Ladies and gentlemen, welcome to the Casino Group 2021 half year results conference call. I now hand over to Mr. David Lubek, Chief Financial Officer of Casino Group. Sir, please go ahead. Thank you. Good morning, everyone. Welcome to Casino H1 Results conference call. I hope you're all remaining safe and well. A few words of introduction before going through our presentation. We have been living with the COVID-19 pandemic for the last 18 months. During this time, as you will see from our H1 results, we have successfully adapted and transformed our food retail business in France. Our profitability has improved and reached a satisfactory level in all our banners. With our mix of efficient urban and proximity formats, a dynamic expansion plan, and a strong food e-commerce proposition, we are now well-positioned to grow sustainably going forward. As for our other businesses, Cdiscount and GreenYellow have communicated their results this week. Both have shown excellent performances in their respective high growth markets based on secure mega trends, tech and e-commerce for Cdiscount and energy transition for GreenYellow. In Latin America, the value of our assets has more than doubled in a year, and Assaí, in particular, has performed extremely well in a tough environment. All of this, as you see, warrant our confident view of the future. Our presentation starts, as usual, with a summary of our H1 financial figures. Page two, the main results highlight, all our geographies showed increased profitability with EBITDA growing plus 11% and EBIT growing plus 24% at constant exchange rate. At the group level, we had stable sales during the semester with the negative impact of the pandemic on our sales in France more than offset by the positive impact of our transformation plans. Net normalized results and net results improved by respectively EUR 23 million and EUR 306 million. Page three, cash flows and debt. We will get back to these in detail later on. Net debt and cash flows in France will now be analyzed excluding GreenYellow, as mentioned during our full year results. GreenYellow will raise debt on its own to fund its transition to an asset-based model and does not rely on Casino for financing, which is taken into account in our new banking covenants. The main takeaway is that cash flows are in line with the usual seasonality of the business and close to the numbers of last year, even though sales were lower in Q2. Gross debt in France, the metric by which we track our deleveraging program, declined by EUR 438 million. Net debt in France declined by EUR 158 million. Net debt after IFRS 5 increased due to the reduction of IFRS 5 following the disposal of Leader Price. Now to the key takeaways from this semester, starting with France. Page four. The first highlight is the success of the transformation plans launched last year in all our food retail banners. Their trading profit margin increased by 81 basis points in H1. This is a particularly strong performance as it happened in the context of like-for-like sales declining by -7% due to a strong comparison basis in H1 2020 and COVID restrictions affecting our operations in Q2 2021, such as curfews, closure of Non-Food sales, and the specific impact in the Paris area with a lack of tourism and the impact of remote working. Despite declining sales due to these temporary effects, our EBIT improved again and reached EUR 166 million in H1 at the French retail perimeter, EUR 173 million including Cdiscount. Following the successful execution of our transformation plans, we are now well-positioned for strong and profitable growth in the second half of the year. Growth is our clear priority for H2, with a more normalized basis of comparison and increase in the size of our network. We opened 353 stores on proximity formats in H1, the sales of which will ramp up in H2, and we plan to open 400 more in H2. The second highlight of H1 is the strong growth of food e-commerce, which has increased by 103% over two years, significantly better than the market, which took advantage again from our key exclusive partnerships with two top technology players, Ocado and Amazon. The third highlight is our financial structure. After a successful refinancing of our Term Loan B, we have extended the maturity of our main revolver credit facility from 2023 to 2026 for EUR 1.8 billion and reviewed financial conditions and covenants. Our maintenance covenant at the end of Q2 was met very comfortably with both secured debt over EBITDA covered with a margin of EUR 369 million on EBITDA. We have also put in place a new RCF at Monoprix with, for the first time, sustainability-linked features. A word about our disposal plan. We announced two days ago an agreement with BNP Paribas for the disposal of our stake in FLOA Bank, securing EUR 179 million of total proceeds while maintaining an exposure to future value creation in fractions payment. We also secured close to EUR 100 million earn-outs from the Apollo and Fortress JVs. Taking this into account, the total value of firm disposal increases by close to EUR 300 million, reaching EUR 3.1 billion. Page five, Latin America. Assaí published its detailed results two days ago. GPA's results were published yesterday, their conference call is planned for this afternoon. I will concentrate on the key highlights. First, as in France, profitability increased again significantly in LATAM with an EBITDA margin at +96 basis points and trading profits growing 84 basis points. This led to an increase in trading profit by 33% at constant exchange rates for the whole segment. Assaí has done particularly well with +22% sales organic growth in Q2, despite the impact of COVID in Brazil. Second, the spin-off of Assaí has been a clear success. The value of Casino's share in Latin America assets has doubled since the spin-off announcement in September 2020 from EUR 1.1 billion to EUR 2.3 billion. Before going into our detailed results, let's move to review of our progress on our strategic priorities in France. Page seven. First, as mentioned in the introduction, we saw the benefit of our transformation plans launched in Q3 2020, with cost savings in store and back office of EUR 30 million per quarter in our food retail banners, increasing their profitability. Looking at the performance of our retail banners using a constant perimeter, that is French retail trading profit, excluding GreenYellow and Vindémia, the net improvement was plus EUR 49 million over H1, that is plus 50%. We have continued the digitalization of our network, with 613 stores now equipped with autonomous solutions and 63% of payments in hypermarkets, 58% in supermarkets done through automatic cashier or self scanning. This is a key edge that allows us to operate with lower cost while simplifying our customer's experience. We have also pursued the growth of our network. 353 proximity stores have been opened in H1, above the initial target of 300, in formats such as Franprix, Vival, Monop', Naturalia, Spar, or Casino Shop. The ramp-up of these stores will contribute to sales growth in Q3. Food e-commerce grew +15% year-on-year on a very strong comparison basis. On a two-year basis, growth is up +103%, far above the market. Our partnership with Amazon has been extended again, with two new cities included in the Monoprix offer on Amazon for same-day delivery, and our click and collect offer targeted in 180 new stores on top of the 600 Amazon Lockers already deployed. As for quick commerce, the offer is now available from 800 stores, thanks to our partnerships with Uber Eats and Deliveroo, and the rollout of our Franprix direct delivery offer. For H2, our priority is clearly focusing on growth. We have a number of brands, all of which are profitable and ready to expand further. On top of the ramp-up of the 350 stores already open in H1, we will grow our network further with the opening of 400 new stores. We also expect further acceleration of our e-commerce operations. Combined with a more normalized basis of comparison for our stores and the impact of our commercial initiatives, we expect these elements to deliver a substantial return to profitable growth in H2. Turning to page eight. A key feature of our strategy is fostering commercial innovation in our various banners attuned to customer needs. This slide presents a number of relevant examples of our innovations, each of them consistent with customer needs in specific areas and clearly aligned with our banners, values, and our CSR commitments. At Monoprix, customer service has been expanded to include concepts around health with personalized advice, local products from a distance of less than 100 km, and an offer dedicated to urban mobility. Franprix continues to expand its network with 160 openings planned over 2021 and 2022, especially in the surroundings of Paris, with a strong emphasis on local services, including evening catering. Casino banners in hyper and supermarkets are also focused on innovation with the introduction of artificial intelligence solutions in store and specialized corners rolled out with partners and startups. Nine small Géant stores have been converted into Casino supermarkets with an offer adapted to local needs and a significant uplift in sales. Page nine, moving to Cdiscount. Cdiscount has continued to actively focus on its key product growth priorities: marketplace, digital marketing, and Octopia. First, the marketplace, with growth of +10% over the semester, +33% over two years. It now represents 46% of GMV. Marketplace revenue increased by 17% over the semester, +39% over two years, to reach EUR 199 million over the last 12 months. Cdiscount's ecosystem of 14,000 vendors and 100 million SKUs is driving the success along with its top-quality logistics facilities, which helped expand express delivery for third-party vendors. As an example, Cdiscount now has 2.2 million SKUs eligible for express delivery, compared to 1.3 million last year. Second, digital marketing, with a very strong +44% in H1, +72% over two years. Cdiscount is taking advantage of the shift to retail online advertising. In particular, with its new solution, Cdiscount Ads Retail Solution, CARS, 100% self-care advertising platform, enabling both sellers and suppliers to promote their products and brands. Third, Octopia, Cdiscount's new B2B business dedicated to turnkey solutions for online retailers. Octopia offers ready-to-operate services to international retailers and e-merchants. It includes all four key elements needed to operate a successful marketplace. Products-as-a-Service, Fulfillment-as-a-Service, Merchants-as-a-Service, and Marketplace-as-a-Service. It has shown a strong start with Product-as-a-Service and Fulfillment-as-a-Service growing +60%, while Merchants and Marketplace -as- a -Service, which have just begun their marketing to customers, have had a very promising start. This led to an EBITDA of EUR 49 million in H1, +148% over two years, stable compared to H1 2020, which benefited from the exceptional inflow of the first lockdown. Compared to the very high level reached in H1 2020, total GMV grew +2%. That is a 14% growth over two years. In H2, Cdiscount will continue to roll out its plans for further high topline growth in marketplace, digital marketing, and Octopia, delivering again significant increase in its EBITDA. On our new businesses in B2B. On page 10, a few words on relevanC of data monetization and B2B retail tech business. A key recent milestone for relevanC has been a signing of a new partnership with Google Cloud and Accenture, which should provide a significant boost to its development. relevanC received the status of premier partner, recognizing its expertise and integrating its solutions within the Google Cloud B2B marketplace. In H2, relevanC will continue to pursue this partnership strategy and plans to accelerate its growth both in France and abroad, notably through the Google partnership. Turning to page 11, GreenYellow. GreenYellow is positioned at the heart of the decentralized energy transition market. It addresses its diversified corporate customer base with its unique model, combining energy saving solutions and self-consumption based on local solar production. GreenYellow has been accelerating on its two key business lines in the last semester. Compared to June 2020, the advanced pipeline of photovoltaic projects is now up 85%, and the advanced pipeline of energy efficiency is up +78%. The advanced pipeline of photovoltaic now stands at 809 MW, with an additional pipeline of opportunities of 3.5 GW. In energy efficiency, the advanced pipeline has reached 350 GW h, with an additional pipeline of opportunities of close to 900 GW h. GreenYellow successfully operates in 16 countries in four continents. It is planning further expansion, with expansion of its operation to Eastern Europe, starting with the launch of its first 4-MW project in Bulgaria and plans for Poland and Hungary. Among its recent projects, the extension of the largest power plant in Madagascar to double its capacity from 20 MW to 40 MW. As mentioned previously, the company is pursuing its transition towards an infrastructure operating model, owning the assets on the long term. GreenYellow disclosed this week an EBITDA of EUR 37 million for H1, excluding asset disposals. That is an increase of 40% year-on-year. Further growth of EBITDA is planned for H2, as detailed by GreenYellow in their recent publication. Finally, page 12, a few words on our CSR policy and commitments, which are an important and long-standing priority of Casino and a distinctive feature of our group. First, we have reinforced our commitment on greenhouse gas emissions reduction. Our goal is to cut carbon emissions by -38% by 2030. This trajectory is in line with the well below 2 degrees scenario. Actions have been taken in all our geographies, leveraging on GreenYellow's expertise. Among the most recent initiatives, the partnership between GreenYellow and Franprix to reduce the carbon footprint of refrigeration units, and carbon neutral refrigerant gases in Carulla FreshMarket store in Colombia. As for Cdiscount, it has now reached carbon neutral status for its deliveries. Second, we have a strong focus on promoting responsible consumption. The share of organic products has increased by 0.9 points in France in H1. Bulk concepts are being rolled out in our various banners in partnerships with national brands. We are reducing our paper consumption with a transition to virtual discount coupons and virtual receipts as well, notably through our Casino Max app. Our commitment to responsible consumption has now been translated for the first time into one of our financial instruments with the inclusion of CSR objectives in the new Monoprix syndicated facility. The margin will be adjusted every year based on greenhouse gas emissions, share responsible labels, and share of vegetable protein products. Finally, we have continued our commitment to solidarity with a new partnership for culture in our proximity store in rural areas with the Fondation Marc Ladreit de Lacharrière, and food drives for students in financial difficulty organized at Casino stores in partnership with food banks. Moving now to our financial results. Page 14, a few preliminary comments on accounting standards. 2020 and 2021, H1 accounts have been restated following the divestment of Leader Price, classified as discontinued operations for IFRS 5 standards. The gradual conversion of the store sold to Aldi is expected to be completed by end September. At end July, around 400 stores were already transferred. As already mentioned in our 2020 full-year results, H1 2020 accounts have been restated to take into account the decision of the IFRS interpretation committee on the enforceable periods of leases. Page 15. The key figures of half-year results are shown in this table, the total change at constant exchange rates. Net sales reached EUR 14.5 billion, stable and organic. EBITDA was EUR 1.099 billion, up 11% at constant exchange rates, with good performance in all our segments. Trading profit stands at EUR 444 million, up 23.5% at constant exchange rate. Underlying net profit group share is EUR -72 million, which is a EUR +23 million improvement compared to H1 2020. The net result from continuing operations up EUR +306 million. Net debt before IFRS 5 was stable, with a reduction of EUR -158 million in France. Net debt after IFRS 5 increased due to the reduction in IFRS 5 with the advance of the disposal plan. Before detailing the results, a few words on Q2 sales on page 16. As expected, net sales were down in Q2 due to two factors. First, a very high comparison basis during Q2 2020, and second, temporary COVID restrictions in Q2 2021, which particularly affected our formats. Over two years, like-for-like growth is up + 6% at group level, with + 12% in Assaí and slightly negative in France at minus 1.2%. Let's move to slide 17 for the detailed analysis of Q2 sales in France. Same-store sales variations was minus 8.4%, with a drop in all store formats that is a bit higher than in Q1. This drop results from 2 factors. An exceptionally high level last year, with growth of +7.9% in France last year during Q2, including +6% for France retail during lockdown. Importantly, tougher health restriction in Fall during Q2 this year impacted some of our formats quite severely. The closure of non-essential product sections weighed on Monoprix and hypermarkets. The decline in tourist numbers and the reduction of Paris customers, accentuated by the third lockdown, affected Monoprix and Franprix. Finally, the curfews forced the early closure of our autonomous stores at 7:00 P.M. instead of 9:00 P.M. or later in the evening, with a negative impact on sales. In two of these restrictions, our banners showed a good resilience, particularly Monoprix and Franprix, which have clearly outperformed the Parisian market. The restrictions were, for the most part, a temporary phenomenon, and after the lifting in June, we have seen an improvement in our sales trend, as shown on the following slide. Page 18. As you can see from the table, all our banners have shown an inflection in their sales trend in the last four weeks compared to Q2, especially in the urban and convenience banners, with Franprix up + 8 points and proximity up close to 14 points. On average, our French banners have shown an improvement of 4.4 points in the last four weeks compared to Q2 numbers. Cdiscount has had a particularly strong rebound of 19.6 points in June-May growth. Part of it is due to anticipated summer sales, and part of it to the structural improvements with a more normalized basis of comparison. As I stressed in the beginning of this presentation, our focus for H2, where the basis of comparison will be normalized, will be a key return to growth, a profitable growth based on the strength of our models. Our first numbers for July clearly point in this direction. Moving to the results in France. Page 19. Overall, France operations recorded a notable improvement in profitability, with an EBITDA margin reaching 8% at more than 100 basis points compared to H1 2020. Trading profit was up 8.2% and reached a 2.2% margin. This increase in profitability, which is particularly remarkable in the context of tough sales dynamic, is due to the success of our transformation plan, which have been delivering consistent improvements in EBITDA margin since Q3 last year. Page 20. Moving on to a more detailed analysis of our results of the France retail segment. Total EBIT at France Retail increased by 8% in H1, including the negative impacts of the sale of Vindémia. Focusing on our current retail banners, that is excluding property development, GreenYellow and Vindémia, which is no longer part of the group. The improvement is even more impressive with trading profits growing + 50% from EUR 97 million to EUR 146 million, and trading profit margin moving from 1.3% to 2.1%. This reflects, once again, the impact of the transformation plan launched in Q3 2020, which reduced our cost base by EUR 30 million per quarter. Combined with the reduction of COVID-19 related costs, this plan allowed our EBITDA margin to move up 115 basis points, despite net sales down - 7.3% over the semester. With the strong level of profitability now reached in all our banners, we are ready to expand our network and take advantage of a return to sales growth to move our EBITDA even higher. Page 21. Moving to the e-commerce segment, Cdiscount. I've already gone into some details on Cdiscount's levers for growth and profits. Total GMV is up +2% on a strong comparison basis, that is +14% over two years. The key indicator that we monitor for growth, marketplace GMV, was up +10.5% over one year, +33% over two years, with marketplace share of GMV growing 4 points to 46% of total GMV. Digital marketing, again, showed a particularly strong momentum as +44%. As mentioned before, we expect Cdiscount to continue to deliver in H2 strong momentum on its marketplace, digital marketing, and Octopia business lines. Page 22. Moving to Latin America. Assaí, GPA, and Éxito have published their detailed results this week. I will focus on the main highlights. In short, H1 showed a strong increase in profitability in all of our business units, despite tough sanitary restrictions affecting sales in Brazil and Colombia. Total sales were up +6.9% in H1 at constant exchange rate, thanks to the strong dynamic of Assaí. Assaí managed to grow both their sales and their profitability despite lower demand from B2B customers, thanks to the strength of their offer to B2C customer and the success of the expansion plan with 19 stores opened in the last 12 months. Assaí grew +22% in Q2 with 9.2% like-for-like and +13.2% expansion. Three stores have been inaugurated in Q2, and 25 stores are under construction in 14 states in line with the plan. EBITDA was up +36% at constant exchange rate, same store sales, thanks to the successful ramp-up of recent openings. Multivarejo sales declined -6.4% at constant exchange rates due to those restrictions imposed to contain the new wave of the pandemic and the strong comparison basis in 2020. Despite these challenging conditions, GPA managed to increase its trading profit in Brazil by 32% in H1, thanks to strong operational efficiency plans. The digitalization of the business have also accelerated, with online sales growing in Q2 by 32% compared to Q2 2020. Grupo Éxito net sales were -3.6% lower at constant exchange rates due to the closure of stores as a result of the pandemic and also disturbances caused by protests in Colombia. Despite these challenging conditions, Éxito increased its trading profit by 15% in H1, driven by higher contributions to our complementary businesses, improvements in real estate, and good performance in Non-Food at Éxito WOW, an innovative model that allows digitally connected hypermarkets to combine digital channels and brick and mortar services. Overall, trading profit was up 33%, that is + 30% excluding tax credits, with EBIT margin up 73 basis points. This led to trading profit up 13.5% in euros at EUR 271 million versus EUR 239 million last year, taking into account a EUR -47 million currency effect. Page 23, underlying net profit. Group share is up EUR 23 million compared to last year, mainly driven by the increase in trading profit. This result includes a negative one-off non-cash impact on financial expenses of EUR -40 million, linked to the refinancing of the Term Loan B with the accelerated amortization of the setup costs of the original Term Loan B. The new Term Loan B, which bears a lower interest, 4% instead of 5.5%, will generate recurring yearly savings in financial expenses of EUR 9 million. Page 24, net results group share. Net result group share improved by EUR 306 million compared to last year, driven by the growth in EBIT and a strong improvement in exceptional items and other financial expenses. Moving on to our disposal plan, page 25. This semester showed further progress on our EUR 4.5 billion disposal plan. The total assigned or secured disposal is now EUR 3.1 billion compared to EUR 2.8 billion at the beginning of the year. First, we secured around EUR 100 million of earn-out from the JVs with Apollo and Fortress, thanks to the group progress of the JV disposals. This level of earn-out is now effectively secured by the disposals already realized or signed by the JVs and has therefore been recognized in our accounts. This level is a minimum, and if the remaining disposals go according to our objectives, the actual earn-out, which we expect end of 2021 or beginning of 2022, should be higher. Second, the disposal of FLOA Bank was announced two days ago. We have signed an agreement with BNP Paribas, which will provide a total cash-in at the closing of EUR 179 million, including EUR 129 million for the disposal of our 50% stake in FLOA, and EUR 50 million for the new partnerships put in place with BNP Paribas. On top of that, Casino will remain associated with the successful development of FLOA's fractions payment activity through a 30% stake in future value created by this business through an earn-out in 2025. This partnership also secures the current conditions for the financing of Cdiscount customer finance. As mentioned before, we are committed to the completion of our EUR 4.5 billion disposal plan in France. Page 26. This page shows the evolution of net debt by entity as usual. In France, net debt excluding GreenYellow declined by EUR 210 million compared to last June. Cdiscount net debt increased slightly by EUR 52 million due to temporary working capital impact. GreenYellow net cash position decreased by EUR 115 million in nine, with a ramp-up of its investments financed by its own resources. We now monitor Casino France net debt and cash flow excluding GreenYellow, since GreenYellow does not rely on Casino for its financing and has been excluded from the computation of our new RCF covenants. Page 27. Moving to the detailed cash flow in France in H1. Structural improvements in our free cash flow generation is a key focus for us, and we have taken a number of actions to that effect. This is evident in our H1 numbers, where cash flows from continuing operations, including lease payments, improved by 51%, driven both by an improvement in EBITDA and a reduction in non-recurring expenses. This is our key metric to monitor the structural improvements of our business. We expect this improvement to accelerate in H2, with the strong profitability of our banners combined with a return to sales growth and the end of our transformation plan translating into lower exceptionals. Free cash flow in H1 after CapEx and working capital was at EUR -346 million, in line with the usual seasonality. The difference versus 2020 of EUR -17 million in working capital relates mostly to Cdiscount, which recorded an exceptionally high level of net sales last year in Q2 2020, boosting its working capital compared to the usual seasonality. This is a temporary effect that should normalize in our full-year results. The rest of the French business delivered a good working capital performance compared to the strong 2020 basis, despite lower sales in Q2, thanks to tight inventory management and the recovery of fuel working capital. As for our CapEx, they are close to the level of last year as we keep controlling the level and concentrate our expansion in franchise. Overall, we are clearly committed to sustained free cash flow generation in France, driven by the good mix of our business, tight exceptionals, and constant inventory management. Page 28. This table sums up net debt variation over the semester. As usual, the variation in H1 reflects the seasonality of cash flows. Overall change in net debt, excluding IFRS 5 and disposals, stands at a level close to the one observed last year, with a EUR 17 million difference explained by the temporary gap in Cdiscount working capital just mentioned. Our cash financial expenses decreased by EUR 63 million as a result of our 2020 buybacks. One word about the non-cash variation number of EUR -458 million. It includes first a variation of EUR -149 million in the segregated account, offset by an equivalent positive flow recorded in other net financial investments. Second includes the negative cash flow of Leader Price, EUR -288 million in H1 2021. This number includes the usual seasonality of this business and the operational losses recorded before transferring the stores, which were aggravated by the pandemic. We expect the conversions to be over at the end of September. After that, all of this in discontinued activities will also be over. Page 29. Going through our bond maturities. Two main observations. First, our bond schedule, which used to be heavily concentrated, is now more normalized, with maturities spread between 2024 and 2027, following our buyback and our two successful refinancings. Our Term Loan B maturing 2024 was refinanced this year with a new 2025 maturity and an interest rate reduced by around 1/3, from 5.5% to 4%. We also issued a new unsecured bond maturing in April 2027, The second important observation is that our near-term maturities in 2022 and 2023 are now totally covered by the amounts available on our segregated account dedicated to debt repayment and the disposals already signed or secured. This means we have no affected debt maturity before January 2024, which is the maturity of our Quatrim secured bond. This bond will be callable as of next November. All of this, combined with the successful refinancing of our RCF, puts us in a very good position to deliver on our disposal plan in an efficient way. Page 30, a focus on our liquidity at the end of June. Total liquidity in France stood at EUR 2.6 billion as of June 2021, including EUR 2 billion undrawn credit lines available at any time and EUR 520 million of cash and cash equivalents. On top of that, our segregated accounts dedicated debt repayments announced to EUR 339 million. The bottom of the slide shows the credit lines as of June 2021. The average maturity was 2.2 years. It has been extended to 4.6 years, as is shown on the following slide, page 31. Our RCF now matures in July 2026 for EUR 1.8 billion. To that must be added the new Monop' RCF maturing in January 2026 for EUR 130 million. Our liquidity is therefore secure for the next five years. On top of the maturity extension, the new Monop' RCF has a lower cost of utilization, and the covenants computation have been reviewed to take into account the improvements of our financial structure and the business plan of GreenYellow. The new debt over EBITDA quarterly covenants are computed on the French perimeter, excluding GreenYellow, taking into account only the secured debt. As you will see on the next slide, this new computation leaves us with a very comfortable headroom. As mentioned earlier, the new 2026 Monop' RCF now includes CSR criteria in margin adjustments based on greenhouse gas emission reduction, share of responsible sales, and share of vegetable proteins. This is the first important step for us in sustainability-linked financing consistent with our longstanding CSR commitments. Page 32. This slide shows the headroom available with the new quarterly maintenance covenants. The secured debt to EBITDA ratio stands at 2.1 x at the end of June, comfortably below the 3.5 x limit, which gives us a EUR 359 million headroom in EBITDA. The other ratio, EBITDA over financial expenses, is also comfortably met with EUR 199 million headroom in EBITDA. Page 33. To conclude, let me sum up our outlook for H2, in line with our clear priorities. In H1, we continued the successful repositioning of all our formats in the challenging context of the pandemic. In H2, with a more normalized basis of comparison, we will take advantage of this positioning to grow profitably. First, we'll focus on growth in profitable formats via the expansion of the store base and the acceleration of e-commerce. We target 400 new stores in H2 in formats such as Franprix, Vival, Naturalia, Monop', and Casino Shop, mostly in franchise. This will bring the total opening to 750 this year, all in profitable and successful formats. We also target an acceleration of our e-commerce operations, thanks to our strategic partnerships with Ocado and Amazon, and specific solutions deployed in our dense network of urban and proximity stores. Second, our high-growth businesses, Cdiscount, relevanC, and GreenYellow, will continue their development. GreenYellow and Cdiscount have already communicated on their plans to finance an acceleration of their growth, which could include market operations. These companies are successful operators, taking advantage of secular mega trends, and they have a lot of opportunities to create significant value through the effective deployment of additional capital. Third, we will maintain an intense focus on cash flow generation with continued EBITDA growth and a sharp reduction in non-recurring expenses in H2. We will also maintain our discipline in CapEx, focusing our growth on convenience franchise formats and e-commerce. These priorities are clearly set and our plans are tightly monitored, so we have a lot of confidence going forward on our perspectives. Thank you for your attention. I'm now ready to take your questions. Ladies and gentlemen, if you have a question, you have to press zero one on your telephone keypad, zero and one on your telephone keypad. Your first question from Arnaud Joly from Société Générale. Arnaud, go ahead. Yes, good morning, David. I have two questions, please. The first one, do you see scope to [audio distortion] costs in France as of the second half of this year? If yes, in which fields? The second question, do you see a risk of a fundamental decline in the food retail market in Paris with the potential decrease in the number of inhabitants and the home working? Have you already seen any negative impact from the development of the Drive piéton, in particular for your convenience format in Paris? Thank you. Thank you, Arnaud. On cost reductions, of course, we always work on cost reductions and optimization. We still have some opportunities for reducing costs by making good use of our technological tools. We've done most of the work, of course, in the stores already with the development of automatic cashier and self-scanning. We are deploying artificial intelligence tools that will allow us to optimize costs further in the back office, and we are also continuing to move on the integration of back-office costs between Franprix and Monoprix. We still have some further opportunities to reduce costs. Of course, we've done a lot of work already. Now, when we see the H2, we will come from the high level of profitability already reached. On that level, we think we can add a lot more, thanks to higher growth. On the fundamental decline of Paris, that's not the way we see things. There has been, in the recent past, yes, a decline of people in Paris, but these are people that temporarily moved during the remote working period. The inhabitants are still there. If we look at the recent trends, we're actually seeing a very strong uplift to Franprix and Monoprix in the recent weeks. We see Paris fundamentally as a very sound market. Our positions there are very good, and as I mentioned, we did far better than markets. The fact that there might be openings of [audio distortion] or things like that does not seem to affect our offer. Importantly for Monoprix, and especially for Franprix, we expect a lot of growth to come, not just from Paris, but from the outskirts of Paris, the suburbs of Paris, which are growing in population clearly. In the past years, it's been clear that the growth of population is not from Paris, it's from the outskirts of Paris. There, as mentioned, we have 150 openings planned for Franprix, mostly in this area. We think we have models that are particularly adapted to these areas. Actually, we see a lot of room to grow in the outskirts of Paris and a return to normal that should happen in next few quarters in the inside of Paris. Next question? Thank you. Next question from [audio distortion] Bank of America. Please go ahead. Yes, good morning. Just one question actually. Two, sorry. The first one, just looking at the French free cash flow. We have not seen many improvements actually in the first half, I understand that there is some working capital impact, what are you expecting going forward and into H2? Should we see a significant acceleration or are you still a bit cautious there? The second question is looking at your like-for-like in France overall. I understand that you had tougher comps, what should we expect heading into H2? Is there a point of time where you believe you can start to develop positive like-for-like again in France? Thank you. On free cash flow, the first structural improvement comes from EBITDA improvement and exceptional cost reduction. That is already obvious in H1 with a 50% improvement, and we expect this to continue on a higher basis. Higher numbers, presumably in H2, both higher EBITDA and reduction of exceptionals. The other items, CapEx, they should be stable over the year, as we mentioned. From one year to the other, the difference between H1 and H2 may change a bit. Over the year, we clearly said that we intended to control the CapEx, at least below last year's number. That is clear. For working capital, there are always variations between H1 and H2. As mentioned, we actually did very well. We could have expected a lower variation of working capital compared to last year on the French retail business because last year was exceptionally well, exceptionally good with very good Q2 numbers. We managed to recover some of the fuel sales, so that compensated the fact that sales were a bit lower in Q2 this year. We expect, of course, we continue to work on tight inventory management, and we expect to have a contribution of working capital. The key improvement, as stated at the beginning of the year, and we confirm that, is we expect a strong contribution to cash flows thanks to EBITDA growth and reduction of exceptionals. That proves a sustainable cash flow generation. In like-for-like, yes, we are already seeing actually positive like-for-like on proximity actually in the last four weeks. Monoprix is already close to zero. It's -1 the last four weeks, so if it keeps on improving, it should get positive at some point in the very near future. The other banners are also improving, so it's difficult to pinpoint the exact time when each of the banners is going to turn positive, but that is clearly the goal. If you look at the recent market share data, it's interesting to see that we were losing a lot of market share two months ago. We were losing -2 points, and it has been increasing period after period. The last period, we are actually in line with the markets. We're not losing market share anymore in most of our formats, except the hypermarket. I think this is clearly the goal, and, I can confirm that to you. Thank you. Next question from Andrew Quinn from Exane BNP Paribas. Please go ahead. Yeah. Good morning, David. I'm going to be cheeky and go for three. Just on that trading, is it possible to give us a two-year stack or just an indication of whether or not the two-year stack of like-for-like in France has improved? Second question would be, is there an earnings impact from any of the recently announced disposals? Just to help us with modeling maybe in the second half and next year. The final one, just on the discontinued operations, obviously Leader Price made a loss, and you're partly responsible for that. Could you pull that out for the first half, and also maybe the expected impact in the second half? Thank you very much. On like-for-likes, what we're monitoring right now is, of course, the reason why the like-for-likes were so negative in Q1 and Q2 is because we had a very high basis of comparison. As the basis of comparison normalize, the like-for-like is up. That's what we're monitoring, and that's why we're looking at the four-week data on a one-year basis because that's going to happen. When we move period to period, the basis from last year normalizes, and that brings our like-for-like compared to last year up, and that's what we're looking at. Plus the fact that the restrictions are lifted. The target is to get to these positive like-for-like by banner progressively, each of them getting to positive, and this is already the case at Proximities. That's the way we're looking at it now. It's really on a one-year basis. It was clearly negative in Q2. It's getting less negative and even positive in some of the banner in the last four weeks. It should get positive with the normalization of last year, plus the impact of the lifting of the restrictions. The impact of the disposals on our results, of course, the EUR 100 million that we're getting, which will be actually, I think, a higher number. EUR 100 million is the bare minimum that we have to report given what's been done. This is a pure -plus. It doesn't entail any additional rents since we're already paying the rents. The real estate has been sold in 2019 to the JVs. The JVs resell these to final buyers, and we get the earn-out. We're already paying the rent, so it's a real estate disposal that does not bring any additional rent. It's a pure positive impact, and it will, of course, allow us to decrease our financial cost. As for the disposal of FLOA, it doesn't impact our EBITDA. FLOA is not recorded in our EBITDA. It had a small net result impact, of course, this will be more than offset by the reduction in financial costs with the cash-in that we have here. As for discontinued operation, we mentioned there was a Leader Price discontinued losses in H1. These were due to the stores that we are keeping to operating just until we transfer the stores to Aldi. As mentioned, this is mostly over. We had 400 stores already transferred at the end of July. That means less than 200 still remaining. We expect by the end of September that all the stores will be transferred. After that, there'll be no more losses. We have just to finalize the restructuring of this. We lost EUR 218 million in H1, as mentioned, on Leader Price, and the number in H2, it should be much lower, of course, because instead of having 400 stores in six months, it will be 200 stores in three months. If you want to make a calculation, you can divide the number of H1 by four and you would get, I think, a reasonable estimate of what that could cost us in H2. After that, it's over. Thank you. Next question from Maria-Laura Adurno from Morgan Stanley. Please go ahead. Hello. Thank you very much for taking the question. This is Maria-Laura Adurno. Just two on my side. The first one, would you be able to provide us the COVID cost that you incurred for the first half of this year and how it compared versus last year? The second question, your net financial charge has come down for France. Just wondering what's the main driver behind this? Thank you. Maria, COVID cost, I've mentioned since Q3 last year, we only have about EUR 5 million per quarter of COVID cost, so that's EUR 10 million for H1. Of course, much lower than last year. Basically, it's about EUR 120 million less than last year. That, of course, contributed to offset the decline in sales. If you look at the growth of our total EBIT, it's basically the loss of sales was compensated by the loss of the reduction in COVID cost and the improvement in EBIT is explained by our cost reduction. We can basically model it that way. Financial charges. The financial expenses as recorded in net results increased, I mentioned it's mostly non-cash, one-off impact links to the refinancing of the Term Loan. When we refinanced our term loan in April, we accelerated the amortization of the cost of the original term loan of 2019 and recognized a non-recurring expense of EUR 40 million. That's included in our net result. It's detailed on page 37 of the presentation. This is mostly a non-cash element. The cash was already spent in 2019. The recurring impact of this refinancing is actually a saving of EUR 9 million per year. That means next year, in H1 next year, we will have an improvement of basically EUR 50 million of financial cost compared to H1 of 2021. It's of course, a very satisfactory refinancing for us since we managed both to extend the maturity and to reduce the cost of this term loan. Thank you. Next question. Thank you. Next question from Clément Genelot from Bryan, Garnier. Please go ahead. Hi. I will have three on my side. The first one is on the covenant. Why did you ask the banks to... To adjust the calculation [audio distortion]. Of course, I understand about GreenYellow need some fresh money to, let's say, carry out its plan. You could have just exclude GreenYellow from the calculation. My second question is whether on net debt. Why is the net debt in France at a discount almost stable in Q2 year-on-year? When we look at the numbers given in the covenants pages, so that's gross debt minus cash. That's almost stable in Q2 year-on-year, while in the same time, you did almost EUR 700 million of disposals over the last 12 months. My final question is whether on future asset sales in France. Have you received any expressions of interest in other assets in France? Of course, that's other assets than just GreenYellow and Cnova. Thanks. Thanks, Clément. The covenant adjustment first, it was, of course, necessary to exclude GreenYellow from the computation of the covenants, since GreenYellow will raise debt to fund its growth. They have a very ambitious plan, and they've communicated the plan, which is to invest EUR 1.9 billion in the next five years, part of which will be financed by their, of course, operational cash flows, part of it probably by new equity, and the rest by debt raised at GreenYellow level. We had to exclude GreenYellow. When we got to the covenant computation, we discussed it with the banks, and we looked at the situation of the group. The assessment was that there was basically no more debt, as you know, in the next two years. The protection that the banks needed was a protection for the security of the debt, not the protection on the overall leverage, which is well under control. The right way to look at it was to look at secured leverage. For the banks, it's important since they have a secured RCF, to ensure that we do not raise additional secured debt instead of unsecured debt. To replace unsecured debt, I mean. We moved to secured debt covenants. They made much more sense. We still have the old covenants. This has not moved on our dividend restrictions. It's still the same. It's a 3.5 gross leverage, total gross debt over EBITDA. That hasn't moved in all our instruments, so that protects all the lenders. We cannot add more debt to pay dividend, for instance. It's impossible. We also have restrictions on the debt that we can raise. Basically, we can raise debt only mostly to repay existing debt, the lenders are very well protected. They saw that the situation had much improved compared to 2019, when we put in place the first covenant financing. It makes a lot of sense to move to the new covenants. Of course, the consequence of the secured covenants that we have much more leeway, much more headroom. That gives us, of course, total flexibility in realizing the disposal plan in a very efficient way. When you discuss with a buyer, it's much better not to be pressed by an immediate liquidity issue or immediate covenant issue, and you can do it in a very confident way, and you can get much better terms with the buyers. That's what we do. That does not mean we will slow down the disposal plan, of course. We are still very committed to do it as fast as we can and reduce the debt to reduce our financial cost. When we discuss with the buyers, we are clearly under no pressure, and that gives us a big advantage, and the banks can see that as well. As for debt, as explained, when we look at H1 versus H1, if you look at the net debt in France, I think that's the simplest way to look at it, excluding GreenYellow, and just on the French perimeter, we came back EUR 200 million, the net debt, [audio distortion], from June to next June. You're right, that is less than, of course, the total disposals that we realized during this period. Our goal is to cover our financial costs with our operational cash flows, our recurring operational cash flows, and that's mostly what we did in the last 12 months. There may be a little gap due to the working capital of Cdiscount, as mentioned. Mostly, we're already there, very close to that point at the end of last year, and that's mostly where we are at the end of June. There are other things that comes below the cash flow, and that explains that the reduction of [EUR 200 million] is not equal to the total disposals that we made. A big part of it, of course, is still the cash burn from Leader Price during that period that we still had to bear. I mentioned the EUR 288 million in H1 this year. There was, of course, a number last year that's probably close to that as well. That puts us around 400 over 12 months. To that, you have to add EUR 70 million last year of unwinding of the [TRS] facilities. We mentioned that, of course, at the time, and that was a net cash cost. Basically, that explains the decrease of the cash position between last year and this year, and the gap between two periods in terms of debt. What matters is that now all these losses are behind us. As I mentioned, Leader Price is mostly over. By September, all the stores will have been transferred, so this is done. There is no [TRS] to unwind or anything like that. When we're looking forward, we see cash flows covering our interest costs with a margin that will grow over time and should allow us to deliver organic leverage and generate net cash flows after financial costs. That's the way we see things. French asset sales, do we have any expression of interest? Yes, we do. Of course, we don't give any details on any discussions that we have or any incoming calls that we get, but I can confirm that we get incoming calls. We get discussions, and we will not communicate on anything before we have a deal signed with someone. That's always what we've done. The recent deals that we announced, we had not communicated on them before. We had just said that at the beginning of the year, we were in the discussions. There were processes going on. We think we did a very good deal with FLOA, because we sold our 50% stake above the equity value significantly. We got EUR 50 million more as part of our new agreement with BNP, and we get a 30% earn-out on the value creation of fraction payment, which is a booming market. You've probably heard of Klarna and companies like that. We think with BNP there's a clear potential to do something very effective there and get 30% of the value created by 2025, so without having to invest in more. It's a very good deal. The next deals that we'll get, we think will be good deals as well. Of course, I'm not going to disclose who calls us for what assets. I can confirm that, yes, there is interest for our assets clearly, and not just the assets that you mentioned. Thank you. Next question from Nicolas Champ, Barclays. Please go ahead. Good morning. Thanks for taking my questions. I have three. First one is, I would like to come back on the working capital outflow in France on page 27. I think you mainly explained it by Cdiscount's negative contribution. Could you be more precise and quantify the impact of the working capital flow for Cdiscount so that we can compute the working capital variation for French retail activity only? On page 36, also, I would like to come back on the significant swing regarding the one-off charges. It seems it basically stemmed from the significant shift regarding the disposal plan. There was a EUR 101 million charge in H1 last year that moved into a EUR 151 profit this year. Could you elaborate on this item, on this significant swing? The last question, I will make another try regarding asset disposal. You had roughly EUR 800 million, EUR 797 million to be precise, million of assets classified under IFRS 5 for French Retail division. Could you elaborate a bit on the nature of these assets? Are we talking about real estate assets, or are we talking about stores? Hypermarkets or supermarkets are included in these numbers. Could you elaborate a bit on this big number of EUR 800 million? Thank you. Yes. Working capital, actually, Cdiscount has published their net results and their cash flows recently. If you look at their press release, you will find the cash flows in H1 compared to last year, and you will find that they have a change in working capital compared to last year that's roughly the same amount, that variation that you see in the table on the French, including Cdiscount. By memory, I think it's about EUR 60 million gap, EUR 70 million gap. Actually. If you look at their press release, last year, they were EUR -114 million, and this year, EUR -183 million. It's exactly actually EUR 70 million gap. If you correct from that, you will see that France was basically the same as last year. On page 36, the approach. Yeah. Last year, we had some depreciations of some assets. This year in H1, what we did is first we recorded as exceptional benefits the EUR 100 million from JV Apollo and Fortress. It's a profit. In accounting, it's an exceptional profit. It's recognized because the fact the disposals that have already been made, make this payment certain. We don't have the cash yet, but it already can be considered as an asset already. We recognize this in the accounts as a profit, exceptional profit. Apart from that, we have some provisions that we took last year, and we took these provisions back because of the revaluation of some of our assets. This is linked with the last questions that you asked. Of course, that point is that last year we had some depreciations, and this year we don't have any. There's the both. The last year, there was a negative, and this year there's a positive. On asset disposals, in the EUR 800, of course, there's FLOA. FLOA was in IFRS 5. It will get out of the IFRS 5 when the deals close in a few quarters. There are other assets, unfortunately, as usual, I can't disclose much more because we don't disclose what's there. I can take the number of different things. We have non-core assets that we can sell. FLOA was part of them. There are others. And we can have some real estate, we can have some specific assets. I'm sorry I can't tell you more because we don't disclose more than the assets that we announce when we sell them. We announced FLOA. We didn't say that FLOA was in IFRS 5 before we announced the sale, but it's now, of course, clear, and it's good in the detailed accounts. That's the only asset that we reveal is in the IFRS 5 because we don't want to show our hands to the buyers. Of course, these are all assets where there is an ongoing process, and we expect these to be sold. Importantly, these assets that are in IFRS 5 should not impact significantly our EBITDA when we sell them, either because as FLOA, they are not recorded as EBITDA at all, or because they are not big contributors to our EBITDA. That's what I can tell you. Next question. Thank you. Last question from Robert Joyce from Goldman Sachs. Please go ahead. Hi. Good morning, David. Thanks for taking the questions. I'll go with three as well. Just on the French free cash flow, a few definitional changes. I'm just wondering if you could give us the equivalent 12-month number for the [EUR 346 million] you give for the French free cash flow. Just saying you said your goal on a 12-month basis was to cover the finance cost. If you have that number handy on 12 months, that'd be helpful for us. Second one, I think I understand Leader Price. It's a little different to how I understood it. Am I right in saying you had a sale price of EUR 648? You said you covered cash outs of around EUR 400 to date, expect another [EUR 70 million] out, and I think you bought back shares, stock, stores around EUR 55 million. Is that the way to think about it? It almost leaves you with a sort of flat sale price. The third one is just on the asset disposals. Just to cover the gap between, I think, what we've guided to [EUR 4.5 billion], hoping [EUR 4.5 billion], and you've got about [EUR 800 million] in the IFRS 5. Is there anything you can say on the assets beyond that? Would you look outside of France for asset disposals? Thank you. Thanks, Robert. Last 12 months, I don't have the numbers right here, but I think you can compute that actually, when you look at We published the net debt last year, end of June, and this year, end of June, and you have the financial cost. From that, I think you can basically compute them. Of course, we will give the numbers at the end of the year, but I can say that you can compare to last year. The gap is mostly on working capital, as I said, and the improvement is on the recurring cash flows. We expect that this year that on this parameter, we will improve. Last year, if that's your question, GreenYellow is now a contributor to our cash flows. In the end, when we look at France retail, including GreenYellow, the goal is, of course, France retail, excluding GreenYellow and including Cdiscount, which should have a small positive cash flow generation as it did last year. The goal is to cover the financial cost. That's the clear goal, and we think that the H1 number actually confirms that we are on the right track for the goal since we increased the operational cash flow by 50%. This is a key driver because in the long term, we know we can target working capital slightly positive if we grow the sale. We've done most of the work in inventory management. We can do some more. We have some more inventory reduction to do. If we look in the mid to long term, we need high operational cash flows to have net cash flows after financial cost positive. That's clearly the goal. We think the numbers that we show here are perfectly consistent with that. On Leader Price, I think your question related to basically what I already said, but yes, we had some cash burn on Leader Price because as part of the deal, of course, we still had to bear the cost of operating the stores before they were transferred to Aldi, and they are transferred by batch so that Aldi can convert these stores to Aldi stores. By next September, it will be over, so there will be no more stores to operate and no more Leader Price structure. We just keep the franchise business, but this franchise business is profitable, so there's no issue with that. Of course, we had to buy back the franchisee, but that's been done last year, so it's not an issue anymore. To sum it up, I would say that we're very happy to have sold Leader Price at a net EUR 600 million, EUR 650 million, which we got from Aldi and [EUR 50 million] that we had to pay the franchisee to do that because it's a business that was clearly bleeding cash more and more. If we hadn't sold it, we would have had to close this business, and that cost us a lot. Getting EUR 600 million for this business was really an excellent deal. Now, from now on, looking forward, of course, there will be no more source of cash burn below our operational recurring cash flows, which means that if we reach the target, which is clearly to cover the financial costs, we'll be able to deleverage organically. As for asset disposals, we have EUR 3.1 billion. We have EUR 800 million under IFRS 5. Other assets are not in IFRS 5, which means that there is no ground to date to classify them as such. To classify an asset as an IFRS 5, you need to have an ongoing clear process to dispose of them. We have a number of opportunities, as mentioned in our full-year results, we have additional flexibility now to realize the EUR 4.5 billion because we have a number of valuable assets in France, and in some of these assets, we have flexibility. We can keep control of some of the assets while monetizing part of them. That is a possibility, for instance. To be clear, the EUR 4.5 billion does not include anything outside of France. The goal is to reach the EUR 4.5 billion by selling assets or parts of assets in France. We are fully confident that the value of our Latin American assets, which we plan to keep. They are not part of the EUR 4.5 billion, to be very clear again. I think this concludes the discussion. If there is no more question, I will wish all of you a happy vacation for those who take some in August. Thank you for your attention. Thank you, ladies and gentlemen. This concludes today's conference call. Thank you all for your participation. You may now disconnect your lines.