Global Custodian Spring 2018 | Page 49

[ M A R K E T collected, how do you distinguish the good data from the bad? How are you able to distinguish patterns and trends in the data sets when there is a lack of stand- ardisation? This is where big data has evolved into its latest form: analytics. “Several years ago, the focus was first on data volumes and complexity,” says Pierre Titeux, chief digital officer, Societe Generale Securities Services. “In order to manage these high volumes, major com- panies and institutions have created data lakes designed to store the different data of the company instead of using multiple data warehouses.” “Then the focus recently moved to use cases and added value from the data, where algorithms, machine learning and data science are required in order to ben- efit from enterprise data.” New tools, new conclusions According to David Pagliaro, head of EMEA, State Street Global Exchange, the concept of data analytics, i.e. the process of examining data sets in order to draw conclusions, is not new. However, what is new is how firms are using new tools and technologies to draw new conclusions. “Most think the Holy Grail for data analytics is being able to drive new reve- nues from it, but in reality they are using it to enhance product development and research,” says Pagliaro. Firms are aiming to evolve their data an- alytics capabilities that can provide them with a golden data source, which can be then be used for multiple functions across the business. Even for those fund manag- ers that operate their own data lakes, they cannot always ensure the data they have is the golden copy. According to SS&C’s Meghew, fund ad- ministrators that process data across all of the middle- and back-office are in the best position to provide this golden source. “This golden source is coming from the administrators, and they are then mar- rying it with other data sets that allows them to build that level of intelligence,” says Megaw. In the global custody world where banks handle huge volumes of transaction data across multiple markets with a variety of client segments, it is down to them to come up with new tools that allow firms R E V I E W | B I G D ATA ] to do more with the data. Asset managers that hold long-term investment plans will need constant data feeds and regular updates, far removed from the reactive data storage function of traditional transfer agents or fund administrators. Handling vast amounts According to Mike Clarke, director of product management for European custody at Deutsche Bank, custodians and other finan- cial institutions have the potential to make the data landscape even more challenging if they struggle to deal with the vast amounts of data in their systems. “A proper handle on these vast amounts of data requires these institutions to focus on their business objectives and on building out their data capabilities step-by-step,” says Clarke. “In doing so, they should employ a top-down approach to data, ensuring they learn as they go and design controls and data quality checks to make sure they have outputs that are depend- able. Starting bottom-up with the full suite of data can be slow, painful and unlikely to show any early business value.” Deutsche Bank has carried out changes to its big data platform, Hadoop, with a focus on being to handle large volumes of data and then apply tools to drive data science and analytics. Furthermore, as seen in our other features on cloud technolo- gies (page 40) and data scientists (page 32), many custody banks are hiring data scientists to build these private data lakes in which all of their own data as well as client and third-party data “This golden source is coming from the administrators, and they are then marrying it with other data sets that allows them to build that level of intelligence.” MICHAEL MEGAW, MANAGING DIRECTOR, SS&C GLOBEOP can be stored in. Developing the tools and technologies on top of these data lakes and provide services that enhance decision-making processes is becoming a larger part of the custodian’s growth strategy. “The industry is ingesting billions of terabytes of data into these data lakes, in which service providers process and make the data accessible through APIs, or hold on to it. The future is to use these data lakes to win clients and grow the business,” says Francis Jackson, global head of client coverage, RBC Inves- tor and Treasury Services (RBC I&TS). Jackson explains the bank is also leveraging its network with Canadian universities to harness artificial intelligence (AI) and Spring 2018 globalcustodian.com 49