The word reconciliation usually makes Operations people wince, and leaves others confused about ‘how hard recon can be’. The legacy perception of reconciliation is a very manual recon process, consisting of complex excels or a basic recon system with excel macros on top. After the manual or Level 1 recon process, another manual and labour-intensive remediation process follows, that may have webs of people involved. It’s a very exhaustive process, with limited control and even harder to get metrics and transparency into the process. But technology has solutions for all these issues, and more.
Reducing the pain in Recon
Recon Technology has advanced, with Configurable Recons allowing the Business SMEs to set up recons to their own requirements. This removes the needs to have additional macros running. In simplest form configurability allows users to add extra columns to the recon to analyse the issue with more detail. Additional configurability allows the users to add additional matching rules and we see significantly improved matching rates from letting the experts control the data.
Recon is no longer a standalone process. Gone are the days when there are different systems for different parts of the problem/solution. The real game changers are systems that cangive functionality to cover the End-to-End Recon process. These means functionality to manage incoming files/messages, ETL on those files/messages, Reconciliation, Workflow,Checklists, Machine Learning, Exception Resolution and Reporting.
Modern ETL (Extract, Transform,Load) tools are less complex to use that Excel formulae, so Operations users find them easy after years of Excel user. Maybe Excel will become a training tool in future.
Tracking breaks day over day is a must have now, especially with the ever increases Regulatory reporting requirements. Having to use a web of people also means assigning breaks to different users. A User Configurable Workflow is also basically mandatory for any reconciliation system in order to track age, commentary & attachments and materiality. This allows companies to simple create reports on any breaks that need to be reported over X days or over Y materiality/NAV Materiality amounts.
Why have a different system for Checklists when all the clicking is done in single system. This now means the sign off process and checklist can be audited together.
Start of the Solution
The game changer for the end-to-end process is Automatic Exception Resolution using Bots and connectors, also known as Intelligent Process Automation (IPA). Recon users recognise the same breaks day over day and in the same way we can define Business Rules so the system can recognise these same breaks. These Business Rules integrate with upstream and downstream systems using bots & connectors assigned to post transactions to resolve the break. A typical example is a bank fee being posted on the bank side, and the business rule recognising that and pushing a bot back to the accounting engine to post a bank fee for Z settlement date.
Users no longer have to wait until T+1to see a view of the Recon post Resolution. Bots also allow us to re-run the recon to see that those actions & any manual actions have cleared the breaks on the reconciliation. The end users now get to see an up-to-date recon without promises of unseen fixes.
We keep hearing how complicated Machine Learning is, and how people are confused by it. Even toasters have ML these days. Why wouldn’t it be possible to set up ML for data that is coming in the same format every day and seeing the repeat/similar issues. There are opportunities to help pre and post reconciliation. The robots want to help to classify/correct issues.
Modern Managers love Meetings and Reports. The meeting part will take longer to solve,but the Reporting Solution is available to play with. The data for the end-to-end process is available,it just needs to be configured to your requirements. The biggest use cases of Reporting are: Monitoring the metrics/health of the recon process, Tracking SLAs live, SLA Reporting, Escalation of Breaks over thresholds and sending Recon Results to Clients. Reporting can be tied into the Recon signoff process and process steps combined.
Recon systems can also store info on data that was manually matched. This allows the system to learn how to create new Automatic Matching Rules to reduce the manual processing time and keep the manual element to a minimum.
Another use is the ability to review the source of the issue, and work to resolve it. With all the data in place, it is easy to review where the issues are coming from, whether another input issue from John Smith or helping explain missed SLA’s due to late files from brokers/clients. Machine Learning on breaks can be very helpful here where the Explainers show the features they use to choose recon break reasons. This can show some training issues required on certain products, or an ETL update required. This global view ensures a much leaner recon review process and thus more time for meetings.
Gone are the days of manual piecemeal processes. Technology is here to solve the problems. Recon can now be the starting point to automatically send resolutions to upstream systems, resolving the reporting dilemma, improve the efficiency of the process by identifying upstream issues and learning new Matching Rules.