About the start
I came to Sberbank in part thanks to the community of graduates of the International Institute for Economics and Finances of the Higher School of Economics. One of my classmates recommended me for the job.
Sberbank is my first job in the banking sector. Before that I worked at the company Lamoda. At Lamoda I learned SQL for working with databases, which really helped me in my job interview at Sberbank. They asked very specific questions, such as “Name the function that finds the first non-empty value”. I named it right away, and I think that played a role in getting hired.
I was always interested in applied mathematics and information sciences, but I wanted to apply mathematical models to social and economic processes, not physics. The field is open for discoveries – invent a model that explains the real behavior of people.
About the treasury and why there are no breakdowns
My department handles risks at the aggregate level, that is, we’re talking about the entire balance of Sberbank. We have to manage assets and liabilities so that there is never a situation when the bank can’t cover its obligations. In professional terminology, this is called asset liability management (ALM) and liquidity risk. We need to keep an eye on how the departments that attract clients and the departments that make loans are performing so that we can match data and understand what terms, interest rates and instruments we need to be profitable.
I am personally responsible for making sure that before 6 p.m. the entire Sberbank group receives the interest rates in eight currencies for their products for the next day. This has to be done no matter what – fires, floods, the servers crash, discs that don’t work – none of that matters. In addition to that, we handle everything concerning internal funding. To put it more simply, we check on how many deposits the new client department took in and how much credit the loan departments gave out, in order to manage the margins. If we see some problems or a way to improve the situation, then we make changes to the funding methods. It’s a rather flexible process because the bank products have a lot of options, and our task is to find the most profitable ratio.
About work instruments
We work with the SQL system for databases, Teradata, Excel and, of course, PowerPoint. After we do all the calculations and construct our models, we present our conceptual ideas to our colleagues. That’s why we absolutely need to be able to do presentations. I work with SQL every day to carry out my day-to-day responsibilities, although it’s not so much a programing language as a data query language. There is nothing the least bit complicated about it. Last year when we did an inventory of macros, we calculated that our department had developed more than 100 macros, which is quite impressive. We have extremely powerful computers with the main programming languages – as I mentioned, Python, R, SQL, Teradata. Sberbank understands that the potential of big data is huge, and there is great interest in artificial intelligence and machine learning.
About analyzing client behavior
I like client analysis. I like to see how people and organizations behave. I juxtapose their behavior with events in the market and in the bank and make hypotheses about why they behave as they do. And I back all my hypotheses up with figures. For example, why do clients suddenly put more money into long-term deposits, or why do VIP clients deposit much larger sums into renewable accounts than the other client segments do? Analyzing client motivation, finding intuitive explanations for the behavior and insights in our data is a fascinating process.
About creativity at work
A great deal must be invented from scratch. I’ve been working here for two and a half years, and in this short period of time many projects have begun that have involved my department and other departments. It’s fascinating to observe the transformation of processes right in front of your eyes. For example, we have a page on our portal where we publish information about interest rates and other relevant information, and I made the dashboard for the page. Every time I open the page, I recall how I thought it up, how I designed it to be striking and interesting. On that page we let users guess the main interest rates that will be set at the next session of the Central Bank, and on average our experts guess more accurately than Bloomberg’s experts. Maybe we can’t use it in our work, but it’s really great!
About the hard part
For me the hardest thing is maintaining relationships and communication with the other departments. It’s not that they aren’t ready to listen. To the contrary, Sberbank in this regard has a very comfortable environment, and I find it comfortable, too. But it’s difficult to put the technical decisions I have in my head into words. For instance, I figured out how to evaluate the medium divergence of one market indicator from another and calculate the bonus for basic risk. But it’s not enough to figure it out. You have to describe the methodology of your decision-making process, choose the right words that present a good case for why it should be done, and come up with all the possible cases when it can be used. That is, formali`z and structure your ideas so that your colleagues both understand and accept them.
Since our department works inside the organization, it is a lot easier to introduce new ideas than, for example, to change something in the Sberbank Online app that is seen by the bank’s clients. The main difficulty is getting the proposed changes accepted by the business bloc, because they might have diametrically opposed interests. Lowering the interest rates is good for some, but raising them is good for others. And then there are departments that calculate indicators for management reporting, and they can’t take everything and use it at the snap of your fingers. Here’s an example: we worked on switching from vintage analysis (vintage means all the deals of the month) to transactional analysis. These sorts of things have to be developed over a long period of time and take into consideration the interests of all parties. My main task here was to prepare an analysis for these kinds of changes. But in general, the level of Sberbank in the sphere of managing assets and liabilities is one of the most advanced – and maybe the most advanced – in the world. My colleagues attend conferences in other banks abroad and we see that we have unique instruments. For example, we have something called “dynamic balance” – a powerful predicting model that evaluates all the future changes in interest rates. We are truly getting to the point when we can directly maximize the bank’s profit and choose the optimal strategy for the bank’s activities. There are almost no other models like that on the market.
About relations with colleagues
Sberbank has a great staff. At any moment you can ask a question, and there will always be someone to answer it or to direct you to the person who can. I’ve never had anyone refuse to help me because they didn’t have time. Thanks to that I’m constantly learning something new – and not just about our bank. I’ve had to study what’s happening on the market and in the economy. In general, it’s great to realize that your bank is a mover of the Russian economy. At the same time, it’s a great responsibility. For example, we were at a meeting to discuss raising interest rates, and one of the main issues was the consequences of that decision for the banking market. Sberbank is a market shaper, and our actions have an influence on other banks.
About my plans
First of all, I want to get my M.A. – I’m doing a joint program with the Higher School of Economics and Sberbank. And second, I want to study client analysis and solve problems related to it. I’m sure that it will only get more interesting, that there will be things connected with machine learning and artificial intelligence, advanced data analysis. Data are right there, lying on the surface, and you just have to find a way to analyze them. And when we do, we’ll find out a tremendous amount.