Curiosity is higher than the market – Economics – Kommersant

The Bank of Russia presented for public consultations the report “The State and Prospects for the Development of Data Management Systems of Financial Market Participants” — after examining more than 100 banks, the regulator found that the development of their data management systems is highly fragmented, while the largest share of respondents (36%) refers to the initial their level of application. In the report, the regulator pays special attention to the problem of data quality – it needs reliable and up-to-date information to make adequate systemic decisions. Experts, however, note that banks themselves may not need such quality of information for normal work.

Only 30% of financial institutions have an effective level of maturity of data management systems, the Bank of Russia found out during a survey. The share of organizations at the “developing” level is 34%, at the “initial” level – 36%. The regulator distributed the groups based on several criteria, including evaluating the attitude of respondents to the value of data and the priority of their quality and analytics. Also, the Central Bank divided organizations into three groups according to the level of discipline and the quality of regulatory reporting: more than half (55%) make an average number of errors and have an average discipline, 26% – a large number of errors with low discipline, and only 18% – are disciplined and rarely make mistakes. . The number of errors during mandatory control, resubmission of documents at the initiative of organizations or late submission of reports was used as metrics.

“The target state of data management systems is to find the majority of financial market participants at a level of maturity that allows them to unlock the maximum potential for using data that meets the requirements for quality, reliability, processing speed and availability, as well as modern technological challenges,” the report says (.pdf). Referring to the work of a number of researchers, the Bank of Russia notes that data have become modern means of production and should be accessible and of high quality.

“At the same time, two-thirds of development teams complain about the poor quality of data, and every second team complains about the lack of data or their presence in an inappropriate format,” the authors of the report complain. To combat this, the Central Bank intends to develop the methodology and model documents. However, market demands are also different. “Someone conducts a detailed business analysis within their organization and needs a reference data model, someone relies on software and service manufacturers,” the Bank of Russia admits.

According to Yury Pakhomov, an expert at the Big Data Analysis Methods Laboratory at the Higher School of Economics, there are various kinds of errors, and it is important to note that they can be caused by various reasons — the nature of operations, imperfect IT infrastructure, and management practices.

There are also many solutions to the problem of data quality – from the introduction of data quality control practices at the stage of their collection to its automation, control of primary sources, etc., and in principle ending with the transition to data-oriented companies, when the collected data is used in internal analytics, believes the chairman of the Association of Data Market Participants Ivan Begtin.

“The quality of the data that the financial structures of the Russian Federation operate, among other things, is quite low,” agrees Oleg Giatsintov, technical director of DIS Group. Some of the issues related to quality can be solved by optimizing business processes, however, their own quality metrics, which are used by financial market participants, often differ from those metrics set by the Central Bank. According to the expert, there are two reasons: first, in the current conditions, an ordinary bank may not have the technical ability to provide the level of data quality required by Central Bank analysts, but often businesses do not need the same high quality. The second is that data quality management solutions (Data Quality class solutions) are often not enough for the required quality level for each of the metrics – data management solutions (Data Governance class solutions) are also needed.

Venera Petrova