Reconciliation ensures all parties have consistent and accurate information about a financial transaction and corrects any discrepancies in that data. The systematic automation of the reconciliation process can reduce costs, improve accuracy, and minimize regulatory risks. It can also make it easier to adopt new technologies, such as the tokenization of assets and the use of distributed ledgers.
Banks can, and we believe must, achieve 90-95% truly automated matching rates in their reconciliations by building new ecosystems of tools, processes, and skills. This will dramatically reduce their costs and, more importantly, improve accuracy and achieve the real-time reconciliation and compliance that customers and regulators require while paving the way for ongoing digital transformation.
The problems with current reconciliation processes span technology, corporate culture, and organizational charts.
Fragmented legacy systems make it difficult to gather and analyze the required data, and to assure that data is accurate. Many banks view reconciliation improvement as an IT project instead of a top-down, strategic project, which makes it harder to get the required executive support and user buy-in. Because they don’t know or have dramatically underestimated how much they are spending on reconciliation, it may be difficult for banks to make the business case to reduce that spending.
Business leaders are also wary of investing more in fixing reconciliation issues when previous attempts have failed. They may lack the skills or tools to manage the required data or bridge siloed reconciliation processes or fear that normal transaction processing will be interrupted during the improvement action.
Many banks have elected to implement stopgaps such as robotic process automation (RPA) to boost their straight through processing (STP) rates. However, this only automates unnecessarily high rates of reconciliation, without reducing the risks of errors or the manual effort required to identify the right team to resolve a conflict.
Automated reconciliations can be expensive, requiring significant time to resolve discrepancies, and they can delay compliance with new and ongoing regulatory changes. They require more scrutiny than current processes, make it more difficult to meet upcoming demands for real-time control and compliance and may prove to be a significant obstacle to the adoption of T+1 settlement of transactions in the US.
Building a Better Ecosystem
With so many challenges to overcome, a true overhaul of reconciliation requires a systematic approach. This includes front-to-back planning, improved management of more types of structured and unstructured data, as well as streamlining and standardizing processes and controls. It also requires improved orchestration of all stakeholders to move toward an enterprise ecosystem that can learn faster and from more cases about how to automate the reconciliation process.
Over time, we expect financial services organizations will increasingly use artificial intelligence to quickly recognize reconciliation breaks and learn from new cases (i.e., “fix once and fix forever”) to eliminate future occurrences and leverage distributed ledger technology to share the same view of the truth among trusted parties.
Five Steps to Improved Reconciliation
In our work with banks worldwide, we have found the following elements helpful in driving meaningful improvements to reconciliation automation levels.
First, create a data transformation layer that normalizes and enriches the data feeds and data attributes used in reconciliation. Be sure to allocate adequate budget and development staff to tackle the inevitable data quality issues that will plague upstream users if not addressed. Also consider use of cloud data platforms to enable scalability and flexibility.
Second, build consensus among business users and IT teams around a standardized, front-to-back target operating model that spans technology and processes.
Third, apply intelligent process automation (IPA) to manual processes throughout the reconciliation process. Don’t forget to also apply such automated orchestration to the exception management workflow, which is a big driver of cost and effort in current reconciliation processes.
Fourth, implement real-time risk management through a reconciliations dashboard and reporting tools that assure the efficiency of the new process, streamline regulatory compliance, and reduce the risk of operational losses and the cost of reconciliation. Performing data analytics on the root cause of failures at a detailed level will help fix problems at the source or automatically suggest resolutions for them.
Finally, fine-tune a plan to deliver the target operating model on time and within budget. This plan should specify critical details, such as reducing parallel runtimes and decommissioning legacy systems, as well as plans for automation, process, and reporting. Delivering the desired long-term solution is as important as reducing the up-front investment and meeting short-term cost reduction targets.
This process will have been successful if it delivers an automated, flexible factory-based model to support automated reconciliations as business and regulatory requirements change.
A standardized, reimagined and straight-through reconciliation process does more than just reduce cost and risk. It provides a foundation for growth and improved competitiveness by providing improved visibility into reconciliation success rates and regulatory status.
By Jerome Dumaine, Vice President, Banking and Financial Services and Global Capital Markets Solutions Leader Cognizant Technology Solutions