR 200 million to any of those actions. It is things which are different. Had we not done the RESET plan, we would have still invested into better improvement of CRS, a better tools for the apps, improvement in marketing, improvement on brand recognition, all kind of things. So I would not link the 2, Andre. And on M&A...

And on M&A, Andre, there's 2 answers to your 2 questions. On Mantra, it is not intended at all to dispose of Mantra for the next 18 months to 24 months. Why? It's because it's so elastic. We've been losing EUR 50 million-plus of EBITDA with the downturn of the Australian market. We need to get it back and we are getting it back as we speak. And you need to have it back in the bank in order for people to buy the positive EBITDA above leasing commitment. So -- and it would be foolish and stupid to sell it today, and we don't have any reason to sell it today. So medium term to long term, we confirm that, I guess, we will be disposing of all the HotelAssets components of Accor because it is something that we must and should do to simplify the balance sheet and simplify the reading of Accor. But for the next 18 to 24 months, it is not marketed at all and it shouldn't be marketed. When it comes to consolidation, I can tell you 150% of my time and the time of the management team is on day-to-day activities we set being finished, implemented. Commercial people, GM to be ready for receiving any new customers, responding to legislations and any vaccine rolled out. So no, we have no time and no will to spend any distractions when it comes to consolidation.

The next question comes from the line of Alex Brignall from Redburn.

I just have 2, please. The first one, you mentioned in your discussing about system growth for the full year that churn was higher and that would be a factor. Can you just give us a little bit of detail on that and the drivers of that, please? And then second the question is on the loyalty and the plans you have for that and the changes you're making, could you just give us a little bit of information on what was that like pre-COVID, maybe the mix between business travel and leisure or anything else that might be significant?

Yes. I'll take the one on the churn. So to be specific and give you a precise number, the churn for the last 12 months is 2.4%. The churn for last year was 2.2%. The churn for FY 2018 was 2.1%. And the churn for 2017 was 3.1%. So that's kind of why I said we are more or less in line with what you would expect. I think we have before said around 2%. So that's more or less the model as you would expect it to be running on a normalized basis, Alex.

And on the loyalty contribution, which is higher than 30% at the group level, for the last few weeks, of course, more than 2/3 of those are being burned and used by leisure customers, which is very much correlated to the summer season. So we'll have a better reading by the end of this year when corporate is going to be back in the fall.

Yes.

Can you say what that number was before COVID in terms of leisure versus business, please?

Probably very similar.

The loyalty numbers on this one, Alex, and why it's not worth spending too much time on those is you've got a large part of what's happening today, which is so much COVID related that it creates numbers that makes no sense. If we come back to what we discussed on currency and all the business that you've got in Southeast Asia and some business that we did get also in Australia, by the way, up to the end of H1 was guarantee business. When you've got guarantee business, you've got no loyalty whatsoever. You don't have any of those that are qualified.

We wish we could. We wish we could.

You're right, because the loyalty -- it's very loyal to COVID. So we wait for the next COVID crisis so that we can increase our loyalty numbers. In fact, let me think about that. But joke aside because it's not -- we really are having crazy numbers because of the phenomenon. So I think as soon as this has come back to a more, I would say, normalized level and more natural channel mix, we will be able to go into the right analysis.

The next question comes from the line of Ali Naqvi from HSBC.

Just wanted to ask on -- just on occupancy. I can see that it obviously varies by region, but there wasn't much of an improvement in Q2. Could you give us any insight as to what was going on in July? And how do you think the RevPAR recovery will balance itself for the rest of the year versus occupancy and rates?

Yes. I mean the rule of thumb that Sebastien kind of gave you on 5 points every month is what -- is mostly driven by occupancy. There is very little situation on the price per se. And so if you want to have a kind of a simple rule, occupancy should go in average up by 30%. And there will be differences by jurisdiction. I mean you should find out that Europe occupancy will grow much faster than the 5% that I just described.

Great. And just on the liquidity and when you get back to being cash positive, is there -- what is the sort of priority in terms of the use of the cash? Would you return it or would you increase your scale?

I mean we're going to go -- the business model is a business model into which share buyback and return to shareholder is part of the expectation of somebody who would invest into a company like ours. And you may recall that in January 2020, we had just launched a share buyback tranche and we were talking about having a great year on that perspective before whatever happened in one. So I think this remains a basis of how we will look at excessive cash or cash to be potentially returned. Now on the other hand, we will do what we are paid with, which is we will also analyze where the greatest value is created. So between share buyback and potentially M&A bolt-on. I think we were doing that with quite a lot of discipline over the last 2 years before COVID. You may recall that we returned over the last 2, 3 years about 5% of the market share every year if you do the computation. And this is very much what is the best market practice around that. And you would -- you should expect us to continue to act rationally and take into account what creates value for the shareholders.

Yes. Another way to say it and people will understand it very quickly. We have been prioritizing for the last 15 months, first, the employees of Accor, which is why we put the ALL Heartist Fund to make sure we protect human capital of this company, which is why we exist. Two, 3 months later, we've been protecting, assisting the owners of Accor Hotels wherever they sit. It's time for us post pandemic to really go and protect and enhance value for the shareholders. That is the sequence. And that sequence is starting now, of course.

There are currently no questions in the queue. [Operator Instructions]

I guess we had an enjoyable time. That was not a question, that was a comment. Unless we have no further questions, which I believe is the case, I thank each of you. It was a pleasure, of course. Pierre-Loup and the team are here to assist. And Jean-Jacques and I will do and hopefully do a great job talking, introducing investors and shareholders over the next few days. Thank you, everyone. Thank you all of you. Bye-bye.

Bye-bye.

Thank you for joining today's call. You may now disconnect your lines.