Commerzbank Hopes to Simplify IT Using a “Complexity Model”

What happens to the cost effectiveness and operational efficiency of a bank’s IT system when performance of individual components change over time? Until recently, Commerzbank and Capco say the answer for CIOs was based on past practice, executive experience and some financial calculations; and was more conceptual than numerical.

A new project at the German bank being spearheaded by financial services consultancy Capco is an attempt to streamline IT by quantifying its complexity. In this work, the bank joins the growing field of firms that are experimenting with IT predictive analytics, or technology that allows banks to pre-determine how a small part of a tech deployment or a single project can improve or hurt the performance of a larger IT portfolio.

"We want to put a number to IT complexity," says Peter Leukert, CIO of Commerzbank, a Frankfurt-based based institution with about EUR 738 billion in assets (about $994 billion), making it the second largest German bank after Deutschebank. "Our basic belief is that if you can't measure something it's hard to manage it."

The two firms have introduced a complexity model that attempts to quantify and articulate short- and long-term impacts of tech initiatives, costs over time, impacts on existing operations and the business as a whole, as well as strategic implications and justifications of high-risk projects.

Definitions of "IT complexity" vary, but it's most commonly thought of as the downstream impact of the interaction between a number of different smaller components in a much larger IT network. Like the butterfly effect, the variables in this interaction could cause latency or other performance issues, making a localized interoperability issue a larger stress on the entire IT network. [The butterfly effect is a part of chaos theory; it’s where a small change at one place in a nonlinear system can result in large differences to a later state. It’s named after the theoretical example of a hurricane's formation being contingent on whether or not a distant butterfly had flapped its wings several weeks before.] “IT complexity increases costs and decreases the ability of a project to meet expectations," says Leukert.

The Commerzbank/Capco initiative is in its early stage, and currently focuses on the institution’s application stack (infrastructure areas such as networks are not yet addressed by the model). Mat Small, a partner in Capco’s technology group, says current measurements focus on the software applications that drive business functions such as trading, CRM and payment processing.

"Based on a time series, we can do simulations or extrapolations and have a dashboard with a measurement of complexity and trends," Leukert says. "We can also do simulations and gain architectural efficiencies for major projects. We can make a more informed decision."

As the model develops, it will incrementally add data inputs to enhance the measurement of IT complexity, expand scenario modeling, and consider how complexity modeling differs across various corporate structures (such as business model, age of business, and size).

The complexity model works by accumulating performance information on various tech components that drive a system that supports a financial function such as trading -- there are as many as two dozen inputs for each component -- and then tracks how this performance changes over a period of time, and also how that individual performance impacts the performance of other components. That accumulated information is run against the complexity model to produce a metric that the bank can use to give a numerical measurement of that component’s complexity.

That measurement can then be aggregated against other tech systems to produce a broader complexity measurement --complexity in payment processing versus complexity in trading, for example. Systems with more functional components can be given greater weight in the calculation.

Component-level information can also be used for forward-looking analysis testing, determining how a new piece of software may impact other components going forward -- giving a bank more information when making a decision on a tech project.

“A numerical value doesn’t tell you a lot by itself. But with the complexity model you can take decision(s) on IT that you are going to make that will change your overall IT landscape; and run that scenario against the complexity model to see how that [new deployment] will affect overall complexity,” Small says.

The model is relatively new, and Commerzbank did not disclose changes in actual tech strategy as a result of its use. The announcement of the bank’s complexity model project with Capco also comes as Commerzbank looks to navigate exposure to the European debt crisis, though the bank did not directly link the complexity model -- which has been in the works for more than a year -- to any larger current corporate cost containment initiative. 

Capco hopes to extend this model to other institutions, which can provide a broader range of data that the tech firm can use to further expand the model’s benchmarking capabilities.

The tech firm is competing in a growing field of IT predictive analytics, which includes firms like Netuitive and Applied Predictive Technologies.

 

For reprint and licensing requests for this article, click here.
Bank technology
MORE FROM AMERICAN BANKER