ROBOTIC PROCESS AUTOMATION: The silver bullet for operations or another red herring?
Investment banks are currently in the eye of a “perfect storm.” The financial crisis has prompted regulators to re-evaluate the entire industry, and as a result, firms now face the conflicting pressures of shrinking margins and managing ever-increasing compliance requirements. Recognizing that existing operational models cannot be sustained, organizations are exploring new ways to innovate, most notably through robotic process automation (RPA). But is RPA worth the hype? Is it the ultimate solution for enhancing control and reducing costs within operations? Or is it just another quick fix that wonít work over the long term? Nick Fry and Lukasz Hassa explore.
There are a number of factors driving banks to adopt RPA. First, despite many years of investment in technological solutions to improve efficiency and decrease costs, banks have struggled to achieve full automation within operations. Many challenges have emerged, including incomplete upstream data feeds, poor quality reference data, multiple golden sources of data, evolving regulation and legacy IT infrastructure that is often tactically updated rather than strategically reviewed.
As a result of these process shortcomings, organizations require multiple manual workarounds to maintain controls and quality.
Second, while there have been numerous success stories related to offshoring and nearshoring operational functions, there have been less-positive experiences as well. Indeed, some firms have repatriated processes to key hubs from low-cost locations because the initial migration caused more issues than it resolved. In some instances, a proliferation of governance mechanisms to satisfy the new global model has emerged, adding complexity and operational risk rather than curtailing it.
What’s more, migrating processes to low-cost locations can appear to significantly trim the operational budget. However, there are often hidden efficiency costs as experience and expertise are lost with the migration.
Finally, the regulatory response following the financial crisis has dealt a double blow to banks. For one, it impacted bank’s traditional revenue streams (e.g., the Volcker rule prohibiting proprietary trading activities). For another, it forced firms to comply with countless waves of regulation, which increases costs. This squeeze on profitability has led organizations to explore new ways to reduce costs and remain compliant, while retaining margins and appeasing shareholders.
RPA OVERVIEW & BENEFITS
RPA effectively provides firms with a digital workforce of software robots. These robots follow repetitive, rule-based processes in the same way that a human would for tasks such as logging onto user interfaces, processing items in queues, managing exceptions, and so on.
Today, firms leverage RPA in three distinct ways: 1) IT widely uses it to answer helpdesk queries and monitor the health of network devices; 2) Internet retailers and companies like Facebook and Twitter are replicating human interactions with chatbots that handle client communications and answer frequently asked questions; and 3) Within operations, organizations are using RPA for process automation.
The goal of process automation RPA is to automate repetitive and labor-intensive efforts to increase operational efficiency and decrease operational cost. Typically, this type of RPA is applied on top of traditional operations IT systems, such as a content management system (CMS) or a business process management (BPM) platform, to extend their automation capabilities. Process automation RPA will further speed up certain back-office work in finance and operations like data entry.
The goal of process automation RPA is to automate repetitive and labor-intensive efforts to increase operational efficiency and decrease operational cost.
By automating operational processes, RPA brings several immediate benefits to the business:
- Reduced Opex—One of RPAís main advantages is lower operational expenditure through reduced man hours, expressed as “full-time effort.” Robots will work 24/7, 365 days per year without a break, sick time or a holiday. This not only decreases operational expenditure but also diminishes office maintenance costs. Further, minimal staff attrition means business and HR can focus their time on other objectives.
- Accelerated time to market—Robots can usually be trained for repetitive processes in a matter of weeks rather than months. This is a distinct advantage when compared with traditional automation change programs because it accelerates time to value.
- Business ownership—Robots do not usually require sophisticated coding. As such, an operations team can train and maintain them. This is often very attractive to business users because it cuts down the reliance on IT to implement changes.
- Productivity—RPA software uses a programmed set of instructions to perform repeatable tasks, dramatically improving the speed at which processes are performed. And since robots work 24/7, 365 days per year, the increase in productivity can be significant. For example, the Cooperative Banking Group case study on Blue Prism’s website claims process efficiency gains of 80 percent.
- Accuracy—Robots are not susceptible to fatigue or human negligence, which in day-to-day operations means significantly fewer fixes and error corrections. Also, software will not make biased decisions based on personal inclinations or pressure from managers and auditors. In the long run, robots improve data and process accuracy as well as banking operations transparency.
- Scalability—Without RPA, firms must hire and train additional staff to cover increases in demand, which is time-consuming and costly. With RPA, adjusting for increasing demand simply means changing software configurations. Additionally, firms can reuse robots across different parts of the organization without expensive revisions and IT architecture restructuring.
- Enhanced analytics capability—By using robots, firms have immediate access to analytical data and can easily simulate how any changes to the workflow will affect the speed and accuracy of the process.
APPLYING RPA IN THE REAL WORLD
Even today, investment banks dedicate significant amounts of labor to perform repetitive, rule-based efforts and operations. These processes, such as managing data exceptions and data entry, are unique to each organization because the front-to-back connectivity sophistication varies significantly from firm to firm and function to function. RPA can add value by automating those repetitive duties and focusing a staff’s energy and attention on more strategic and fulfilling tasks.
Consider this example: A large investment bank has an upstream database that takes feeds of instrument data from various industry sources such as Bloomberg and Reuters. However, the business case had never been successfully made to build an automated feed to the downstream equities platform. As a result, a manual control process was implemented within operations to ensure all changes and additions to the instrument database were reflected in the downstream system. With the introduction of RPA, the management team was able to quickly turn this manual process into a robot-owned function, with the robots retrieving data exceptions from the source system and entering the information into the downstream system. This enabled headcount to be freed up for more valuable tasks within the division, such as improving customer service.
RPA is also ideal in situations where a data entry conduit is mandatory. Some legacy platforms require data to be entered or actions to be conducted via the user interface in order for the process to occur. In these scenarios, RPA tools can synthesize the actions of the user and mitigate the need for manual intervention.
Additionally, RPA is perfect for scenarios where a firm is implementing a solution for a regulatory deadline, but full automation of the process cannot be achieved within the designated timeframe. RPA can “plug the gap” and avoid manual alternatives while the strategic solution is developed.
If a process is fundamentally flawed or broken, and it is converted from human-owned to robot-owned, then the underlying data quality or connectivity issues that existed in the old world will also exist in the new.
IMPLEMENTING RPA: OVERCOMING THE HURDLES
As with the introduction of any new software product, RPA doesn’t come without challenges. Time must be set aside for the adaptation and stabilization of the newly adopted platform. Organizations must carefully plan the introduction of RPA due to the low level of standardization across the enterprise and a number of proprietary in-house processes.
It is important to note that RPA is also likely to be affected by a quickly changing technology landscape and will require updates and bug fixes. Process automation RPA is highly sensitive to any changes in the user interface (UI) on which it operates. Therefore, it is critical that any changes to the UI are communicated to the owners of RPA and all changes are tested to ensure the RPA does not break upon the production release of a new UI.
Other potential challenges include limitations with existing infrastructure, automation and data ingestions. RPA will successfully automate a process performed by humans, but it will not magically resolve data quality issues or existing platform limitations. If a process is fundamentally flawed or broken, and it is converted from human-owned to robot-owned, then the underlying data quality or connectivity issues that existed in the old world will also exist in the new.
Further, while RPA is faster than human operational speed, it is actually slower than API-integrated software. Firms should always ultimately look to process data through a fully automated route via an API, particularly given the fact that many of these repetitive processes within operations exist merely to cover limitations in the front-to-back connectivity and infrastructure, and should look to eliminate them altogether rather than replace them.
Many RPA tools cannot process 100 percent of data formats either. Firms may be required to build adapters using technologies such as C# or Java in order to transform such data into a readable form for the RPA tool. If this is the case, firms should perform the necessary cost-benefit analysis to ensure the effort required and any essential adapters would not be better spent on the implementation of full straight-through processing (STP).
Just as the industry landscape for operations has evolved significantly over the past 30 years, so has the need for a far more efficient and controlled approach to covering inadequacies in the front-to-back trade lifecycle. RPA provides a solution for the management of repetitive processes that is cost-effective, efficient, productive and scalable—and can often be applied within weeks rather
However, traditional process automation RPA should not be seen as the “silver bullet.” It is just part of the solution for investment banks, and firms should always seek to ensure that they do not lose sight of implementing strategic change. Optimization and synthesization of front-to-back infrastructure will help close control gaps and eliminate the need for many repetitive processes altogether. This should continue to be the ultimate goal. Where the application of RPA becomes very interesting, however, is when it is combined with more advanced, cognitive artificial intelligence tools to deliver intelligent process automation (i.e., robotic automation that not only processes pre-programmed functions, but also makes autonomous decisions using a sophisticated rules engine based on learning algorithms). While many traditional technologies are still not advanced enough to fully automate processes with complex decision-making, the application of tools that learn and think, as well as do, opens up a myriad of new automation possibilities.
While many traditional technologies are still not advanced enough to fully automate processes with complex decision-making, the application of tools that learn and think, as well as do, opens up a myriad of new automation possibilities.
To evaluate RPA and its viability within the business, firms should answer the following three questions:
Is RPA the only answer? While it’s true that many legacy operational processes cannot be easily automated by existing software, there have been cases in which operations departments have sought to apply RPA to processes that, with little cost or effort, could be fully automated using available industry platforms or even existing tools within their in-house technology suite. It is crucial that firms conduct the necessary due diligence before embarking on the RPA route.
Can RPA cover all in-scope use cases? As emphasized above, some RPA tools struggle to cater to all data formats. Therefore, it is vital that the right level of analysis be performed to mitigate the risk of building technology on top of technology.
What is the primary reason for the process the RPA will manage? If the RPA will be used to cover a control gap inadequacy in the front-to-back infrastructure, then firms should, in parallel with the definition of the business and functional requirements for the RPA implementation, produce a roadmap for how the process can be eliminated altogether.
In summary, RPA can be extremely effective for increasing control, improving efficiency and reducing cost if it’s applied in the right way. However, before implementing it, firms should perform the right level of upfront analysis to ensure that significant benefits will be realized for the short, medium and long term.
Nick Fry is a Director and lead post-trade subject matter expert based at Sapient Global Markets in London. Nick has worked for 23 years in the capital markets industry, with a background in investment banking operations followed by financial services consulting. His extensive derivatives subject matter expertise, deep knowledge of the operations domain and first-hand experience of leading operationsí process re-engineering initiatives draws from his time at Deutsche Bank and Sapient Global Markets.
Lukasz Hassa is a Manager based in London with extensive experience in middle-office operations, trade lifecycle management and enterprise data in the capital and commodity markets. Throughout his career at Sapient Global Markets, he has leveraged his prior semiconductor industry experience in the delivery of technical solutions across trading, data and analytics. Currently, Lukasz is a project manager for a major bank’s MiFIDII program.