AUTO1 Group SE (AG1) Earnings Call Transcript & Summary

Hello, good afternoon, and good morning and good evening to international participants. I'm Philip Reicherstorfer, Group Treasurer at AUTO1 Group. Welcome to our first AUTO1 Group Capital Markets event. We will start as always with the presentation and then have the opportunity for a question-and-answer session. The presentation today should take about an hour, and we have scheduled a similar time for Q&A. So hopefully, we'll be wrapping up in about 2 hours from now. [Operator Instructions] Before I hand over, I must make you aware of the safe harbor provisions at the beginning of this presentation. These will apply to any forward-looking statements made by management during this call today. Today, Christian Bertermann, our Co-Founder and CEO of AUTO1 Group; and Christian Wallentin, our CFO, will provide you with a deep understanding of our retail and merchant segments. It has been 5 years since the IPO and our business has changed materially. We think now is the right moment to lay out historic financial segments and long-term targets for both our merchant and retail businesses, and we're excited that you're joining us today. With that, over to you Christian.

Hi, everyone. Thank you, Philip. Welcome to this event. AUTO1 Group is Europe's leading vertically integrated digital automotive platform for buying, selling and financing used cars. Together with Hakan Koç, we founded AUTO1 in Berlin in 2012. Since then, we have traded more than 6 million cars across 30-plus companies, generating EUR 8.2 billion of revenue in 2025. Today, we're going to walk you through the AUTO1 model and what makes it structurally different from any other player in Europe. We'll show you how our 2 segments: merchant and retail interact and compound one another, each making the other stronger over time. We will lay out the long-term financial targets for both segments and explain specific drivers for unit and margin progression on the back of the immense opportunity ahead of us. We're building the best way to buy, sell and finance cars. Our mission shapes everything we do. We didn't grow up in the old school used car business. We have backgrounds in tech and see a very large, very fragmented and deeply inefficient market that we can change for the better. Since day 1, our approach is to systematize and digitize the industry at scale. That is why we built our C2B buying business. Europe's largest consumer purchasing network and our sourcing engine to make car selling fast, transparent and fair. It is why we build auto1.com, our merchant platform and the backbone for B2B buying across Europe. And it's why we launched Autohero in 2020, the first trusted only digital car buying experience in Europe. On top of that, we have built market-leading integrated financing products for dealers and consumers alike because financing is such an integral part of this market. The sequence of how we built the company matters enormously. We started with supply, building a consumer sourcing engine first in 2012. One year later, we launched Automon.com, bringing inventory to the dealer side and starting to accumulate the pricing data that sits at the core of our AI pricing models today. Only in 2020, after 8 years 8 of building the foundation, we launched Autohero. We invested more than a decade into becoming the European leader in digital car sales, AI-powered pricing, and European logistics, physical car infrastructure and innovative power financing products. The results speak for themselves. From 230 cars in 2012, we grew to more than 842,000 cars traded last year and a record 249,000 units in Q1 of this year alone. Group gross profit per unit has increased steadily since 2012 and is on a very remarkable track since 2018, increasing every single year. We crossed EUR 100 million adjusted group EBITDA for the first time in '24 and almost reached EUR 200 million in '25. We are now at the point where we are no longer just building, we are leveraging. We drive volumes and profitability simultaneously. AUTO1 Group is one of the most exciting investment opportunities in Europe today. First, we operate in one of the biggest markets of the world. used cars is a massive, highly fragmented industry, and the vast majority of customers are unhappy with the current buying and selling experience. Second, our superior vertically integrated business model that is clearly apart from the traditional brick-and-mortar approach in used cut trading, generating superior customer experiences, growth and profitability at scale. Third, the outstanding customer experience we deliver is winning market share across every customer group we serve. Fourth, the pan-European infrastructure that took a decade to build and cannot be replicated is our structural moat. We built in a very scalable way. and we have a proven expansion strategy for every part of our business. And fifth, underlying our business is a very robust financial model based on the advantages of vertical integration, it delivers market-leading profitability at scale. Let us walk you to each of these now. Over to you, Christian.

Thank you, and a warm welcome from my side as well. Let's start with a deep dive into the European used car market and the massive opportunity ahead of us. Europe's used car market is one of the biggest consumer spending categories globally for most European families, buying a car is the largest or second largest financial decision they will ever make, depending on if they choose to own their home. The used car market is almost twice the size of apparel and 4x the size of the electronics market. On a value basis, that is EUR 700 billion in annual transaction volume with EUR 100 billion in financing. The EUR 100 billion financing pool deserves a particular mention. Every used car transaction is also financing opportunity and the economics of embedded car financing are compelling. At slightly above 3% market share, we are already the largest player in Europe with 97% of this market still ahead of us. What makes this market so extraordinarily attractive is the combination of size, fragmentation, stability and its unhappy customers. There are over 250,000 dealers in Europe and the top 20 on less than 6% of the market. The main reason is that the traditional brick-and-mortar approach in used car trading as scale limitations. On top of that, Europe is not 1 market. We have different languages, tax rules, registration regimes, transport networks and very local demand patterns that form barriers to easily scale Europe-wide. European used car volumes are very stable at around 27 million to 28 million transactions per year and have a long-term CAGR of around 2%. The underlying general need for mobility is what drives the size and stability of this market in the long run, a market that is built into the fabric of everyday European life. Yet, it remains one of the least consolidated lease digitized consumer markets anywhere. And then maybe most importantly, almost 4 out of 5 consumers do not enjoy the car buying experience. They dislike the haggling, the opacity, the lack of trust and the physical car buying journey, which can be exhausting and frustrating at the same time. Now I'm handing over to Christian to walk you through our superior vertically integrated business model.

Thank you, Christian. And let me first apologize for the slide. The truth behind that small front is that the advantages of vertical integration are creating numerous benefits for our customers, so they barely fit on one page. We source directly from consumers across 9 European markets through a seamless digital evaluation and funnel, combined with our dense network of drop-off branches. As a result, we control our entire supply without being dependent on auctions or third-party wholesalers. With our unique approach to carbuying we are creating a superb selection of cars for our buying customers while offering market-leading prices to our selling customers through our European demand generation engine. On pricing, our AI pricing engine is built on over 6 million actual realized transactions accumulated over 14 years. We generate a competitive price for selling within seconds and our matchmaking technology connects supply to demand across the entire EU efficiently. This is how we are realizing best prices for our buyers and sellers a huge competitive advantage. On logistics, our bespoke logistics network with more than 170 logistics hubs, more than 300 logistics partners and a dedicated last-mile delivery fleet for Autohero is the largest of its kind for cars than our on. Thanks to it, dealers can receive their fresh purchases incredibly fast, both nationally and cross-border, including all paperwork being handled professionally. At the same time, the network allows us to offer very fast, convenient and reliable delivery times for our retail customers. On our physical infrastructure. Our physical infrastructure across Europe is a moat 14 years in the making. Today, we operate 12 in-house production centers with a combined refurbishment capacity of over 248,000 vehicles. Every car that goes through the Autohero refurbishment process, befits the same high-quality standard, adding to the high trust the Autohero brand stands for today. Together with more than 750 drop-off and more than 150 pickup points, our physical infrastructure is a unique asset that enables more than 70% of the European population to reach a drop-off point within 15 minutes drive from their home. Our trusted brands are a key component of our vertically integrated model representing the better, superior way to buy, sell or finance a car. They decrease friction and offer customers peace of mind. Our consumer selling brands in our dealer brand, auto1.com, are based on 14 years trusted transactions with private consumers and dealers alike. Autohero is the fastest-growing European auto retail brand with already 35% aided brand awareness. We believe that building an unparalleled brand experience no matter where our customers can get in contact with our brands is a massive future demand driver. On financing, our in-house merchant financing provides dealers with the capital they need to grow their business across 8 markets. Consumer financing is embedded directly in auto hero purchase flow and can be completed easily and stress-free in under 5 minutes. Both are funded to our own ABS programs at institutional scale, which means we can pass competitive rates through to our customers internal driving conversion on finance transactions. All of these different layers of our vertically integrated model are structural advantages compared to the traditional brick-and-mortar approach in used car trading. They directly translate into increased value for our customers. We have spent more than a decade to develop our business model and our unique operating approach. Vertical integration gives us control over price beat and customer experience or in other words, every aspect of the transaction in a way no one else can match. Christian will now explain our products in merchant and retail in detail, and how they are contributing to our exceptional customer experience.

Thank you, Christian. Let me take you through each of our segments and showcase to how we create value for our customers. We are operating 2 segments on 1 integrated platform. Autohero, our consumer growth engine, changing the way people buy cars. Carefully selected vehicles from our consumer sourcing business, go through 1 of our 12 large industrialized production centers for inspection and reconditioning. They are then listed online on Autohero, extremely convenient to check out and available with integrated financing and the 21-day return policy. We're delivering to the buyers store and/or making the purchase available for nearby pickup quickly. It's an e-commerce-like experience applied to one of the largest purchases they will ever make their new car. The Autohero growth trajectory tells the success of our value proposition. Since launching in 2020, we have grown almost 60% compound annual growth rate. In parallel, GPU has grown from EUR 362 in 2021 to EUR 2,605 in 2025. These great results are based on a superior value proposition, our operational discipline and excellent execution from our teams. On the merchant side, we have built a pan-European wholesale market leader for used cars. Resource vehicles directly to consumers across 9 countries and sell them to 1 of our more than 54,000 buying dealers spread across more than 30 countries. At the core of the business, it's a unique matchmaking engine that, for instance, matches the supply of Volvo in the Nordics with demand in Spain or sourcing Beneau in Germany and selling it to France. No other used car player does this at our scale. We continue to grow strongly with a 12-year CAGR of roughly 50% since 2013. And our unit economics continue to improve steadily with merchant GPU growing to EUR 976 in 2025. Autohero, our retail segment launched in 2020. In 5 years, we have grown to over 100,000 units delivered across 9 markets, making it the fastest-growing used car retail brand in Europe. Our MPS stands at 69%. That is world-class core for any consumer product, let alone for car sales. Aided brand awareness reached 35% in Q1 2026, up 9 percentage points year-on-year. Our ambition is simple, to make Autohero the go-to brand when buying a used car. Brand drives organic acquisition, organic acquisition lowers cost per unit, lower cost per unit increases contribution margin. That is the compounding flywheel. The skills we have built to get here are not easily to replicate. We learned how to source at scale, price at scale, refurbish at scale and deliver at scale continuously improving our unit economics on the way. These are the capabilities that will power Autohero to category leadership across Europe. We create value for our retail customers in multiple ways. We offer a vast selection of cars listed across Europe at any given time. Every single one has gone through the same rigorous standardized refurbishment process at one of our own production centers. Cars come with a detailed uniform condition profile AI-powered damage detection and 12-month standard warranty. Our scale is structural pricing advantage that compounds further with transactions. The more cars we process to lower our refurbishment and logistics cost per unit and the more competitive our pricing becomes. The same logic applies to our seamlessly integrated financing. The bigger we get, the more competitive our financing rates can become. Additionally, AI pricing gets smarter with more data. In retail, there's still plenty of potential for improving price precision by a bigger data set. Alto hero offers maximum convenience. The entire purchase journey is completed online or on our app without a single phone call if you don't want to make one. Financing is approved in minutes. We deliver to your door within 10 days in 1 of our iconic glass trucks or the cars available pickup at numerous locations across Europe, often as quickly as within 48 hours. You enjoy a 21-day full money back, no question asked to return policy in case you change your mind. Altogether, this is the best way of buying a car. Auto1.com is Europe's #1 wholesale platform for used cars. Over 54,000 unique buying dealers, 30-plus countries, 50,000 cars available at any given time. These are the results of 14 years of continuous investments into supply and demand, platform technology, AI pricing and physical infrastructure. Auto1.com is effectively the clearinghouse of the European used car markets. As mentioned before, Europe is not a single car market, but over 30 national markets with different price levels, tax regimes, registration systems and consumer preferences. We have reduced this complexity to a few simple clicks for our dealer partners. With us, they can tap into any country supply. Cars are sold to the dealer in Europe who values it the most with paperwork, logistics, payment and financing handled by us. Every car flows to its highest value use across the continent. The value that we offer our dealer partners is based on 3 key pillars. We offer the largest EU-wide selection at great prices. Around 50,000 cars listed daily across 30-plus markets, consistently quality graded an AI price in real time, dealers buy at market value and with full transparency. Prices are adjusted to market conditions in real time based on more than 6 million realized transactions historically. As part of the fully digital end-to-end experience, auction bidding payment, transport and documentation is all managed centrally in our platform. Dealers can focus on their customers. We handle everything else. Our AI-powered search and recommendation tools help them to find the right case in no time. Through our in-house merchant financing product, dealer's get instant working capital to buy more cars without tying up their own cash. It is the only fully integrated sourcing and finding solution in Europe, and it is a key driver of frequency and loyalty on our platform. Over to Christian, who will walk us through how we built the infrastructure to scale and what our expansion playbooks look like in practice?

Thank you. We operate the largest European vehicle drop off and delivery network seamlessly connected to the biggest logistics infrastructure for cars. With a weight of 1 to 2 metric tons, our goods require a unique logistics chain. The physical network that we own and operate forms a very strong moat and is one of our many structural competitive advantages. We operate more than 750 branches across Europe, where selling customers drop out their car and papers. 82 of these locations now serve additionally as Autohero pickup points. where retail buyers can collect their Autohero car. This is the by Autohero co-branding model, which we believe is a powerful first step of bringing our 2 consumer brands closer together targeting a seamless customer experience. 77 pickup locations are operated exclusively by Autohero, mostly in areas where the retail delivery volume is already very high. Additionally, we operate 12 large production centers that currently run at around 50% utilization. Together, these form are incredibly strong physical mode more than 10 years in the making. We have a proven playbook for scaling our business. While the drivers are different for scaling supply, merchant or consumer demand we possess unique knowledge on how to invest, apply and monitor each driver. These playbooks were developed over years and were optimized with every success and failure we went through. The same also applies to our infrastructure. We successfully scaled our logistics network, a number of branches, our production centers, our retail delivery infrastructure in line with the needs of the business and improving unit economics. Every driver and market has a lean rollout playbook that is executed disciplined and precisely when needed. Our strong organizational knowledge for scaling input trials enables us to do both at once, building quickly, while we operate with increasing leverage at scale. Now we come to the heart of today's event, our long-term segment targets and the drivers behind them. For the first time, we're disclosing full historic segment financials for both merchant and retail and will explain unit and profit drivers within each segment P&L in detail. Let us start with merchant. Merchant is the wholesale market leader, our cash generation engine and as we will showcase in the next minutes, a business with a very large growth and profitability runway ahead. Across the last decade, 3 trends make up the development of the merchant business. First, strong unit growth. units almost multiplied 25x since 2014 from 29,000 to more than 740,000 cars sold in 2025. The business grew every single year with the exception of COVID and the strong profitability push in 2023. Over the years, it was tested by very different macro environments and has emerged from each 1 stronger and more profitable. Second, structurally rising profitability per car. Gross profit per unit grew from EUR 749 in '21 to EUR 976 in '25. Total gross profit reached EUR 723 million in '25, up from EUR 416 million in '21. Third, a substantial increase of adjusted EBITDA per unit sold, EUR 113 in '21, EUR 257 in '24 and EUR 320 in 25 million. The profit of every car we trade has almost tripled in 2 years. Let me walk you through 2025 totals on the next slide. Revenue was EUR 6.4 billion in '25, up from EUR 4.2 billion in '21. Gross profit was EUR 723 million. SG&A was EUR 484 million, of which EUR 104 million was marketing, EUR 269 million operations and EUR 110 million overhead. That results in adjusted EBITDA of EUR 239 million for the Merchant segment, a 3.7% margin up from EUR 158 million and 3.1% margin in '24. This is an absolute profitability improvement of 51% year-on-year. The merchant business has a considerable amount of fixed cost leverage visible on the overhead line, with total overhead just growing roughly 3% per annum over the full period from '21 Marketing is getting similarly efficient at scale, a testament of the strong brands and consumer trust we have built within merchant. Let me now translate the historic merchant financials into per unit terms and show you the massive long-term opportunity. Since 2022, we have increased GPU every single year, arriving at a level for gross profit per unit of EUR 976 for 2025. In the long term, we are expecting merchant GPU to reach levels between 1,080 and EUR 1,200 per unit. On marketing, we have seen a continuous downward trend over the years, with marketing hitting a level of EUR 141 per unit in '25. In the long term, we are expecting marketing per unit to reach levels between EUR 140 million and EUR 110. On SG&A in total, we have observed EUR 653 in '25 . We expect SG&A to decrease to between EUR 600 and EUR 480 in the long term. And consequently, we're expecting adjusted EBITDA per unit for merchant of EUR 480 million to EUR 720 for the long term, a strong future upside and a continuation of the adjusted EBITDA per unit ramp over the last few years. We are combining these long-term targets for merchant unit economics with a growth rate corridor of 10% to 15% for [indiscernible]. We believe that this corridor represents a prudent target that we feel comfortable achieving. Our goal, of course, is to be on the top end of that range and ideally outperform over time. Let's now go into the details. Start with the units and lay out the total addressable market for merchant. We will first go through the TAM logic for demand, and then we'll lay out an equally simple approach on the supply side. We are categorizing the 250,000 dealer pool of Europe into 5 categories: enterprise, large, medium, small and local dealerships. Based on our own data and models, we roughly know the share of cars that each type of dealer is looking to buy externally as internal sourcing is typically limited. The relevant amounts are indicated in the table column external sourcing demand. In total, these amounts aggregate to 10 million units per annum. This is the amount that based on our modeling, the total dealer base of Europe is looking to buy from other sources than their own. The current market share that we reached within our -- with our 2025 merchant volume is around 7% of that with higher shares in the small, medium and large segments. We believe that we are best positioned to grow that share to between 20% and 25% in the long term and still have 75% of the market to go thereafter. Let's look at the total addressable market on the supply side now. The Continental European car part is 190 million vehicles. 24% of car owners are interested in selling over the period of the next 12 months. a rolling indicator. This equates to roughly 45 million cars potentially up for sale. Out of this amount, based on our own customer data and models, we assume between 20% and 30% of car owners are generally interested in our C2B selling product. This number is a result of several factors, among them, for instance, the age of the vehicle they own and the preference for comfort in the sales process. Consequently, the resulting TAM for our C2B product is 10 million to 15 million units per annum. We assume that we can increase our market share to 20% to 25% of that TAM, which roughly equates to 2.5 million to 3 million units bought per loan and our long-term outlook. After laying out the TAM logic for both demand and supply, let's zoom in a bit and explain the high-level input drivers for scaling units towards those long-term targets. We'll start with supply. Broader marketing reach is a key input driver for us on the supply side. We already have considerable brand strengths with over 60% aided awareness in our largest sourcing markets. This makes it much easier to grow from here. However, unaided awareness is much smaller at only 22% for Europe, population weighted. This means there's a lot more investment potential for higher marketing reach in the future. An increase in awareness generally leads to an increase in the number of customers interested in selling via our CB product. As an additional catalyst ahead, we're executing a plan to step-by-step integrate our Auto hero and C2B brands with the goal to fully integrate our C2B brands into Autohero over a multiyear horizon. We're expecting significant synergies from this 1 unified brand for buying and selling over time. On top of that, we will continue to expand our branch network. The logic is straightforward, we've proven many times over. The shorter the drive time to one of our drop-off points, the more customers are interested in selling to us. Today, roughly 70% of our customers are within 15 minutes drive time of a branch. We believe that we can shorten this distance further and enable a much larger share of the population to be in close reach by building 1,000 by -- building a number of 1,200 to 1,400 branches in the long term. That would enable 90% of customers to reach us within 13 minutes of drive time. So far, we have built over 700 branches with a lean, standardized rollout playbook. For both drivers, we have more than a decade of experience in building and growing them, both our supply side investments fully within our control with returns we can measure precisely. Let's now focus on the input drivers for merchant demand. We have steadily scaled our buying dealer base from 20 dealers in 2012 to 54,000 buying dealers for last year. When we look at our demand base from a cohort point of view, then we can see that the dealers who stick with us are increasing their basket over time in a very stable way. This is the result from investments into 3 main areas: one, greater investment into sales and platform, better coverage of our dealer base, new platform features drive new dealer acquisition, higher activity and better conversion. More dealer demand also directly increases our pricing power on every car. Two, merchant financing has developed into a key demand driver. Dealers using our floor plan financing solution typically grow their basket with us by 40% to 60%. Financing eases our dealers working capital constraint and increases loyalty to our platform. Fulfillment, number 3 is a similar important driver. Every investment in faster, more reliable delivery increases conversion, basket and retention simultaneously as dealers can turn their inventory faster and can come back and come back for replenishment quicker. An important aspect of these unit drivers is that they reinforce one another. Greater investment in sourcing by broader marketing reach and our expanding branch network increases the level of supply and therefore, selection on auto1.com. Greater selection together with larger investments into distribution, attract more buying dealers and larger baskets. More demand for more dealers means better prices for sellers since pan-European demand allows us to pay more than locally in many cases. Higher volumes and better prices enable again more investment in sales, platform, finance and fulfillment, reinforcing this flywheel effect. This is a genuine network effect in the physical market, which is where most marketplaces have network effects, but no operational leverage because they do not touch the product. Most operators have leverage, but no network effect because they are local. We have both. And we know of our infrastructure serves both sides of the market and every car that flows through at data that improves pricing for the next car we trade. While the flywheel accelerated, merchant GPU increased from EUR 749 in '21 to EUR 976 in '25. We expect merchant GPU of EUR 1,080 to EUR 1,200 in the long term. We expect EUR 50 to EUR 100 improvement from better trading, constantly improving trading systems root each car to its highest value channel and buyer, improve demand forecasting optimizes our selection in real time, indicating needed volumes per specific car type dynamically. Improved AI pricing continues to increase seller and buy a conversion in parallel across 30 markets. Every car retrade makes the next trade more informed. Overall, we're expecting a compounding return on our investments into trading, data and technology. We expect a further EUR 50 to EUR 120 of improvement from financing and other products. Every transaction on auto1.com is an opportunity to attach value, whether it is dealer financing, logistics services or car and document handling. These attached revenues having robust margins because the transaction, the customer and the infrastructure already exists. The incremental cost of attaching a financing contract or a transport to an existing trade is minimal and penetration of these products remains at an early stage. Now let's go to the cost lines of merchant. Let us start with marketing. We are setting a long-term target of EUR 140 to EUR 110 per unit for marketing. This target corridor is based on 3 buckets of drivers that increase marketing efficiency over time. We are expecting the expanding branch network, the additional retail purchases triggered by our quickly expanding retail business and our industry-leading C2B NPS to contribute to higher selling conversion over time. We believe that our multiyear brand unification plan will unlock substantial synergies between the 2 brands. And in the case of marketing costs for buying a car, lower the cost of marketing per car purchased further and in line with historic trends. We expect retail trade-ins to become a more material source of cars purchased over time, essentially blending down the average marketing cost for car purchase as they have purchased at near 0 marketing cost per car. Let's take a look at the operations line next. So the variable cost of buying and selling a car in merchant. We are expecting operations cost per unit of EUR 350 to EUR 300 per unit in our long-term outlook. We believe we can improve operations cost per unit by EUR 20 to EUR 70 in the long term while we scale the business further. We expect EUR 10 to EUR 50 of improvement in operations from one, applying AI process improvement and automation through physical evaluation process with the goal to shorten the overall evaluation time and two, higher utilization of our drop-off network capacity over time. Additionally, we're expecting EUR 10 to EUR 20 of improvement from further sales and customer service process optimization, leading to efficiency gains. The last, merchant cost lines overhead. We are expecting overhead costs per unit of EUR 80 to EUR 110 in our long-term outlook. The strong reduction is in line with the trend of the past 3 years in which we have kept investment into overhead rather stable on a total base. Drivers of this development are: one, we took the majority of the investments for building up the central functions already in the past years; two, going forward, the relative growth of investment needed for central functions to continue to perform is limited; and three, there is further upside by applying AI and automation technology to central functions. So that completes merchant. Let's now switch to Autohero. The structure of this part is very similar to merchants. So we're first going to look at the historic disclosure and the long-term segment targets a view of the TAM and then detail the different GP and cost drivers for the long-term unit economics. From our start in 2020 to over 100,000 cars delivered last year and Q1 setting another record. Autohero is the fastest-growing car retailer in Europe. In parallel to unit growth, gross profit for the quarter has steadily grown to EUR 82 million for Q1 '26, driven by a very strong ramp of retail GPU. Adjusted EBITDA per unit sold is on a remarkable trajectory as well, starting with negative EUR 4,100 in '21 based on heavy upfront investment. It is on a strong trajectory towards breakeven with last figure, diminishing to negative EUR 410 in '25. This trend is also visible in the adjusted EBITDA margin progression per unit. Starting with negative 29.3% margin per unit improved to negative 2.4% in 2025. Let's look at these numbers on an absolute basis. Over the 5-year trajectory, our retail revenues tripled to EUR 1.76 billion. Retail gross profit reached EUR 268 million in '25, almost 18x more than in '21, driven by strong improvement in trading, consumer financing and the sale of attach products. SG&A was EUR 309 million in total, of which EUR 92 million was attributable to marketing, EUR 69 million to production, EUR 55 million to operations to consumer financing, EUR 52 million to logistics, EUR 34 million to overhead. Adjusted EBITDA was negative EUR 42 million a minus 2.4% margin. Let's look at these values on a per unit basis and framed a massive long-term opportunity for retail. Since '21, we have increased GPU every single year substantially with '25 GPU reaching EUR 2,638. In the long term, we are expecting retail GPU to reach levels between 3,880 and EUR 4,470 per unit. On marketing, we have generally seen a strong downward trend versus the 2022 and '21 numbers and have observed a level of EUR 650 to EUR 900 over the last 3 years. In the long term, we are expecting marketing per unit to reach levels between EUR 710 and EUR 540 per unit. One thing to point out here is that we are including the marketing cost for buying retail cars in this value. as indicated in the table. For total SG&A per unit, we saw values of around EUR 3,000 for the last 3 years. We expect total SG&A per unit, this includes marketing now, of EUR 2,430 to EUR 2,060 in the long term. Consequently, we're expecting adjusted EBITDA per unit for retail of EUR 1,450 to EUR 2,410 in the long term, a strong future upside and a continuation of the adjusted EBITDA per unit track of the last few years. We are combining these long-term targets for retail unit economics with a growth rate corridor of 20% to 40% per annum. We believe that this corridor, similar lead to merchant represents a prudent target that we feel comfortable achieving. Our goal is, of course, to be on the top end of that range and ideally outperform over time. Now let's turn to the Autohero total addressable market. We believe that 15 million units per annum are directly addressable with the current market footprint that Autohero has. We expect another 5 million used car transactions from the C2C market to be addressable over time as we generally observe a trend of decreasing C2C shares. Our biggest market, Germany, for instance, shows that trend nicely in recent years with C2C shares falling from 41% to 24% over the course of a decade. The reason for this trend is that consumers are increasingly choosing what dealers provide, trusted selection, warranty, financing and convenience. Autohero is best positioned to capture this huge overall demand pool, offering the most trust and convenience at the best price euro-wide. Now let's look at the different drivers of retail long-term unit economics, starting with trade GPU. Trade GPU increased from EUR 346 in '21 to EUR 1,860 in '25. We expect a trade GPU of 2,400 to EUR 2,680 in the long term. We expect EUR 290 to EUR 450 of improvement from better sourcing driven by a bigger retail database resulting in higher pricing position and lower error rates. More data can improve pricing materially as already demonstrated in the merchant segment. We also expect the share of trade-ins to increase strongly with more scale, and believe that enabling cross-border sourcing is a positive trade GPU driver in the long run. We expect to add EUR 240 to EUR 360 per unit from better trading optimizing selection with more scale, more precise demand forecasting, improved trading systems, new platform features, rising brand recognition, combined with positive word of mouth and faster delivery are all strong positive drivers for trade GPU over time. Now over to Christian Valentine for details on GPU finance.

Thank you, Christian. So consumer financing has become a key value driver in the retail segment. We started by attaching external bank financing and earning a referral commission has to kick back line on this page. Then we built our own captive finance business and the GPU per car stepped up considerably from just EUR 12 of internal interest per unit in 2021 to EUR 210 in 2025, an 18-fold increase in 4 years. The reason we can capture that margin rather than collect a referral fee is that our own product is simply better for the customer and a better product is what leads and lets us keep the economics. Financing is embedded directly in the Autohero flow approved in minutes with terms built around the specific car and customer because we own both sides of that transaction. That converts better, and it's why the captive lines keeps taking share from the referral line year after year. Our captive markets today are Germany, Austria and Spain. The long-term outlook is EUR 870 to EUR 1,100 of finance GPU per retail unit, 4 to 5x the 2025 captive finance levels. We get there on 3 levers: rolling up from 3 markets toward all outer markets, increasing attachment within those markets and letting the loan book mature because financing income is earned over the life of each loan. So today's origination build a stock recurring high-margin revenue at compounds for years. And strategically, financing changes what Autohero. A car sale is a transaction, a multiyear financing relationship is a customer. It deepens retention, drives potential trade-ups, the next purchase and turns Autohero relationship-driven business fueled by data and knowledge. This is one of our largest value pools in our retail model and it's already proven market by market. Let me now take you through the levers underneath our long-term GPU finance target of EUR 870 to EUR 1,100 and what to expect on each of the business as the business scales. I won't pretend this slightly simple, and that's partially the point. The complexity you see here mirrors the complexity of the rollout itself, building a captive finance business across market is genuinely hard operationally and structurally, but that difficulty is exactly why it's defensible. Let's start with the detachment. Group-wide, we are around 40% in Q1. In our capital markets, we already exceed 50%, while internal -- external only markets run at 20% to 51%. So the group figure is the blend of high internal attachment and lower external markets. Spain is the latest proof that the model travels. It has ramped according to plan over the past year and is now trending higher very quickly with over 20% of customers now taking our internal financing offer. Our long-term target is 50% to 60% attachment rate across the platform. On how the book grows, Think of it as a sizing tool, Autohero units times attachment, times the average loan that gives you the annual originations, the new lending we write in a year. Then you have to multiply that by our origination to AUM multiplier, you get the total loan book outstanding. That multiplier was almost 1.4x in Q1 and moves towards 2.5 to 3x long term as the business matures. Put plainly, we're earning on this year's new loans plus all the still outstanding loans from prior years. Net interest margin then tells you what you actually earn on the book, 5% today. Austria at 5.2%, Spain at 5.8%, with a long-term range of 5% to 7%. In our risk, this is a low risk by design. The Q1 cost of credit was 1.2% and with a long-term expected range of 1% to 2%. The structure of the region, it stays low, is twofold. First, we know the asset and the customer better than any bank could because we bought the car, we inspected it, refurbished it and priced it. And that same data continuously sharpens our scoring and underwriting. So the book is expected to get better as it grows. Second, we are the disposal channel for the collateral. When the loan defaults, the car come back to our own remarketing engine, so we expect our loss given default to be structurally lower than the generic lenders. Then the 2 separate efficiency stores. One is operating costs. OpEx per loan is EUR 74 today and heading towards EUR 50 at scale, driven by platform automation. The other is capital. The business is structurally capital-light. Through our securitization program, our own equity in the portfolio steps down with each generation of our securitization structure from 16% in the warehouse to 5% in our outstanding Finance Hero 2 structure to 1.5% in our just announced Finance Hero 3 enabled by vertical risk retention. As that structure rolls across the book, we expect the whole portfolio to settle in that 1% to 2% range long term, trending towards the 1.5% that we expect Finance Hero 3 to deliver. Each of these levers is already working in our captive markets, which underpins our EUR 870 to EUR 1,100 long-term GP outlook. Now back to you, Christian.

Thank you, Christian. The third driver of retail GPU is GPU other products. We're expecting GPU other products of EUR 610 to EUR 680 in the long term. We're expecting EUR 160 to EUR 210 of improvement from higher attach rates from our premium warranty products and an increase of warranty duration. Additionally, we expect the launch of our subscription model for royalties and improved bundling of royalty and financing services as long-term contributors. We're expecting EUR 60 to EUR 80 of GPU other products increased from smarter selling of second wheel sets in Europe and EUR 20 additional contribution from various other attached products like registration, insurance or maintenance. Let's now switch to the cost lines of the retail long-term P&L and start with marketing. We are expecting a combined marketing cost per unit for retail of EUR 710 to EUR 540 in the long term. This value includes the marketing investment needed for retail selling and the market spend needed for retail buying. We see 3 main drivers to reach this level over time. One, we believe that our plan to establish Autohero as a selling brand will turn the existing C2B marketing funnel into a low-cost buyer pool over time. Additionally, we are expecting a higher share of repeat buyers over time, lowering overall marketing cost per car. Two, we're expecting our brand strength to compound over time. lowering overall cost per unit in retail. Brand once built, is a durable demand generator that does not need paid acquisition. Together with increasing consumer readiness to buy cars online our brand audience itself grows every year, which increases return on every euro spend in brand. On top of that, our industry-leading NPS reinforces trust and word of mouth reducing the share of customers acquired through paid advertising over time. We're expecting that our growing retail customer base deepens our advertising insights into segments preferences and needs as a third level, unlocking, sharper targeting, messaging and execution and in turn, lowering retail marketing cost per unit over time. Now let's go to production costs. We define production costs here as the combination of 2 components: the cost of materials, parts and external refurbishment work recognized in COGS and the SG&A portion covering labor and production center costs for work done in-house. One thing is worth noting, as we bring more production in-house cost shift from COGS to SG&A over time. You can see this in the chart However, the total comes down, and that's the important message here. We are expecting a production cost of EUR 930 to EUR 870 in the long term. We're expecting EUR 10 to EUR 30 of improvement in COGS to optimize spare parts procurement and a further lowering of the share of external work. we're expecting EUR 120 to EUR 160 improvement in SG&A production to AI-powered workforce planning, a complete rollout of our patrietaryCAr audit inspection technology lean process improvements and structurally lower mechanical complexity with growing EV shares. For Logistics, we are expecting EUR 340 million to EUR 290 of logistics cost per unit in our long-term outlook. This number includes payroll and other OpEx for logistics. We expect EUR 170 million the EUR 220 of improvements through: one, the densification of our production and pickup center footprint, reducing driving distances by up to 1/3. And generally, shorter distances mean lower transport cost per car, faster delivery times and more satisfied customers. Two, synergy effects of combined flows and dedicated fleets between inbound and outbound logistics at scale; and three, higher utilization of pickup locations with growing scale. Now let's look at the operations cost per unit line. We're expecting EUR 450 to EUR 330 of operations cost per unit in our long-term outlook. We're expecting EUR 80 to EUR 160 of improvement in retail sales and customer service by applying AI processed automation, for instance, using agents in nonbusiness hours. We're expecting EUR 20 to EUR 60 of improvement in purchase operations to higher utilization and purchase process improvements. These are the same drivers we have outlined above for the purchase portion of the merchant operations piece. The final cost line is retail overhead. We're expecting EUR 130 to EUR 100 of retail overhead cost per unit in our long-term outlook. We are expecting an overhead cost per unit reduction of EUR 200 to EUR 230 per unit. In the long term, the strong reduction is in line with the trend of the past 4 years in which we have kept investment into overhead, stable on a total basis while scaling units strongly. Similarly to merchant, drivers of this development are we already took the majority of the investments for building up the central functions in the past years. Going forward, the relative growth of investment needed, more central functions to continue to perform is limited and there is further upside by applying AI and automation technology to central functions. Let's close the RTO section with an important effect to know about. The speed at which we grow retail is a headwind to short-term unit economics. As a high share of SG&A per unit cost, of course, roughly 60 to 80 days before that -- before the corresponding revenue and gross profit is realized. In other words, sourcing, marketing, production, inbound logistics and the purchasing part of operations costs occur in the P&L when we buy on recondition our fresh parts. On top of that, a portion of the sales marketing builds up demand that lies in the future. These customers are customers that started to be in the market for buying a car but will take weeks and months for the final decision. Put together, this means the faster we grow, the more of these costs we carry for cars not yet sold. And the bigger the short-term headwind to unit economics will get. In numbers, this means growing at a 20% rate per annum is a roughly EUR 250 per unit headwind. 30% means around EUR 350 and 40% around EUR 450. As a rule of thumb, every additional 10% of growth adds about EUR 100 of short-term headwinds. We generally believe it makes sense to grow faster given where we are right now, and given that we're expecting a major step in all of the drivers outlined before, when we approach a critical threshold of 1% retail market share. We expect that critical threshold to be somewhere between 250,000 and 300,000 retail transactions per year. This is a perfect segue for our milestone group targets. While we so far laid out long-term targets for both segments and the corresponding drivers in detail, we also want to give you a better sense for what the business will look like on the path towards these long-term targets, which we call milestone targets. We are not linking them to any specific year, but to the number of units that we think will enable these levels of unit economics per segment. When we will reach these will ultimately be driven by the sequence of growth rates over the coming years. So while this is not formal guidance, this is roughly where we expect to be as a milestone on the path towards those long-run targets. You can see here there is a range of growth rates and no specific date, but it should be helpful as you think about the trajectory to help build your models. Based on a merchant growth corridor of 10% to 15% annually, our milestone target for merchant is 1.2 million units per year at a GPU of EUR 1,025 or above, delivering EUR 400-plus of adjusted EBITDA per unit. Based on our retail growth corridor of 20% to 40% annually, the milestone target for retail is 300,000 units per year at a GPU of EUR 3,300 or above delivering EUR 800-plus of adjusted EBITDA per unit. The group milestone target roughly corresponds to the low end of our 5% to 9% margin target while the low end of our long-term targets would be within that 5% to 9% range and the high end of the long-term targets would exceed that range. But we think giving you goals in absolute euros is more useful for moderates than percentages. And while we absolutely stand by our prior percentage targets, going forward, now that we offer a lot more disclosure, we think keeping the focus on neuro targets makes it simpler for everyone. After 14 years of investment, into our vertically integrated business model, we have established an unmatched platform that maximizes value for car buyers and sellers across Europe. We are incredibly excited to continue our journey towards these targets outlined today, and with that, unlock the massive potential in one of the world's largest and most fragmented markets. Thank you very much for your attention. We will now go over to the Q&A section of this event.

Thank you. And if we could just have the readout of how to ask questions and then meet yourself.

[Operator Instructions]

Thank you. And actually, we'll start with 5 questions that we got from Joe Barnet-Lamb from UBS. Unfortunately, I think he always seems to have technical issues with the Zoom tool. So I'm just going to ask the questions on his behalf. The first one would be the timing on targets. What is the time line for the milestone and long-term targets? If we apply a midpoint of the growth target covered out to 2025 units, this would imply you will achieve your milestone targets in 2029. So should we think about 2029 as the target year for the milestone targets? How about long term? We are then having a question on cash conversion. The guidance you have given is obviously helpful, but there's nothing disclosed with regards on cash flow -- cash conversion. Over the long term, what cash conversion are you targeting? And what are the building blocks. And we are then coming to the, I guess, auto retail market in general. There's a decent amount of debate around the auto retail market at present. Can you help us understand your volume targets, what you assume for the underlying markets? I think then back to our business, production capacity and CapEx. With the some new production capacity on Slide 20, it's roughly EUR 250,000. Obviously, a milestone target is EUR 300,000. Can you talk about expectations for production facilities going forward? And also pick drop-off locations and how that plays into CapEx? And finally, on customer penetration, I think, especially in the merchant segment. When we look at Slide 27 and just share external dealer sourcing, it's obviously belle from smallest to largest customers. Is that a natural shape you expect to remain? Or are there product gaps and other specific blockers that are currently eating your penetration from the smallest and largest customers. I think Christian [indiscernible] over to you.

Yes. Thank you, Philip. Thank you, Joe. Maybe we should also glad to start answering some of the questions directly. Philip and maybe we do like 2 at a time or so because [indiscernible]. I guess too much to write down, maybe it's simpler. But yes, time line on target, I think, yes, that's pretty much something that I referenced at the end of the presentation. So we're not giving any specific time lines in a number of years for the milestone or for the long-term targets. However, we have given you a growth rate corridor, which I think you can definitely work with. So assuming the low end or the high end or something in between will get you to a specific year. And we believe that this is the best way how to represent an answer this question. I think now we're like the first 2 pretty much. So again, we're operating with annual growth rate corridor assumptions here, which are Yes. In the case of merchant, not too far apart in the case of retail, the spread is a bit bigger. You know where we are currently in terms of growth rates. So yes, you can pick pretty much like a growth rate assumption here that you find relevant. On the cash conversion, I think that's best for Christian Wallentin to answer.

Thank you. So we introduced 2 pages on cash in the appendix, and we will introduce those as well in the in the version that is on the web page. So we just want to do that. In summary, since we turned adjusted EBITDA positive, we have generated a cumulative EUR 367 million of adjusted EBITDA, as you can see on this page here and generated EUR 104 million of free cash flow. So that's a conversion of almost 30%, so 28%. So in the long term, as growth normalizes, we see this increasing to 40% to 50%. And I actually think we will come back to this page when we get questions on ABS structures and cash flow, but let's pause that for now.

On the development of the auto retail market, I mean, we've shown you a couple of numbers on the market volumes in the slide deck, we expect that, over time, the market will return back to its long-term CAGR of 2% from current levels. We do not see any like major disruption affecting volumes here. So yes, based on our models, I mean, we could work with a stable market from here, but we think it's going to Also, our long term, the assumptions we laid out on the TAM on the different ones. They're based on the current working on the current market size numbers. So we could work with a stable market from here, but we think that the market will return to its long-term CAGR of 2%. And on production capacity and CapEx. So for the -- and I think your question was on the milestone. We assume that we will have around 20 production centers. So we're going to then build yes, roughly 8, 9, 10, something like this and the CapEx needed for each center is somewhere between EUR 2 million and EUR 4 million. As a reminder, we're working here with brownfield and our greenfield opportunities. So this means we are converting existing facilities, which is low-cost approach. And yes, it's working fine.

And then, I don't know [indiscernible] briefly want to talk about the customer distribution on Page 27.

Yes. Can you repeat that question, maybe...

So I think the question is basically that we will continue to focus on the medium-sized customers? And are there any special barriers for us to work with the largest and the smallest customers. Maybe we can move to Page 27. I think it's the next one.

The next one? Or the one before -- no the one after that one. The merchant [indiscernible], right? No, I think you need to go back on ever. This one. I think this one is the -- the one in question. Yes, I mean we have higher market shares in the small, medium and large groups because these are just making up the bulk of the volume, right? So that's why we are concentrated more on those. We think enterprise customers are an interesting segment. But yes, we would need to adjust and invest and develop kind of our AUTO1 sales platform a bit different for them. So this has been something that was on the -- what it is on the table, but it's not something that we, at the moment, prioritize, given how much growth potential we have in the large, medium and small. So we would think that the structure of those shares that you currently see here grow in line of -- in line with the current distribution. So if we double it, then we'll double it, but relative from the value where it is today.

And we will now move to Andrew Ross from Barclays. And I think, Andrew, you had about 6 questions or so, so maybe we can split them into 3 blocks of 2.

So in the interest at 1 sanity after those 5, I'll keep it to 3. Can you give us a sense of phasing of the improvement in per unit that you're talking about between 2025 and the Milestone year when that is. Is there kind of a back-end weighting to it as you scale across investments you've made into brand marketing and intestinal being able to understand. And then I guess that leads into question 2. Can you tell us what level of units you would expect a retailable breakeven on an EBITDA basis on IFRS. Second question. I mean the third question, let's come back to that slide in the appendix on the cash conversion. I can't see that yet on the website. It would be quite helpful to go through that in more details on kind of here the answer is a conversion of 40% to 50% from EBITDA into free cash flow. But can you walk us through in more detail the CapEx and then, I guess, particularly in the working capital and how much capital you're expecting to absorb versus kind of inventory, receivables and I guess, I mean, going the other way on payables, that would be helpful to understand in more detail.

Okay. So blot #1, Andrew, so the phasing of improving unit economics from where we are today and to the milestone target. So as indicated in the script, we would think there's a stronger progression on each of those drivers, the closer we're getting to the critical threshold. So somewhere between 250,000 and 300,000 units, we would expect to get a majority of the improvement. Yes, this is what I would say as a trend. So as we approach those units, we're getting stronger ramp of all of the drivers outlined above. And before that, we'll also see improvements but of a lower absolute improvement. And the reason, for instance, if you think about in logistics, it's just that we need to through like approach of certain density of the transaction network. So logistics, I think, is an easy example to understand that now at 100,000 units and then a 2.5 to 3x of those units, the driving distance is just smaller. And there is more potential to bundle the inbound and outbound fleets, and that leads to this ramp and reduction of the delivery and logistics cost per unit. And that's something that then is really like kicking in as a stronger lever the more dense it gets. So it's not a linear improvement. I think that's what I'm pointing out here. Yes, what level of unit will we be breakeven in Autohero. I think if you look at the numbers, you can see that we're pretty close. If you calculate in the headwind to short-term unit economics and the growth rate for 2025, then you can assume not growing would actually be on an adjusted EBITDA breakeven already. However, we choose to grow because it's exactly off the answer to question one. And yes, if you look at the trajectory, then we have been on a constantly improving track and I would say, we're close. But again, it depends also on the level of growth. And we try to maximize growth under the site target of overall group profitability. So that's what we're trying to manage in the best point way. And yes, I hope this answers your question.

So let's move to the more on the cash question. So this is related to both our operational cash flow and then the ABS structures and the net debt. So I think this will be a slightly long-winded answer, but I want to do this because we received a lot of questions historically from various levels of knowledge. So I'll tag through it in a structured way. So first of all, I think most importantly, we are cash flow positive and our projections, we self-fund our growth. So that's the takeaway. So just to illustrate that point, we ended 2022 with slightly less than EUR 550 million cash or so on the balance sheet. We had EUR 652 million cash on the balance sheet at the end of Q1 '26. So we raised no debt other than the ABS funding against our assets and no equity in that period. So that is EUR 110 million cash generation from 2023 to Q1 '26 on top of what we invested into very strong growth in our business and expansion, particularly in our capital finance activities. So now to the -- so I'll go through the facts of the explanation. We have 3 types of ABS structures for inventory, merchant finance and consumer finance receivables. So one, nonrecourse funding from the ABS facilities collateralized by assets, so receivables and the inventory. So those you see on the balance sheet and nonrecourse means that the corporate entity of AUTO1 is not live of credit losses. And that is lenders cannot force an event of default on AUTO1 if the underlying assets underperform or lose significant value. So these trucks are funded by banks and public investors who like the risk of cars and also in the auto financing secured by cars. Three, there's no need to use cash to repay or put any more cash into these ABS structures. The underlying receivables and the cars being sold or what paid the banks and investors with interest. So if needed, the debt is self-liquidating against specific inventory and financing receivables. The structures have been successfully tested in the toughest environment, meaning COVID and been riding through that with stat. So therefore, very logically, this non recourse debt is not part of, one, corporate net debt or to the cash flow we need to fund as we grow. This is also supported by the rating agencies for large U.S. car dealers. We have no rated peers in Europe, by the way. They always exclude inventory financing from debt ratios as they view it as working capital like item. So the agencies also stripped out captive finance assets and liabilities from the corporate credit ratios under the captive finance policies. So this is the difficult hear now that I'm going to say. So the economic reality is what I described now, and it's not as obvious in accounting for a couple of reasons. So when we consolidate these structures, cheaply to control the servicing of the underlying assets we'll be absolutely want to do because we see better credit performance if we control the contact with our clients. So the accounting rules do not reflect the economic risk of the setup given the only risk that what was put into these structures originally at inception. And under IFRS, we show any increase in assets like inventory or captive finance as a cash outflow in operating cash flow, while the related ABS funding is shown in cash from financing. Hence, the impression is that we are cash flow negative from IRS perspective, from an economic and management point of view, and hopefully, also your view, the assets and the funding are directly linked. So as out to 1 corporate and shareholders, we will only fund the net proportion of these assets, which have not been funded by the banks or ABS investors. Hence, we focus on something we call 21 managed cash flow. It looks at the net movement of the assets and ABS facilities. On this basis, we've been cash flow positive since at least 2023, and you think in the example that I had in the beginning, proves that point. So 3 conclusions on this. So we're doing this in a capital-light way. We only invest cash at inception into the equity of the structures. The large majority, depending on which structure we talk about is 80% to 99% depending on if it's the -- for example, the [indiscernible] 3 structure out in the market now. will be worth risk retention, that's up to 99% is funded by the banks and external investors. So there's no further claims on us than the assets in that specific SPV. Conclusion number 2, we are cash flow positive, and we self-fund our growth as these are evergreen ABS structures, and they scale with the business growth. Three, accounting rules do not show the economic reality of the business as we forced to consolidate them even though we control the service don't call -- because we control the servicing of the assets. So that's just the facts of the matter and the conclusions from them. So we are in order to simplify this for you and for people that follow us, we're publishing managed cash flow with the results on a web page on a quarterly basis. We will also include more explanatory slides in the earnings presentation. We have already started that. So please look out for information there that will go into more details. In addition, we're also considering having a modeling session in the future to explain these technical questions in a more structured way. So today was very much about the segments and giving you that segment detail. and we'll consider to go back to these more technical modeling topics as well. So with that, I think, Philip, if I missed something that you think is relevant and please add on.

I mean I think it's a very good and I hope easy to follow introduction to the topic for everybody. And maybe just 2 comments. I mean, one, this is actually new disclosures. If you actually followed our web page, we were always publishing this Excel spreadsheet but the quarterly earnings numbers, which included this AUTO1 non-IFRS cash flow. It's also normally contained in the highlights section of the financial report we are publishing each quarter. But I think going forward, it's something we will also take into the earnings presentation to really make sure that everybody understands the point and sees our view on cash flow. And I think the other thing just to mention because I actually don't think a lot of you realize this is just how also operationally, we are linking the asset side and the ABS fundings. So this is not the case to be just every 2 weeks or so, collect our information and then go to the banks and get an advance rate. But we literally -- if we have to pay for a car that we are purchasing today, we actually tell the funding SPV in the morning, look, this is the cars we are paying for today. And this is the cars for which we got paid yesterday. And so this is the net change that we need to fund today. So it's fully integrated. And then, for example, if a merchant select use merchant finance, we're literally on the same day just have a transfer of money from the merchant finance ABS structure to the inventory ABS structure. So again, that runs fully automated and fully integrated, and this is why we are really focused on saying that -- saying that those 2 always belong together. And I think one of the challenges we always have from a tech and structured finance perspective to make sure that as we develop new products, we always keep that pipeline of making sure that as we generate the assets, we immediately also raise the refinancing and operate in this really efficient and capital-light manner. And that is what being hopefully everybody will kind of agree that looking at this AUTO1 cash flow and the net changes in the inventory and capital finance assets makes much more sense than looking at the assets separately up in the operating cash flow and then the ABS funding down and the finance cash flow. But I think this is really the point about -- well I think we are so comfortable in our ability to self-fund the business. And I think just don't really often understand a lot of the questions about when will you be free cash flow positive because my perspective, we have been free cash flow positive for 2.5 years.

In other words, IFRS was not built for used car dealers.

Yes. After that, a long discussion explanation, I think James, you had -- James Tate from Goldman Sachs had 2 questions.

Well these are all questions from Andrew. Or did he have more?

I think he said he only had 3 because of the other question, I think had been answered.

I think we had one on the CapEx as well. I mean, we historically have guided to 20 to 25 basis points and that has proven to be quite generous historically, so we never really gotten there. So that ties to the number that Christian gave in absolute terms as well.

And so that was meant to be 20 to 25 basis points of revenue or about $20 million to $25 million probably this year.

Yes.

James?

I've got 3 questions, please. I guess, firstly, and following up slightly from Andrew's question on retail EBITDA per unit. Could you give some color on where you think you'll end up this year or where you're trending so far through 2026 in terms of retail EBITDA per unit compared to the minus EUR 400 in 2025. That would be really helpful. Secondly, in terms of the SG&A lines for the retail business, noticed operations and production costs per unit have gradually increased over the last couple of years. I guess, firstly, what's driven this? And then you've outlined how these costs decrease over the long term, but in the near term, do you expect these to have peaked in 2025 or is there further investment required here to drive growth? And then lastly, Christian, towards the end, you mentioned that strategically, you think it's best to drive faster retail unit growth. So is it fair to assume that retail units should continue to grow towards the top end of the 20% to 40% corridor over the next couple of years?

Yes, thank you, James, for these questions. So as I told I can understand the curiosity of retail GP of retail EBITDA per unit for this year. We're going to disclose that. I think what we can say so far that I mean Q1 was a very good development. But as also coming back to your third question, right, we try to balance growth and profitability to the best way possible on the way to the milestone target. So this growth rate corridor that we have given 20% to 40%. If you look back at the last couple of quarters, we definitely have seen increased reach rates within that corridor also north of the top end of the corridor. So we try to best balance profitability and growth as we go forward. And yes, also, as indicated in the script, Definitely, we would want to be on the high end of that corridor. But at the same time, we also want to see some okay to good progression on the EBITDA per unit in order to make the full group's EBITDA per unit growth. So in that sense, now that we have the full disclosure. We can also say that, obviously, the reference for growth in Autohero that cost them more as a negative short-term headwind as illustrated also depends on the profitability and cash generation from the merchant business, yes. So let's say, like if you're advancing their base in better, and we can also let retail grow faster and the other way around. Now in Q1, merchant GPU was as explained a little bit down. So then this introduces also like a bit of a site condition where we say, okay, how much can we grow in retail given EBITDA. So this is kind of all the variables that we try to manage in the best possible way. We're leaning towards the higher end of that growth corridor. And because we strongly believe that we will see substantial improvements on retail unit economics, the closer we get to the critical market threshold, market share threshold of 250,000 to 300,000 units per year. So I think it's too early to think about retail EBITDA for the year, but we expect that we are continue our progress that we have been showing now for the last couple of years and balance it nicely with the growth. On the SG&A production cost slide, maybe we can quickly go there, Maria, I think the point that you asked -- can we go to the slide, please, Maria? Yes, this one. I think the point that you asked was really the SG&A production portion per unit that has slightly increased in 2025. So overall, we can explain this again, the blue part that you see and the blue part with COGS production per unit, in '21 and in '22, that is essentially the ramp-up of the internal refurbishment capacities anything externally how we did it at the beginning in '21, I think also in the quarterly reports for those years, you can really see that we reported on the -- that we reported on the internal external shares and how we progressed on reducing the external and increasing internal share. Pretty much this is the financial story behind building the internal refurbishment capacities. So the blue dark-blue bar going down. And then we started to increase our SG&A production per unit, which includes a strongly payroll, but also overhead cost of a production center and the total capacity, the on-top capacity that we found in rent and in facility. And this portion has increased slightly in 2025 because of lower utilization. So we built up more capacities towards the end of the year that we are now utilizing in a better way. And then special effect inside that is also that there's temporary workers that we are -- from time to time needed to use over the last couple of years, which are really expensive, much more expensive than our core personnel in case we did not -- we were not able to hire enough capacity fast, and that's something that we've are in the process of managing much better. But that explains kind of these numbers. So the important part for us is that we have arrived on levels that we consider very good versus '21 and '22, and we're seeing further upside. But not even that much for an upside because ultimately, it's a process, it's a quality standard that we think EUR 100 to -- EUR 30 to EUR 190 are possible for the full block. I hope this answers your question.

And with that, Wolfgang Specht from Berenberg.

I have 2, if I may. First, on the refurbishment and you explained. Can you give us an idea if you have to do extra investments, not only at the centers, but also on the drop-off locations to get better diagnostic of the cars and, let's say, keep the weak horse out of the system? And second point, your largest driver for gross profit per unit is definitely retail finance, how do you expect to cope with competition from retail banks that have been really weak in this discipline over the last years. But supported by AI tools and better front end maybe retail financing could get attractive for a couple of your customers as well. How do you want to keep customers in your system?

Yes. Let me take the first question. So within our current evaluation process, we're not expecting incremental investments that we need to do or, let's say, material investments that will be changing any of the numbers that we just set out on targets. We're not expecting that we need to materially invest in our drop-off network to increase the quality of our cars. So every retail park that we are buying for a retailer that has the target channel retail is being checked at the production centers very thoroughly with all the equipment needed. So this is where we have it. There's a small share of cars that we sort out. But overall, for us, this is a more efficient way of dealing with that problem. So those cars will then be reevaluated and sold back to merchant. And this is partly more efficient solution versus having many production centers, time 700. We have tried this, and we know that this is the more optimized and streamlined workflow. Second question, competition from retail banks.

Yes. Thank you, Wolfgang. So you're correct that this is an important value driver. And I think the answer is that we see that when we use external bank, one, first of all, it's a protected sales, so to speak. So this is Autohero units that we sell. So we don't let anybody finance that specific car. So if somebody wants to buy a car, and this goes back to the superior retail experience that we have, which is driving our growth as well. So when we are selling it, we see that the internal, so captive finance solution is much better product. So this is driven from both being better embedded and quicker to prove that drives higher detachment rates. And that's not possible for an external bank to replicate in our system. So I think we're comfortable that we can offer a better product without any buying the car without finance and then going financing it somewhere else. So we certainly believe that the attachment rates that we have in the captive markets in Germany, Austria and now growing in Spain as well, is indicative of the future potential because it's reflecting a better product.

And I think that that's Nizla Naizer from Deutsche Bank.

Thank you for the very helpful presentation as well. I have 3 questions from my end. First is on used car pricing. Could you kind of tell us what sort of assumptions for used car pricing over time you are incorporating when considering these absolute long-term targets that you're giving us? In other words, if the average prices of used cars go up? Is that incorporated in the range of the EBITDA that you've eventually given for each segment and vice versa. Some color there would be great. My second question is on your milestone 300,000 Autohero car target, does this include launching in new geographies and moving into different subsegments of the consumer-facing used car sales that you're doing, in other words, maybe lower-priced used cars? Or is it with the current sort of trajectory of used cars that you're selling? And lastly, similarly to reach the merchant milestone volume target, will you have to start buying cars from new markets? Or will the 9 sort of sourcing markets that you currently have be sufficient?

Yes. Thank you, Nizla. 3 very good questions. So yes, we generally see a trend of rising used car prices over time. That depends on how the new core cohorts are developing. But we deliberately have baked that in the long-term targets. So I think it's very hard to really map out kind of how the used car prices will develop over every single year that we have modeled behind. So that's why we chose to make it simpler for everyone. And the answer to your question is, yes, we have incorporated any price trends. So it's not like that. Now if used car prices are increasing more than we would expect our milestone or our long-term GPO targets to increase further. We want to make sure that we keep things simple and for everyone involved and try to get the ASP part of the equation. On the milestone targets, we're not assuming any new geographies for retail. We're also not assuming any new geographies on the sourcing side for merchant. We might want to experiment with cheaper used cars in retail, but with actually very young used cars or also with different forms of owning the car, which might be subscription, but that's something that so far is only an idea and it's also not baked into the target. So late as a requirement. So for the targets, it's really the current business setup and its long-term profitability potential.

And then Marcus Diebel from JPMorgan.

Just 2 questions left. One question again on those targets. Christian, I appreciate you obviously don't guide for certain years, and it's okay. Just to really understand that these growth rates, 30% to 40% and 10% to 15% growth. These are not CAGRs, i.e., if you reach the milestone target 1.2 million cars, you feel comfortable that inversion that on that number, you're still growing at 10% to 15%. And then similar in retail, once we reach 300,000 cars you're comfortable to grow that number still by 20% to 40%. It reads like this, and maybe it's 2 simple question, but I just wanted to clarify this. And then maybe the second question is more specific is on the GPU and other in the retail business. A lot of growth comes from premium warranty. Could you just tell me a little bit more how it works because you incur clearly the fees for the premium warranty at the beginning and the cost come in later? How do you actually reflect that in the GPU numbers, just a more technical question, I guess. Just these two, the rest has been answered. Thank you.

So on the CAGR growth rate question. So yes, that's indeed more how we mean it. Like you just said it. So we also assumed that after reaching the way point that these growth corridors are still valid. So while I think some market participants will assume slightly declining growth rates typically by any business over time. So we stand behind those corridors and think they are intact for the Waypoint or will remain also the same for the Waypoint and then also beyond. And yes, on the very specific GPU other question. So obviously, this is a bundle of products that we're selling here. but specifically to warranty provisions, we are, to my knowledge, but please Christian and Philip chip in. To my knowledge, they are pretty much a recognition of the revenue over time and also provisions over time. So they are being booked as cohorts and the revenue, but then also the assumed cost and provisions for the future cost of those warranties, there directly booked in there. But maybe Philip...

That's correct.

No, no. It's clear. I want to make sure both is pro rata, but that sounds like it's okay.

Yes, yes.

So I think the important answer is the gross profit contribution it's not just the revenue, but it's also the provision for the expected claims that we need to pay out.

And then clearly, you have those others in there as also set of wheels registration, insurance, maintenance, that sort of thing.

And then we got Mourad Lahmidi.

So I have 2 questions. The first one is on the internal financing. So on the slide where you showed the attach rate, which is 40%. I just want to make sure that the 40% is on this market where you offer the internal financing and not on all markets, are there any markets where you have restrictions in terms of offering the direct financing? And also a question on how the finance when it's accounted internally is accounted for. Is it the net present value of the future net interest margin? Or is it the current year net interest margin? So that would be my first question.

Okay. Just very simply, Mourad, I think on the first question, the tile that we're showing at the attachment rates, as we said, that 40% is actually external or internal financing across all markets. So that just shows the fundamental demand for the product. So 40% of all our customers are refinancing today. And then we're kind of showing that, for example, in Germany, Austria, we're having the end channel product actually achieving above that total level, [indiscernible] a proof point for our ambition of 50% to 60% in the long-term attachment rate on financing. And then if you're looking at the external only markets, they're ranging from 20% to actually also over 50% in one of the Nordic markets. And then to clearly replace the external with internal given that we will capture the full profit pool. And then the net interest margin or the gross profit contribution that we have in that is the actual net interest realized in the period on the actual complete loan book. So we're not just having some net present value booking or forward looking. And that is why we're saying if you actually build up that loan book, that obviously then generates an annuity stream over the next 5 to 8 years.

Okay. Very clear. So I have a second topic, which is the strategy of integrating C2B with retail more and more going forward. Can you please elaborate on how this will help the business grow faster or deliver higher margin or being more efficient.

I mean, in a nutshell, Mourad, there's a lot of customers in our sales funnel also interested in buying a car. When they are evaluating a car so, let's say, like 1 month to 6 to 8 weeks before they are actually then finally occurring with their sale. So let's say, we have a lot of customers on the C2B side, or a lot of customers -- potential customers evaluating their car, while they're in the market for their new car. So they typically tend to solve that problem or that new car first because you don't want to be without car. And then thereafter, they are finding the best selling solution for the residual car if the dealer did not take their old car and the trade in. So there's a big pool of customers that may be synergy level 1 that are interested in buying a car when they are already in the process of finding out the value of their old car. So now all of these sessions and traffic at the moment happens on C2B. And of course, if it was happening on outer under the site condition that Autohero stands for both and that customers in Europe actually understand that Autohero can also buy cars and there's a lot of future synergies, similar synergy we see in the physical space. So we have 700 branches plus that we operate, and they're all branded with the C2B branding and the C2B brand is very strong. However, if it was an integrated or at some point then full Autohero branding, then we will also get the benefit of additional awareness coming from the physical representation of the brand. onto the sales funnel. So it's just better if everything were to happen on the 1 brand. However, the C2B brands that we have, they are high-performance conversion machines, if you want to. So this means that we will address this and execute this very carefully over time. The synergy potential that we see is expressed by the combined long-term reduction for both marketing numbers in the long-term targets for each segment.

That actually brings us to the end of the question list as well. So I guess, just on time as well. So I think Christian, thank you very much. I think that's a very insight, Phil, but also I think everybody quite happy session. So I think that's lot to digest, obviously, Maria and myself and the team, we are available if you have any questions or issues. I think we also got quite a full schedule of conferences and meetings coming up for the next 2 weeks. I'll probably see quite a lot of you in any case. Otherwise, we got the Q2 and first half numbers then coming up at the end of July. So thank you very much for participating for listening in. I hope this was a useful and interesting session. And thank you, Christian, Christian and Maria for all your support.

Thank you so much, everyone. Yes, I think indeed it was very useful and helpful session. We hope that we have made it a bit simpler for you. and everyone else out there tomorrow at AUTO1. Yes, we're absolutely excited to continue to march to these long-term targets. But now without further talk in, let's close the session. Thank you very much. See you soon.

Thank you everyone.