[Non-English content] Good morning. Good morning, dear spectators, dear audience. Welcome to the Investor Day. Today is a very special day for us. Sberbank is celebrating 185 years, its anniversary. Happy birthday to Sber. On this day, we've decided to hold a very important conversation. We're going to summarize the three-year strategic cycle, which finalizes this year and t hen we will talk about the further development of the story that we started almost two centuries ago.
We're going to talk about the priorities of the company, the new opportunities, and new challenges. You will hear from the top management, the CEO, Chairman of the Executive Board, Herman Gref, First Deputy Chairman, Alexander Vedyakhin and Kirill Tsaryov, Senior Vice President, Head of the Technology Development Block, Andrey Belevtsev, and Deputy Chairman of the Executive Board, Chief Financial Officer, Taras Skvortsov and t hen, live, we are going to discuss the details of Sber's strategy by 2029 with the CEO of the bank and with the investment analysts. We've invited them here in this hall. Let's begin.
For the latest three years, we've had a time of serious economic challenges, which means this is our time. Salute! Show us our successes. We've delivered on all of the goals that we've set for ourselves. We have paid record-high dividends. We've increased the number of shareholders. We've sent almost RUB 4 trillion to the federal budget, and we have improved the lives of our dear clients. We have turned our business model from products to the human being. Family, travel, and everything you need is in a single space.
SberPrime is a flagship multi-subscription available for your family. Almost 110 million people trust us. 3.5 million companies trust us. Together, we are composing the symphony of success. Great success, but that's not all, right? Salute set the winning rhythm. We have built one of the most powerful anti-fraud systems. Sber is the second-biggest acquirer in the world. We have launched the flagship sovereign model, GigaChat. We are building an AI-native company based on our own proprietary technology.
Sber has been supporting projects which are changing the lives of people: education, science, healthcare, development of cities, culture, and maintaining cultural heritage. Sber is now 185 years old. Today, the world is accelerating, and Sber is accelerating along it. But it's not enough to be catching up with the world. We want to set the tempo, to set the pace. Fasten your seatbelts. Are you ready? Sber. We carry on.
Good afternoon, dear friends. Today, we are not just celebrating a date, any date. We are 185 years old. 185 years of continuous trust. Over this time, generations and entire eras changed. Technologies and the way of life changed. But through all the storms, one thing never changed. People have been trusting us with their most precious things: their savings, their dreams, and their future. Thank you for your trust, and thank you for being with us. Let's see what we've come to this date with.
The last three years have become a kind of a real test of resilience for us. Ultra-high rates, a decelerated economy, growing tax burden, and unprecedented sanction pressures. Geopolitical tensions have reached historic heights, and uncertainty has become the new normal. Any other business in our place could have just frozen up, but Sber took a different path amid all the tumult. We have not been just surviving. We continued to invest in the future. We have not been hiding from reality. We have been creating it. We launched services that take over the routine and create solutions that increase wealth.
We did not just maintain our positions in financial and non-financial markets. We strengthened them by supporting 1 million people in these turbulent times. Over the past three years, we have done the impossible. We released three generations of our own language model. GigaChat is not just a technology. It is our digital sovereignty. We proved that we can create not just financial products, but fundamental technologies of the future and a t this crucial historic juncture, we have won. Despite all the external chaos, we have achieved all of our financial goals.
The return- on- equity is above 22%, and this is the pulse of a healthy and powerful organism. The capital adequacy ratio is above the regulatory requirements. We have created an unbreakable margin of strength. Dividends at 50% of net profit annually, and this is in line with our promises to you. Over the three years, we paid over RUB 2.4 trillion to our shareholders. Just think about it, RUB 2.4 trillion. That is what Sber paid out over the previous 20 years. We did not just preserve the money, we created a new reality of yield.
Summing up the implementation of the current 2026 strategy, we see how its goals and principles have been translated into concrete business results. First of all, we have strengthened our client base. Tod ay, it comprises 3.5 million corporate clients, which is 300,000 more than three years ago. Client engagement with the loyalty program has increased significantly. One in three clients is a member of the program. The bank's share in the corporate lending market is 32.4%.
The loan portfolio today is almost RUB 33 trillion, up 41% since late 2023. 55% of new lending transactions for medium and large businesses are now executed fully autonomously from application to disbursement. The loan portfolio for medium, large, and largest businesses that was disbursed using AI reached around RUB 8 trillion. Another crucial stage was the transition to real-time customized offers to each large and medium-sized business client online. The tailoring is based on pricing as well as the selection of products and services. 53.5% of transactionally active small and micro business clients choose Sber as their leading financial partner. It does not end there.
Most importantly, introduction of AI has been changing the economics of entrepreneurship. More bold ideas can be implemented over less time and with fewer resources. This is no longer just a tech trend, but it is a new reality in which the rules of doing business are changing. An AI assistant and AI agents are coming to the forefront, taking on more and more tasks, from analyzing information to performing concrete actions. We are actively developing our assistant at two levels, as an executive or a business owner, and separately for a company employee.
The first level is a business assistant or a human assistant for business development. Its business decisions are linked to personal goals, plans, capabilities, so the assistant should understand the situation with the business, help choose the next step, and autonomously execute instructions if, of course, the owners of the business allow to do so. Today, the business assistant mode is already available to clients in Sberbank's internet bank. In the chatbot, you can obtain analytics, request analytics, run checks on a counterparty, set up a payment, and confirm a transaction.
We are also developing an assistant that will understand the entrepreneur's goals to work with a large amount of company data and organize the execution of current and urgent tasks. For example, "Open another point of sale. Can I afford to do so?" the entrepreneur asks themselves. The assistant will analyze revenues, expenses, and liabilities, calculate how the opening will affect the cash flow. It will offer a variety of options. It will help identify the need for funding and come up with a realistic action plan.
Once approved by you, it will bring in the necessary agents and services to search for premises, recruit personnel, and prepare documents. It will follow up on execution, and it will get back to you if you need additional input. The second level is GigaCowork, an assistant for the company employee in daily tasks. It is an environment for employees and agents to work together in the company. Here, they will be able to work with corporate data and documents, allocate tasks, and execute them in the company's systems. The managers, their teams can offload some of the day-to-day work to agents, thereby solving their problems faster.
This way, the assistant becomes a business development partner for the entrepreneur, manager, and the entire team. With GigaCowork, you can help prepare a client offer, study sales dynamics, compare terms of different suppliers, launch recruitment, and more. The employee sets targets, the agents can perform the necessary operations, and the person checks the result and makes the final decision. Let's say the team is tasked with developing a commercial offer. Agents will collect the information on the client, check for products availability, calculate alternative scenarios, and prepare documents.
Employees can approve parameters and choose the optimal scenario. As a result, the company can scale its work without increasing the load on the team. Employees free up their time to communicate with clients and to develop their business. The business assistant and GigaCowork form a single loop. The idea that the executive discussed with the business assistant, along with the associated context, can then be implemented by the team in the shared workspace.
The next step is organizing interactions between agents of different companies. For this purpose, we are developing GigaNetwork. At the St. Petersburg International Economic Forum in 2026, we already presented a successful pilot project. Agents of buyers purchased industrial equipment from a supplier, and the entire cycle, from the request to payment, took less than a minute. We strive to ensure that sophisticated, elaborate technology that used to require huge resources would become an affordable tool for a wide range of business people.
For Sber, this opens up new opportunities for growth because relevant business solutions create additional incentives for entrepreneurs to become our clients. The wider the range of everyday tasks that we help with, the stronger the connection with existing ecosystem partners. Support of procurement, sales, and company development forms the basis for increasing the payment turnover, account balances, and transactional income. A deep understanding of business needs can scale lending while keeping risks under control. In addition, the services themselves, AI services themselves generate additional revenue. This is about the synergy of leadership and technology and finance. By helping our clients grow, we are increasing Sber's income and strengthening the Russian economy.
Good afternoon, dear shareholders, dear investors. Summing up the results of the strategic cycle 2024- 2026, in retail business specifically, the key results are: Sber strengthened its market positions and created new growth opportunities. We expanded our presence in the everyday life of people, we updated the payment infrastructure of the country, and started the mass implementation of AI in client services. Let's start with our basic lending services. During the strategic cycle, the portfolio increased by RUB 5 trillion and exceeded RUB 20 trillion. Our market share in three years increased 4 percentage points to 50.7%.
In the previous cycle, we added another 5.5 percentage points. Over six years, we completed a journey from about 41% to more than half of all retail loans to private individuals. This is a significant change in our position in financing the lives of people. We strengthened our positions in key areas of mortgage, credit cards, car loans, cash loans, and in niche products such as education and salary loans. In savings, we maintained our leadership and significantly increased the amount of funds raised.
Private clients' funds increased by about RUB 12 trillion to RUB 35 trillion, and our market share increased to 45%. This is not a new record, but a recovery of our positions in that market. In the investment market, we not only recovered, but also increased our market share by 6.7 percentage points to 27.8%, which is above our pre-crisis levels. These results were achieved in the environment of high stakes, of high key rates, and serious constraints. It is important for us that this growth is ensured by building long-term relations with clients, the quality of management, risk management, and efficiency of capital use.
We have about 111 million people with us today. More than 43 million people use SberBank Online daily, and more than 84 million use it monthly. On average, a client uses almost six of our products every day. Our goal is to be useful in more life situations, and consequently, we have prioritized a number of important areas. Payments come first. They are a daily contact with the client and the basis of relations with the trading business. During the strategic cycle, we changed the way people pay for purchases in Russia. We turned payments with a smile into a part of everyday life.
In Q3 2026 alone, clients made 69 million payments by smile. By the end of the third quarter, the stock of biometric terminals reached 1.8 million devices, which is 80% of our terminals. The speed of biometric transactions was quickened by more than a half. We brought back the convenient contactless payment by iPhone. The Whoosh technology is available for connection by other banks, and clients of different banks, other banks already use this solution to make daily purchases.
115 million payments were made in the third quarter of 2026 with Whoosh. Cards, smartphones, payment rings, biometric watches, the customer chooses the most convenient method. This year, we introduced Neo, our own terminal with embedded AI. It adjusts to the buyer's position and lighting conditions, allowing you to pay in installments, use bonuses, and leave a tip. We value not only convenience and variety for our clients, but also the security of each and every payment. We are constantly improving fraud protection technologies.
Over the three years, we prevented a theft of about RUB 1 trillion. The next important result will be developing the ecosystem and turning SberBank Online into a place not only for financial transactions. We have brought together the capabilities of the bank and the ecosystem. In the app, you can buy goods, groceries, plan travels, purchase theater and cinema tickets, all with payment by usual tools and Spasibo loyalty points. In 2025, these features were used by more than 50 million of our clients. We combine the capabilities of the bank and the ecosystem around everyday tasks.
For example, the auto and home sections collect financial and non-financial solutions for car and real estate owners. This approach is also being developed in DomClick, and every client can customize the SberBank Online app to their own needs. The ecosystem services connect Sber, SberPrime, and Sber's Spasibo loyalty points. By the end of Q3 2026, Sber served 128.5 million people. This is more than the number of active retail clients of the bank. The ecosystem also attracts people who do not use Sber's banking services yet. For us, this is a great additional opportunity to introduce them to the bank and build relationships with them.
As of September 2026, SberPrime had 22 million subscribers. Each client of SberPrime uses more than seven ecosystem services on average. In this strategic cycle, we began to develop family banking, the joint management of money, savings, and planning of large purchases. The child's card is available from the age of six, and parents can control the child's expenses and help them in the development of financial literacy. Financial capabilities are complemented with care for health, geolocation of loved ones, and protection from fraudsters. It is important for us to help our clients effectively manage their savings and prepare for their pension.
For example, the PDS long-term savings program in the first two years brought together 6.5 million participants, and the PDS assets at Sber have already reached RUB 865 billion. We pay great attention to the convenience of investment management and want investments to become permanent, simple, and mass-scale daily solutions for 1 million customers. We understand, of course, that the competition in the investment market will only increase, so in this strategic cycle, we paid a lot of attention to the development of our services, including Sber Investments.
On the 1st of September, we launched the easy transfers between brokerage accounts. We have been preparing this very thoroughly, and today we can already say with certainty that our clients are voting for Sber Investments. Since the 1st of September, the service has already been used by more than four million people. We see that transfers to Sber exceed transfers out of Sber. Third area, AI. More than 12 million clients use our AI-based solutions every month. The expense assistant helps to find opportunities for savings, investment services, analyze the portfolio, and make a more informed decision.
In 2026, the tax assistant has helped clients recover more than RUB 27 billion. We see the potential of AI in personalizing services and increasing their efficiency. The next step is an AI-powered personal assistant capable of executing everyday tasks and supporting the achievement of long-term life goals. The person determines the goal and makes the key decisions, and the assistant takes over the analysis and selection of options and routine operations. For example, in some financial use cases, the assistant will be able to link personal financial capabilities, commitments, and long-term goals into a single financial plan. As technology evolves, it will be able, with client permission, to connect the necessary services and perform the agreed-upon actions.
Good afternoon. I am pleased to share our results across the technology areas of the strategy. We've come a long way in this time. I will identify three systemic results. We maintained our constant focus on the independence and reliability of the IT landscape. We completed the key migrations to our own technologies and increased the stability of our systems. We also took our own GigaChat models to a new level. They have learned to reason and act. The most noticeable application solutions are being implemented by Sber's agents and processes. We built new products based on the same technological base, and they're already being used by our clients.
Let me describe each result in order. I'll start with the foundation. Almost 100% of databases and server operating systems of the bank are now our own services and databases. All critical information infrastructure systems have been migrated to domestic solutions. Migrations of this scale usually mean a risk to stability. We did it the other way around. Over three years, the volume of changes in the bank's IT landscape increased by more than 2.5 times to 6 million per year, while the number of critical incidents decreased almost 2.5 times.
This was achieved through systemic leadership and the development of digital twin. Reliability is built into the landscape itself. Mandatory requirements apply to systems and agents, and incidents are resolved automatically. This is the base underpinning everything I'm going to talk about next. The second big topic is proprietary models. Over these three years, we released three generations of the GigaChat model. GigaChat 3.5 Reasoning was the first model that we taught to reason with reinforcement learning.
The model tries to solve problems itself, receives an appraisal, and learns from it. Today, our models not only respond but also act, search for relevant information, work with external tools and systems, and are trained in Russia from scratch. We are independent of external vendors and can guarantee data security. We're also developing our own models for generating images and videos. In October, a new generation of the Kandinsky 6.0 model was released. Kandinsky is among the best open-source models in the world, evidenced by independent benchmarks.
The assistant is also based on these models, our GigaChat assistant, and it is a fully-fledged assistant integrated into many of Sber's products, including the bank's mobile app for everyday tasks. More than 20 million people regularly use it. Schoolchildren and students are among the most active. With GigaChat, they can understand, they can figure out complex topics and do their homework. The third big thing is AI applications and agents. We started with ourselves.
Within Sber, we launched more than 1,100 AI-based initiatives. Each of them is aimed at a real economic effect. The initiatives span over all key functions of the bank, from the contact center and sales to procurement, finance, and legal support. So that this many directions do not turn into chaos, we are building a single environment for all the bank's agents. Agents are assembled from common components, skills, knowledge, memory. They are executed in a single secure environment under quality and risk control, unified quality and risk control. This makes the creation of agents faster and cheaper, and their operation safer. This is the answer to the main risk of mass adoption of AI and unmanaged technical debt. We are scaling agents without losing control.
The development cycle itself is changing. The engineer sets a task and sets implementation results criteria and an execution plan. A system of agents takes up the checking responsibility. All our engineers use agent development, and more than 50% of the new code in Sber is created with AI. Over the past three years, we achieved significant results. By the end of 2026, we will implement half a million changes in products. This is double the improvements of 2023.
And the lead time from setting a task to implementation decreased by 40%. This means we are developing banking products faster, we are introducing new opportunities for clients faster, and we are doing more without detriment to growth in resources. Everything we have built for ourselves, we offer to the market based on the model we have compiled a single proposal for different roles and tasks within the company. We start with personal experience. An agent is an autonomous agent for managers and specialists who connect to the mail and calendar, remembers the personal context, and suggests a solution.
And when you need the context of the entire company, you get the help of a GigaCowork platform where employees work together with agents and each other in a single loop. GigaCode is a tool with which the developer writes code together with the agent, and business users create applications without programming. These three products form the core of our offering to businesses. The offer has already found demand on the market. The agent, GigaAgent, has already become a common tool for our customers. A typical scenario is, a typical use case is an agent prepares a project report analyzing deviations, collects meeting materials, or checks counterparties, and does it on schedule without reminders.
We launched GigaCowork in July this year. In the first month, more than 1,000 companies connected to it. It works with conventional business systems, and more than 40 ready-made connections to corporate software and external sources are available from 1S to the SAP database. Among pilot implementations of the country's largest industrial company, there are tasks ranging from document flow to financial reporting consolidation to geological data analysis and production planning and Sber's development tools, including GigaCode and GitVerse, are now used by more than 30% of Russian developers.
Today, AI works not only in the digital world, but also in the physical one. To close the autonomy cycle of the company, we are developing physical agents as well. We created our own anthropomorphic robot named Green. Its brain is a physical AI model. Green-VLA is now top three at the main competitions in the field of physical AI at the International Conference on Robotics Autonomous Systems and e arlier this year, we introduced the GreenWAM model. It controls robots, predicts how their actions will change the environment as a result of what they do.
We have already piloted robots in real-world operations, such as a robot for selling goods in stores. According to the Ministry of Labor forecast, by 2030, Russia will need more than 3 million additional workers, and more than half of them are needed where there is physical labor. To be ready for this, physical and artificial intelligence technologies must be created today, and we are doing precisely that based on domestic solutions. Over this time, we have accumulated a great deal of expertise within the bank on how to develop models, how to create agents, and how to work with them, how to build the work itself in a new way.
And right now, we are continuing to explore the new architecture of models and collaborative work of many agents on one task. We are actively transferring our experience to the market. We made our Kandinsky and Green-VLA models publicly available under MIT License. They can be used free of charge, including in commercial products. We will talk about the latest achievements of Sber and more at the AI Journey Conference. Last year, it was viewed in more than 190 countries, and we invite you to participate in it this time. Today, we have a reliable landscape of proprietary models and products that businesses choose. We made it, and we are going on.
We are not looking at the rear side mirror, but we are looking at the windscreen. In the last several years, the AI went ahead. Now you can give assigned tasks to agents and get results in our life and in our businesses. The complexity of the task that the agents resolve are getting more and more complicated and this is a new high-tech revolution, not a hype, and it has already started. All great revolutions have the same logic. I think this technology removes one deficit and immediately creates the next digital age.
The digital era made the revolution instant and quick, and the deficit, the lack is neither in data or reflection, expertise, or coordination and ty his is the deficit that artificial intelligence removes, creating the prerequisites for new agent economy. What does this change for every one of us? Today, a person searches, compares, switches between dozens of services, and verifies execution themselves. We spend most of our lives on this digital chaos, sometimes forgetting what it is all about. We are tired of being the operators of our own lives.
The agent economy returns the most important resource, time and freedom, and creates the conditions for growth of our own agency. The person sets a goal, determines what results they want to get, and all the other routine tasks like search and comparisons and control is done by the agent and t his is the essence of our human-centric vision. We help people to realize and achieve their true goals by expanding their human potential. What does this mean for businesses? Agent economy is the economy of results. The new and most expensive deficit in it is trust and responsibility.
Sber is also responsible for this result, has always been responsible for this result. It was a historical strong position, the trust of the financial capital, decades of competencies in risk management, and this gives us the right and responsibility to claim the role for the driver of the new economy. This is our key objective for the next three years. We have identified three priorities on which we will build the Sber of the future. The first one is an AI assistant, your personal navigator, who will help you solve problems in your daily life based on your interests and long-term aspirations.
The assistant must take over 90% of the organization's execution. You're just saying what you want to accomplish, and that's it. The second is the environment of trusted inter-agent interaction. You know, to entrust agents with serious tasks such as transferring money, signing contracts, verifying management of assets. People and businesses require a guarantee. Who is going to be accountable if something goes wrong? We need mechanisms for verification of permissions, reliable calculations of financial protection in case of errors, and that's what we know how to do, assess risks and take accountability.
We, as Sber, intend to become the main provider of these protection services in this new economy. Third, some ecosystem of the best products and services. The smartest assistant is useless if solutions cannot be implemented in real life. Our products will be selected not only by people, but also by agents, including those created by other companies and f or them, we too must be the best choice number one. Therefore, we need to win in terms of terms and conditions and quality and 100% reliability of execution. Our APIs will become standard for machines. These three priorities are inseparable.
The assistant helps to formulate your dream. Financial protection gives you the courage to delegate it. The ecosystem of the best products and services turn this all into a reality. For this new world to become possible, Sber itself must change. We need two basic conditions. The first one is the technological base. The race for AI leadership is a matter of national sovereignty and our survival. We need the best AI models, and we also need to have an error-free, failure-free IT landscape that includes a great cybersecurity, the leading cybersecurity.
The second one is the team. On the one hand, we must continue to improve and protect our today's business. On the other, we must create fundamentally new businesses at the highest possible speed, and it requires each of us to double our efforts. What does this strategy mean for those who have invested in us? We do not select between the future and the present. We intend to combine high returns with aggressive investments in the future, maintaining leadership in the current markets and explosively and disruptively create new sources of revenue. ROE at least 21% annually.
Capital adequacy at least 13.3%. Dividend policy remains unchanged. 50% of profit under IFRS is paid out as dividends. The financial model of Sber 2027-2029. Sber's ambitious transformation goals based on a sovereign AI model require us to have impeccable financial discipline, system-wide risk management, and capital management.
This is the approach that will help to pave the way for long-term sustainability of Sber's business model. Over the next three years, we expect a moderate economic dynamics with an average annual GDP growth rate of 1.5%. We also expect inflation to slow down next year, with its stabilization close to the regulator's target in 2028-2029. This forecast will allow the central bank to lower the key rate below 10% already in 2028. The regulator's cautious approach to easing of the monetary policy, but rather strict requirements in terms of the prudential regulation of the banking sector are a limitation for the acceleration of the credit activity.
So we expect a stable trend in the corporate lending segment, an increase in the range of 10%-11%, and a slight recovery of the rates in retail lending segment to 9%-12% in the next three years. The growth will be mostly observed in secured lending, in mortgage, car loans for the next three years. As for the funds raised from individual customers, we forecast the annual growth of the sector by 11%-13% on the background of the gradual decrease in rates. The growth of funds of corporate accounts in 2027 will be considerably less than that of retail, but there will be a recovery up to 9%-11% in the next two years as business activity in this market is growing.
Sber will grow on the par with the sector. External conditions will result in moderate pressure on the business margin. The net interest margin will decline gradually throughout the period, but will remain above 5.5% by the end of 2029. In fees and commission business, the revenue dynamics will reflect our previous decisions, including eliminating one-off commissions and growing investments in the client loyalty. Therefore, the dynamics of the net fee and commission income will not exceed the inflation rate for the next three years. The most challenging year will be the next one, 2027.
The financial results of other activities and businesses, including non-financial businesses, will reach the positive area or the positive zone and, in the aggregate, contribute significantly to Sber's results over the next three years. As regards the cost of risk, we expect it to normalize to about 1% by the end of the forecast period. The Corporate segment, at the same time, will be under great pressure, especially in the next year. The cost of risk for retail loans will remain stable. Some improvements are possible thanks to the change in the portfolio structure in favor of secured loans.
An important part of the strategy for expanding is the use of AI. We are now not only talking about cost cuts, but also about creating new sources of income. We see the possibilities in two areas. First is the transformation of traditional financial products towards greater personalization, and secondly, the building our AI-driven corporate solution for our partners. As the world experience shows, the development of this technology is impossible without significant investments. The scale for us will be driven by the dynamics of the equipment prices and the need to maintain a high level of reliability of systems and processes.
The accompanying increase in depreciation will put some pressure on the ratio of operating costs to income. At the same time, traditional cost elements, including salaries and several others, are expected to grow at a much slower rate than in the previous periods, thanks to annual efficiency gains from AI technology. The constant focus on operating efficiency will help us to keep cost-to-income ratio within the range of 30%-35%. A separate challenge for the entire sector will be the announced tightening of the capital adequacy requirements for banks, including going back to the buffer of the conservation, the plans to increase the systemic importance buffer, and the introduction of increased capital requirements for investments in a number of assets.
We take into account the approved requirements of the Bank of Russia for buffers, taking into account the schedule of their introduction, and under this environment, we do not change the target capital adequacy level of N20 at the level of 13.3%. So the combination of external conditions and regulatory challenges for the group's strategic business development goals require us to achieve an ambitious goal for ROE at the level of at least 21% per year. We are talking about our new 2029 strategy horizon and it will enable us to pay out 50% of our net profit as dividends to shareholders, and the other half can be invested in long-term sustainability of Sber business model.
Our forecasts for 2027 are on the screen. We expect a gradual recovery of economic activity following the reduction of inflation and some easing of the monetary policy. All financial metrics are in line with our strategy targets for the next three years. We are moving on to the next stage with a strong business and a unique technology base, and a great team that can handle all of these challenges. We are not going to build up your confidence with promises, but with the results that you will feel in your life. Thank you for being with us. Happy birthday, Sber.
And to conclude, I would like to share one thought that should be our common benchmark. A great Russian engineer and thinker, Peter Kapitsa, once said that the means do not determine the goal, but they determine the scale of its achievement. We have all that we need, the capital, the technologies, people, and we know what our goal is, to make each person's life free from routine and e very business in Russia should be free from routine tasks. Our current means are unprecedented, and thus, the scale of our future victories will be the same. Forward to the future.
The Q&A session with Herman Gref. Viewers, we are continuing our Investor Day. We have a dialogue with the CEO and Chairman of the Executive Board, Herman Gref. It is going to be an exciting dialogue and w e have invited some analysts who have been with us for a long time and know our investment case. Olga Naydenova, the Senior Analyst of Sinara Bank, a great expert. Evgeny Kipnis, Senior Analyst of Alfa-Bank. Evgeny is very much interested in AI, which is very relevant to our discussion. Andrei Melashchenko, the Chief Economist of Renaissance Capital. So Andrei has a unique gift. He is a macroeconomist and a great industry expert.
Timofey Martynov, the Co-founder of the Smart-Lab Retail Investment Association. This is someone who is a great influencer among long-term investors and those who have only opened a broker account. Our strategy is published on our website in a special section for the shareholders and investors. If you want to read it, use the QR code, scan the QR code on the screen, enjoy the strategy and read it out. The strategy is about AI that is focused on interests, not on interests really, but on the goals of people. It is very ambitious, and it would be great to look into that in detail. Evgeny is an expert. He has tried the new agent economy. I am sure you have some questions.
Thank you very much. First of all, I would like to congratulate you and your organization with the anniversary. You have been heading the organization for a long time, and I wish you and your company all the success in the coming years. Talking about AI, I think 2026 has become the year of AI for many people. This technology has become basically the extension of their arms, of their hands and w e understand that this is a major disruption and paradigm shift, yet many people are still pretty cautious about AI.
They do not trust all of the decisions. What is the share of decisions made autonomously in Sber, and what will be this share by 2029? Who do you think will be responsible before the regulator, the shareholders, the customers if an AI agent makes a mistake? Where is this red line where the autonomous solutions will not be used? Where is that boundary?
Thank you very much for the questions. I would like to make some corrections to what Anastasia has said. We do not have the goal to implement AI. If we didn't have to do it, I wouldn't do it. Our goal is to build a human-centric organization. That's our goal, a human-centric organization. It is only possible if we fully implement AI technologies and fully embrace it. We do not think that technology is some magic thing or some godlike technology.
We are obsessed with our customers, and everything that is good for our customers, everything that can move our customer experience to a new level is sacred for us. If you need to fully embrace and develop and implement AI for that, we will do that. If we will see that we can cancel AI implementation across the bank, I would do it just as easily as I approve its implementation. You said that 2026 was a major shift, and it is true. I agree with you. We see that in professional forums and workshops, always focus on AI. That's on top of the agenda. Here we're talking about the culture, science, education, social security, all of the other.
Any topic, any industry discusses AI as the number one topic. For us, AI has been on top of the agenda for the last decade. So, answering your question about the autonomous nature of some of the solutions, around 99% of decisions in retail is made by AI. And by the end of the year, our goal was to move 75% of the decisions on corporate clients to AI, but now they are 80% automated. That's the new achievement. And our goal is not to give you these high numbers in AI-driven decisions.
The logic is different. AI-driven, AI-based decision is more rational. It's quicker, it means less losses. The AI is not usually conflicted with various biases when it makes a decision, so it's more rational. By the end of 2029, we think that our organization will be fully autonomous, but it is a manageable autonomous nature. Even as AI grows as quickly as it grows today, we do not think that by 2029 any organization can be fully autonomous. That's the first thing.
The second important point is that the role of the people will grow dramatically. Might sound weird, but that's it, because AI cannot be entrusted with redesigning, rethinking the foundation of the business model. Therefore, anything that we can routinize, and that means run tasks and many change tasks, and I'm talking about coding, developing new products. All of these things will be delegated to AI and gradually, but everything that is related to disruption, innovation, new business models, and learning our customers and learning what needs they have, and delivering totally new capabilities to the customers. Everything that I've just mentioned will be human-centered, human-driven, so the role of the highly qualified talents will gradually grow.
Thank you very much. A question from me, from our large institutional investor. It is related to the news that we get from American and Chinese markets, and one piece of news is that many large players decided to close access of the external agents to their own ecosystems. They were motivated in it with the protection of personalized data and cyber risks, but there is another reason highlighted by analysts, that the companies protect the margin drivers of their business. How will Sber respond to that, and will Sber work with external agents?
All in all, this position reflects both problems faced by the market today. Problem number one is the low level of maturity of technology. No bank today can afford to have third-party agents in their systems, because at the end of the day, the bank is responsible for any mistakes, for keeping money secure. So, this position at the early stages is completely rational, is completely justified. Another thing is that large players will certainly be defending the existing business model, and smaller players will all the time be suggesting ways to disrupt it. So, there will always be this friction between large and small players in the market.
Speaking of trends, the main trend is that at the end of the day, we will have this gradual permeation of agents into the system, the increasing interoperability of agents, b ut before that happens, the technology of multi-agent systems must reach a certain level of maturity. And we also need an adequate environment for it, for transactions to be performed securely. This is why, in our own strategy, we've established this way of transitioning to the agent-based economy, and we believe that trying to go against the trend is completely useless. I mean, you can slow down the trend for a while, but you can't beat it, you can't win over it.
Agent-based economy is already here. It's this new reality that we live in, and short-term, it might give you a lot of benefit, but long-term, unless you change, unless you transform your business model, you will lose. So s trategically speaking, we are focusing on building an organization that will be very thoroughly adapting its business model to the agent-based economy, and that will be making profit, not with the traditional things, the fees and commissions, for example, but with the accuracy, reliability, trust, and the ultimate result.
Our entire strategy is about this preparation for the next three-year period. I don't believe that in just three years, we will have this explosion in the market where the maturity of technology, the maturity of infrastructure will be such that we will be able to make this shift, because we need a lot of other changes, including changes in regulation. But I do believe that three years from now, we will wake up in a new reality, and we have all of the conditions for it, for this mass transition to the new state of the world. Interesting periods lies ahead of us, apparently. Olga?
Yes, thank you. You've been very actively investing in AI technology for your own needs, for Sber's needs, and for the needs of the wider audience of the general public. The efficiency, the effectiveness of investments into your own product offerings, personalization, for example, in this regard, we know how to calculate the efficiency and effectiveness of this investment, but what about the investments into the wider AI applications? I think it's a bit more complicated, it's a bit more difficult to assess the effectiveness and scale of these investments on your strategic horizon. How will you be approaching the investments and assessing their efficiency and effectiveness?
Thank you, Olga, for the question. I wouldn't say it is difficult to assess those things. All of these things are quantitative, they are measurable. You're saying wider AI applications, I think you mean generative AI, right? Unfortunately, generative AI is not breaking even. I mean, this is not just our case. This is the case of everyone, all of the companies around the world. But in our case, the situation is that we started adopting generative AI a long time ago, and our investments, we have returns on those investments. If we hadn't invested in the previous cycle, today we would probably feel very insecure. We would not be as comfortable as we are now.
But all in all, the returns on investment into AI are positive, not as positive as it could be if we behaved like a company that decided to skip over this investment cycle. For example, Apple, they're making great profits. Capitalization-wise, they are number one in the world, and they have not been investing into generative AI. Instead, they are using, lagging very much behind, they are using other parties' artificial intelligence, third-party AI. This is just a different strategy.
I do not know which strategy is better, the big tech strategy, the strategy of Google, Microsoft, and other companies of this caliber, or startups. It is hard to call them startups anymore, like Anthropic and OpenAI, or the opposite, the position of mature companies that skip over this investment cycle and launch it once the technology is mature enough. We have gone for the latter option. We are among the companies that are developing the entire cycle of AI.
One of the reasons for it is that we are the only player in the market developing the entire technological stack related to domestic sovereign models, meaning models for robot control and management, models for video and image generation, a big scope of models that we use for healthcare needs, and of course, LLMs. All in all, over the past three years, with AI adoption, we have made roughly RUB 1.5 trillion so t he positive balance is significant. But in the subsequent years, it will go down, at least in the next three years.
We will have to invest even more aggressively, and the returns, at least in the next two years, will not be growing as exponentially as the expenses, b ut we see that our strategy is not deliverable without these investments. This is where we come to categories that are not as measurable, I would say. The CEO of Google was asked by the company's investors, why are they investing so much into AI? And he said, "I feel that underinvestment would cost us much higher than direct investment of this scope," so we feel the same. It is a feeling, of course. I mean, you cannot put a number on a feeling.
We cannot precisely predict the exact returns on investment into generative AI, b ut in our case, I do not think in the case of Google, they do not really have any alternative. It is a company that, in my view, has done the most for AI development in the world. So they cannot just stand up and leave that market. We cannot leave that market either, because we have been doing this very actively for the past 15 years, and because we are the only ones doing this in this market in Russia. Russia must have its own AI competencies, its own AI.
Another important thing is that at Sber and around Sber, we see a lot of specialists emerging who know the ins and outs of this technology. We see more and more researchers. We have our own research groups, very powerful ones, and AI research institute, our lab that works on AI adoption. Another thing is that AI is about security today, cybersecurity, combating fraudsters that use cutting-edge models, the most advanced models of AI. This is a very big, very serious challenge after Mythos appeared in the market.
As we have seen, the world has changed quite significantly. Maybe not everyone has seen it quite yet, but we do. Without sophisticated skills and competencies, it is very, very hard to ensure reliability of your own business and secure the interests and funds of our clients. All of this taken together means that we will continue to invest heavily in AI, because we do not see an alternative.
A follow-up question, if I may. In principle, technological development, as discussed earlier today, helps optimize a lot of bank processes, but it requires new competencies, pretty expensive competencies at that. What is your take on Sber's team by 2029 in terms of its headcount and its structure?
It is hard for me to predict these things right now because last year we disrupted this trend. We transitioned from exponential growth of our tech team to reduction of headcount, specifically people engaged in run activities. Meanwhile, the engineering part had been growing heavily, and we cut 20% of the engineering team, but w e see, nevertheless, that we will have to increase the amount of resources involved in certain business sectors. For example, we have recently made this decision to significantly reinforce our team that works on Mythos. This program is called Mythos-r eady.
We had to reinforce this team that works on our own proprietary attacking models because we need to transition from defense to offense, to attack ourselves primarily, because otherwise you will not be able to figure out whether you have inside any vulnerabilities, any fraudsters already. You have to be able to as good at attacking as fraudsters. And of course, this has a lot to do with the modern IT landscape. I mean, the fingers of one hand would be enough to count all of the companies in the world that have this cutting-edge modern landscape, modern platforms and i n our case, it is very difficult to guarantee 100% cybersecurity, even with our infrastructure.
So, we have had to reinforce the team that works on ensuring this security. Engineers that do the transformation of the platform and the engineering team that look for vulnerabilities and develop state-of-the-art models to prevent the exploitation of such vulnerabilities have been reinforced. This also includes fast patching, as we have seen. If there is a vulnerability, exploits are published within a second, and the patch is ready a couple of months from then. That is how we live.
The entire world has been functioning like this for a while now. The average term for delivering patches at companies such as Microsoft and others used to be one or two months. Now it is impossible. The zero-day vulnerabilities, there are so many of those, and this is impossible in principle. You need to change the entire philosophy of how you work with vulnerabilities, with exploits. This means complete rethinking, complete revisioning. Only people, only human beings can do this at this point, but I mean, not just human beings, but human beings equipped with state-of-the-art AI.
So, we will be growing in terms of headcount for physical AI. In particular, our models have been performing pretty well regarding international benchmarks this year in China. Our robot control model came second at the international competition in China, where the entire world was represented. So, regarding physical AI, I think we will be growing than anything that has to do with state-of-the-art world models that we are working on right now. This will, again, require more people on board. More than 80% of code is already being written by agents.
We are creating a huge amount of requests for change, many, many millions a year. I mean, at some point we saw 20,000, I think, and we were happy with that. Today, it is millions per year, and the speed is only increasing. So, with routine work, the number of engineers will be going down. As for these new areas I identified, the intelligence-intensive areas, they will only be growing. Thank you very much. I suggest we segue from AI and AI-related challenges to our separate business areas. Retail business. Andrei, over to you.
Yes, I wanted to discuss the topic of demography, the current trends, including the reduction in headcount. Potentially, it can be a challenge for the economy, for Sber, and Sber has been dealing with this through increasing efficiency and productivity, offering solutions for its clients as well. But the demographic trend is also a challenge in terms of client base. Over the past years, the pace of growth in the number of retail customers have gone down.
I believe that the clients that grow up with Sber consume fewer long-term consumer goods, and maybe they have different needs in terms of improving their living conditions, but they are active at savings, and they may be potential clients for wealth management and the funding base. So, what are the demographic trends that you have embedded in your strategy? How have your client needs been changing, including the needs for key banking products, and how have you been responding? How have you been dealing with these challenges so far?
Thank you very much, Andrei, for those questions. We've long believed that the aging of the population in terms of client base is a problem, and at some point, Sber's strategy was rejuvenating its client base and transitioning from older clients to younger clients. But then at some point, we saw that this strategy is completely incorrect. We started to fight for pensioners as our clients. Well, we didn't really have to make much effort because no one considered them as a lucrative segment. So, we acquired roughly 60% of the client base in terms of pensioners, senior citizens, and then it turned out that this is the most profit-making part.
This is the most profit-making segment. These are the people who have money. They have been working all their lives, and by the time they retire, in terms of deciles of age groups, the wealthiest group is people aged 60 or older, and I do not agree that they are reducing their consumption. It is just that the structure of their consumption is different, but they are not decreasing it. Maybe they don't need a new flat or a new car for themselves, but they buy it for their children and grandchildren.
And these are the people who spend much more on their lifestyle and maintaining their health. So, in this regard, it is a golden age group for any business if the business can keep up with this shift in the consumption structure and make relevant offerings. It's very often the case that they may not necessarily spend that much time looking for goods and services, but if you learn how to work with this, if you learn to make relevant offers for their lifestyle, for their life needs, in terms of client base, I would say this is the best, the most lucrative, and the most profit-making segments.
So we are seeing this shift, 1 percentage point per year, our penetration in this segment of the market, of the senior citizens, is growing, and our profits from this group of customers are growing as well. We have to improve the customer experience because right now, there is still room for improvement there, and our services are now We start to tailor our services for the interests of this group of the population, and we were doing the things that we have never done before.
We are investing in the research in longevity, and we are looking at the most advanced methods, longevity methods, and we are confident that on this market, there are major disruptions, and in the nearest years, they will be more pronounced. Making your life longer and making your healthier life longer, everything related to anti-aging technologies and beauty technologies, I think they are very promising. This shift is We are going to witness this shift in our country, in Russia, but in terms of the profits, our profits from this segment will not be negative. It will be, I think it is likely to be positive.
Thank you. Thank you very much. We have talked about the retail. Let us talk about the Corporate. I think that the state of the corporate customers is of most concern, and your forecast means that there will be still growth in the Corporate segment, and the cost of risk is dwindling a little bit. This forecast, in what way it is related to the macroeconomic forecast, including the output forecast? In what way is it sensitive to the macroeconomic forecast parameters? How do you think the situation might unfold if the decrease of the key rate is much slower than expected?
Well, this is something that we focus on in our research, and this is on top of the agenda when we have sometimes arguments with some analysts and our regulator. We think a major deterioration in the corporate sector in the second half of this year, and this trend will continue in 2027. We face a major geopolitical challenge. Every day, some industries are losing money. Maybe it is not critical, yet companies have to spend their resources to restore some of their production facilities.
We see the increased tax burden and the positive key rate, we see it at around 8%-10%. All of this combined cannot say that we have great outlook and a pretty strong RUB on top of that. With this combination of factors, it is obvious that the economy will face a challenge. Therefore, we have created additional provisions in some of the sectors beforehand, including the offline trade, housing development loans, coal, wood processing and w e are entering 2027 with a pretty good safety margin and I think that in these sectors we are going to see some problems next year, some difficulties. We do not think that next year the cost of risk will go down.
I think the cost of risk will go up. In our view, 2027 will be the peak. If the trends remain and if the rate goes down gradually, by 2027, the situation will normalize. I think that the cost of risk will be around 1%. The end of this year and next year are going to be challenging periods for the banks and for the customers, for the clients in the sectors that are mostly affected by this challenging macroeconomic situation. So 21% of ROE that you promise within the horizon of the entire strategy sounds pretty ambitious. It is pretty ambitious, but overall, we will do everything to achieve these parameters.
Thank you very much. Timofey, I would like to make you join the discussion. Thank you. I will step back from the strategy a little bit. Sberbank, I reiterate, is a pretty predictable organization. When I was going to this event, I asked my analysts in my team, "What do you think about the future of Sberbank?" They said, "Well, it is pretty predictable. It is pretty clear." The things that are not clear are the things that you cannot control, like the external environment. One of the major problems there in the external conditions is the situation with the budget and the fact that Sberbank is a profit-making company, makes a lot of profits.
How can you fight back against Ministry of Finance? How do you fight back against some ideas that Sberbank should pay some additional taxes on super profits?
Well, to have the tax on excessive profits, you need to have excessive profits. Many people do not quite understand the nature of the business and how businesses make money. The returns of the bank and businesses, its profitability, the ROE is 21%. That is what we have discussed. If we look at ROE across all the sectors, even with our ROE, we will not be even in the top 10 of the industries because there are some industries that have an ROE of like 70% or more.
In this regard, in our case, we obviously, like in terms of the overall volume of the profits, we are the largest organization in the country, but it is mostly related to our market position, to the scale of the financial sector. This sector in Russia can be compared with only oil and gas sector. In our case, we are talking about the efficient expense management, cost management, having cost-to-income ratio at the level of 30%. I do not think you will find many organizations around the world that can manage their costs so efficiently.
This is thanks to the fact that we invest a lot in technologies, and technologies are this safety net that we have. Since we invest a lot in technologies, in the next couple of years, we will have a depreciation problem that will accentuate We see this problem among large big tech corporations, and we start to feel that. In the next three years, the depreciation will accelerate because there is a lot of investments in AI, in data centers. This equipment will depreciate, and it will affect the company's balance. We have the full understanding with the Ministry of Finance.
I think the fact that the Minister of Finance is the Head of Supervisory Board, chairman of Supervisory Board, is of great help. He understands the company from inside out. He understands this is the economy. At the same time, he understands our philosophy. And our philosophy, our approach is that we have to pay all the taxes, and we have to maximize the dividend payouts to the extent as possible to make sure that we meet all the obligations before the state and before the shareholders, 50% of dividend payout.
We are a boring company. We try to be a boring company in terms of the financial results and in terms of our products and technologies. We try to be the company that surprises its customers. In terms of the financial results, we are the largest company in Russia. The best that they can offer to their investors is, that we can offer to our investors is the stability of the results, so w e try to stay committed to this principle. We try to maybe promise a little bit less to our shareholders that we can deliver, but we should never promise something that we cannot deliver. We cannot fool our customers and our investors.
Capital adequacy ratio, 13.3. Our dividends is the tax for these successive profits because the state is the largest shareholder, and everything that is larger than 13.3 is taken away. There is no company in Russia that would pay the same amount in terms of the percent from the profit and in terms of the volume of the dividend for such a long time, such a stable period of time.
We are not going to pay less dividends. You are talking about long term?
No, I am talking about year- by- year. If you take everything that is more than 13.3, which is the minimal capital adequacy ratio, any other additional taxation would lead to the fact that we will start paying less dividends. So in this choice, it is pretty clear that for the state, it is much better to make us stay at the same level of contribution to the tax system so that we can pay as much dividends as we can, as we do right now. In terms of the minority shareholders, I mean, that is some great news for them because they are together with the Ministry of Finance, which is the majority shareholder, and they are interested, both of them, in the volume of dividends paid out, and it is a guarantee that the dividends will be paid out.
But again, this is not limited to the management position. I mean, actually, the top management team thinks it is pretty normal. It is business as usual. We should give as much capital as we can. The capital should bring money to the shareholders, and it is a great honor for us to maintain the same level and volume of dividends paid out. We will do everything we can to make sure that we pay as much as we can a nd the fact that our shareholders love us, we need to make sure that we cherish this trust, this love, especially among the retail shareholders, because it is very easy to lose the confidence, but it is very hard to build it. It is not a figure of speech.
The obsession with our investors and customers is the basis of the culture of the business, the corporate governance. The Executive Board should be responsible for the operating management of the company. Others should be responsible for the strategy and being able to adjust the strategy and making sure that the top management team is executing this strategy, and this is the role of the board of directors. I do value a lot our cooperation with the board of directors. We are very open, we are very sincere about every line of the business.
Any director can request any information, any data about any business, and we will provide all the data, all the information that we have, because trust is one of the cornerstones of great cooperation, of great work, and we have zero businesses that would be close, that will be unseen by our directors, by the owners of the company and the shareholders. We support this environment that we have the roles that are delineated. We have to be fully accountable to Supervisory Board and we are committed to that.
We are open, we are sincere. We show all the problems to them. We have not concealed anything. We have a special audit committee, and they have many hours of meetings, so w e have a fantastic audit and control system. We have guys who built fantastic system to detect possible deviations from the risk metrics. This service is accountable to Supervisory Board. the s upervisory Board is approving its plan and listens to the at every meeting is reviewing the results. It is a great team, and we do not conceal our problems.
This line of trust coming from both sides is very important, and it keeps us up to shape. We have been discussing the strategy, and we made three sessions for Supervisory Board. the Strategy Committee had a several-hour meeting when we were discussing our strategy, and then Supervisory Board approved it. There were no questions on Supervisory Board because we had a series of consultations before that, and every member of the Board had the opportunity, after reviewing the text, to get a management team to ask any questions and to give any comments that they need to voice. We are working with Supervisory Board team very closely.
Supervisory Board members are very interested, are very valuable. We are very attentive to their criticism. We try to provoke their criticism because the shareholders and Supervisory Board can criticize us. That is easy, that is cheap. Supervisory Board members can criticize us, it is free for us. Shareholders, it is a problem. But the customers, when they criticize us, that is the most costly criticism, and we have to make sure that we minimize it. We have to talk about macroprudential regulation. We see that there is a number of regulatory initiatives, the differential add-on on the system-wide importance, the weighted risk for the corporate and retail businesses. Basically, these are the measures that are asymmetrically affecting the sector, and they put Sberbank and state-owned banks and the largest private banks in the less privileged position if you compare them to smaller banks.
At the same time, we understand that without the largest banks, the system would not have enough resources to meet the needs of the economic players on loans, and we might face a situation that when the loans might not become costlier, but it might not become cheaper in line with the gradual decrease of the key rate, so w hat should change in the economy to make sure that the Bank of Russia approach to regulation changes?
First of all, we determined the following: the buffer of 1.3% to N20.0 is something, is the buffer, the add-on that we will be committed to. I think it is irresponsible to go lesser than this level. We have discussed it with the members of Supervisory Board, and the updated dividend policy is actually focusing on that. We are going to pay 50% apart from the cases when we do not have enough funds, enough capital to support this add-on.
What would be the add-ons?
We cannot predict that. So 1.3 is something that we are going to be committed to. In the next few years, the ROE will go down. We say the ROE is 21%, the ability to generate capital to make sure that we pay dividend payouts in the volume of 50% of net profit, and at the same time keeping the minimal level of the capital adequacy ratio. Why I am talking about the minimal CAR? European banks is like 17%, Chinese banks 19%. We are the largest bank. Our capital adequacy ratio is 13.3%. At the same time, we are the mostly capitalized banks in Russia. We are seeing that the other largest banks are very much limited in terms of their business capabilities. They are moving out of the market. They do not have enough capital.
It is amid these terms, the regulatory requirements. If the regulatory requirements for capital adequacy increase, we will have to reduce the amount of dividends paid. Obviously, we would have to. Our market position, as you said, will not be infringed upon because all the other players would feel much worse in this situation. The more efficient one will gain a competitive edge. If it is a way to combat us, this is the worst possible way to come up with in the current climate.
If our competitors' capital adequacy would be around 15% or higher, then yes, this would be doable, b ut in this case, we will turn out with a certain market bonus. So, I am not concerned about our position in this regard. I mean, we need to expect a lot of risks, but we are one of the most efficient players in the market, one of the most high-tech players in the market, so we will certainly deal with tighter regulation. As for the rest of the banking industry, well, the regulator will have to think long and hard before they go for such measures.
Thank you. One more question, if I may. We see in the banking sector this trend to expand ecosystems, meaning banks are developing organically and non-organically. They are expanding their ecosystem to meet as many client needs as possible on different platforms, but the Bank of Russia is de-incentivizing investments in such activities. Historically, the non-financial business of Sber has performed negatively, but your next three-year strategy provides for positive results for profit-making of such businesses. I mean, where will be this intersection between ecosystem expansion and tighter regulation, and what are your plans on making your non-banking, non-financial businesses even more profitable?
Well, they have quite tight norms. They have quite stringent requirements for such activities, for such risk-sensitive limits. But you know, of course, the introduction of sanctions put a lot of pressure. It hit us very hard in terms of expanding our ecosystem, and we are investing quite little in those things. I mean, we are talking about just very, very targeted, focused acquisitions. We will carry on with our strategy, with our philosophy, but we will be abandoning the business areas that are not generating mid-term any ROE. We will only stay, we will only maintain the areas that are of strategic value for us, the areas that hold strategic importance.
We have started doing this earlier this year, and in the next two years, 2027 and 2028, hopefully, we will be able to put these things in order to make sure that our expectations can be delivered by 2029. You will see, I mean, you have seen this year that we have significantly revised our e-com model. We used to have a lot of warehouses generating loss for us, but the situation with our warehouses helped in our situation. By the end of this year, I think 100% of unnecessary warehouses will have been written off, sold to marketplaces. So, this used to be a big source of loss in e-commerce, and the same goes for other areas.
Of course, the use of generative AI is really helping with entry into a lot of other markets. I think the time of heavy load investments into the ecosystem is behind us, and we do not plan to continue. The cloud-based business, with its exponential growth, this is where we will be solidifying our position, because we need it for our business, we need it for the businesses of our clients, and of course, we want to be choice number one in terms of AI-based service provider for our clients. Anything that has to do with that, not that many acquisitions required there, but we are talking more about accelerated growth and building new data centers, and maybe acquiring a few players from the market.
Thank you. Timofey, you had a question?
Yes. As a shareholder of Sberbank, it seemed, I think I heard a number of things, a couple of things that I would like to ask you to clarify. You've just said that if the Bank of Russia raises its norms, its regulations, you said that we will have to bring down the dividend payouts. It's just that when you said it may sound a bit alarming for those in the audience watching us. I just wanted to clarify, this is not the baseline scenario, right? The baseline scenario is 50% of your profits for the next 30 years paid as dividends and t his is a highly likely scenario, right?
Yes, you're absolutely right in asking for that clarification. Thank you for drawing my attention to it. Look, we are saying that it is the only necessary prerequisite for it, but not the only one possible. I mean, if the requirements go up, we have this one, two, three barrier, this add-on, to go on top of the add-ons that will occur. Even if this add-on occurs and the profit generation by us is enough to deal with this one and three threshold, we will still maintain the 50% payout. The only scenario we have to give it up is where we have this increase and we fail to form the buffer. So, two conditions. But it's not like an automatic switch, right? These two conditions have to converge. They have to coincide. The reliability of the bank is more important than anything else.
Thank you. And one more clarification. It sounds, once again, a bit alarming. In 2027, you said we will have a peak in terms of risk. The trend is clear, but still, I just wanted you to maybe give some hope in terms of how limited this problem will be, the corporate sector in 2027, because it sounds a bit concerning when it's just there, when it's said. Will there be a crisis? I just wanted to ask you about the possible ceiling, the cap of that problem.
So, speaking of Sber, we have included it into our strategy already. As I said, this year, we have had accelerated growth in terms of provisions for the sectors where we see potential problems next year, so that we're not faced with anything unexpected. I want us to remain boring and conservative and predictable in terms of financial results. So we try to foresee such things, and if those risks don't materialize, it will be a pleasant surprise, right? But the likelihood of that happening is pretty high.
So this is why, for the sake of our financial results, we will be adhering to the strategy, and hopefully there will be no surprises for you in terms of financial performance. But yes, this problem is likely if the parameters that we observe today persist. I mean, in terms of regulation. Its impact on the rest of the banking sector is hard to predict today because, in any case, the situation is not going to be very comfortable for the bank next year. You have to analyze it case by case. Some of the banks have good capital buffers. They have good provisions because they are preventive, they are proactive. Some of the banks will struggle. They will have to optimize their business to sustain capital adequacy with an increased share of resources that they will have to accumulate to form provisions.
Thank you very much, Mr. Gref. One more question. I just have to ask you this. From the Shareholders Club of Sberbank, almost 1 million members in the club. So we are about to go on holiday around New Year. It is time to reset, to prepare for the new stage in life, and draw inspiration somewhere. What would your advice be for the shareholders? Watch something, listen to something, maybe read something?
Let me put it this way. If you want to watch something, we have been sponsoring, together with the Cinema Fund, this wonderful movie called Arsenyev. It will be available on our Okko platform. It is a TV series, eight episodes about the Russian Far East, and an absolutely wonderful, glorious, I think you can call him a statesman, called Vladimir Arsenyev. I am sure many of you read it, read this book in your childhood. I did. I know I did.
I devoured that book called "Dersu Uzala." Yevgeny Mironov, the Russian actor, lived, literally, he lived in the Far East for almost three years. The quality of the series is such that each episode is like a full-fledged full feature. I think it is really, it does deserve your attention because it will give you a much better idea of how our country was developing in its Far East, in its attempts to sell the Far East to other countries, including Alaska.
Music, for me, a lot of inspiration comes from music.
That is true. I do not think I can recommend anything new. We will be following your playlist then. Yes, classical music and a good sound system.
What you can read?
You know, I have this interesting young writer, author on my radar, or rather a psychologist. Let me look up his book. It is a very, very recent one. I have just read it myself. Let me look it up. Let me look up the title. It is called, if you translate it roughly, it is someone who is in distress, a person in distress. Evgeny Latypov wrote it. Also, I really love this book by Robert Sapolsky. It is another of his books about the freedom of will, and it is called something along the lines of, "It is All Decided," or "It is All Predetermined." "Determined: A Science of Life Without Free Will." It will be very useful, including in your work with AI.
Thank you very much. Thank you, dear audience. Thank you for watching us. Some final remarks, a final message for the audience, Mr. Gref? I do not really like all those final messages. I think you should stay optimistic. Stick to your optimism.
I believe in the power of human mind, and I think everything is going to be fine. 2027 is going to bring a lot of clarity, I think, to our lives. People deserve to live better, so let's try to be optimistic. We will be able to accomplish everything.
Thank you very much. Once again, I would like to remind our audience that we have prepared a small surprise for you. It is a video based on an actual story, based on real events, reported to us by our employees, a video about the first steps of a baby with the support of Sber. Thank you.
Can you wait for me here, please? I'll be right back.
I'll do it myself. Mortgages, loans, real estate transactions.
Hello.
Hello. Do you have an adult with you?
I have a question of my own. I want to ask you a very serious question.
Okay. How may I call you?
My name is Fyodor.
Fyodor, what is your question?
I need a loan.
A loan for what?
I want to buy a scooter.
A loan for what amount? I have it here. Right. What is your source of income?
So, RUB 100 for doing the dishes, RUB 50 for taking out the trash. I also have this as collateral.
Why don't you just ask your mom?
I want to do it myself. Mom said that I should learn to be independent.
Okay. We will consider your application, but please note that you will need a surety.
I can be the surety. Do you know the borrower? Do you trust him? Eight years and six months that I've been trusting him.
Oh, you have saved RUB 4,000?
Yes, I did it myself.
Okay. You will give half, and your dad and I will add another half. Great. And for the next goal, you can save money with your own card in your name.
So, after the scooter, I will be saving money for a bike. I support that. Deposit.
This video was created based on real events, and this story reported by our employee, Kristina Isaeva. Sber, helping even the littlest ones in achieving their dreams. Thank you for your attention.