Financial institutions (FIs) recognise that the industry is at a tipping point where cloud is now the default option with buy-side firms maximising the benefits of cloud technology by adopting a managed service model.

By migrating to the cloud, FIs can reduce costs, increase flexibility and re-allocate resources to more strategic projects.

This is highlighted in a recent WBR Insights and IHS Markit report ‘Taking Data Management to the Next Level: Key trends at buy-side firms with $10-100bn in AUM’.

Cloud offers great potential for buy-side firms, and Finextra spoke to Andrew Eisen, global head of solutions for financial services at IHS Markit about the report, the benefits of adopting a managed service model and outsourcing, and how the cloud can help map varied data sets that need to be analysed for insight.

Cloud as the default

Eisen highlights that “five years ago, there was a general scepticism about whether financial services providers, businesses and other large corporations with sensitive data would be comfortable using or outsourcing infrastructure to the cloud.

“However, most of the firms that we work with have adopted a cloud-first strategy. In boardrooms, it’s not a discussion of if they’re migrating to the cloud, it’s how fast and how to deal with the shift from CAPEX (capital expenditure) to OPEX (operational expenditure) and the change in people and their skills,” Eisen says.

Referencing the finding that 52% of buyside firms use the cloud for data management today, Eisen predicts the figure could increase to as much as 80% by the end of 2021. The growing move towards cloud is reflected among IHS Markit’s own clients with up to 70% of new clients selecting a managed service over a traditional enterprise software deployment.

Embedding cloud capabilities into native solutions gives FIs the ability to scale in a cost-effective way that has not been traditionally available, for example high-performance computing (HPC), where tens of thousands of machines can be spun up to run complex calculations to complete a task. When complete, FIs are able to turn it off and stop spending for it.

“Cloud is not just changing the underlying cost and the availability of the infrastructure, it’s also changing the logic of how FIs go about solving problems. It has been an amazing facilitator for innovation,” according to Eisen.

Differentiation through outsourcing

Managed services often bridge new technology and the change in processes while freeing up scarce capital or resources to allow FIs to focus on the parts of the business that differentiate them from their peers. Eisen explores how outsourcing managed services impacts risk management.

“In financial services, with new regulations and regulatory risk on the horizon, institutions are increasingly offering up parts of their business to third parties to run. They now must explain how it affects their risk profile, especially as traditionally, the consensus was that managing services in house reduced risk.

“Cloud technology providers have scaled and established credibility, and because they have such specialisation, their ability to eliminate risk in the cloud is better than the FIs have on-premises.”

A note on hybrid models, Eisen states that this is “almost a religious conversation.” A hybrid model is for those who do not want one vendor to have control over their infrastructure.

Ultimately, however, it requires increased capital to allocate workloads across different cloud providers, and FIs will not be able to take full advantage of the features, functions and expertise that is being offered to them.

This also suggests that there is value in physical infrastructure. Eisen describes an “interim where the eventual goal is to get most of the operating model in the cloud but the FI is tied to those physical assets, which are a fixed investment in software and capital.”

Single source of truth

While FIs may opt for variations of cloud provision, there is a single source of truth for operational data, especially where machine learning models are concerned. “Data management, as we define it, is about making sure that the data can be trusted in operational processes. You have a record of the truth that can be traced through the lineage of governance of that decision-making process, so in an audit, the best-known data is available for that moment in time, for that particular task.”

However, as Eisen explains, with machine learning and portfolio analytics, the data that has been deemed “bad” would also need to be considered, as this could affect the positive outcome of the model that is being built.

“Data management for the concept of mastering the single source of the truth is for business data or operational data. When you start getting into theoretical potentials, you can have multiple versions of the truth,” Eisen concludes.