Reconciliations — New Paradigms for an Old Problem

I have been in financial services many years building and leading various post-trade infrastructure areas in operations, settlement, treasury, and collateral management. One of my reasons for joining a fintech firm was to think about new ways to tackle old problems. With all the innovation that has occurred in the industry, there are still what appear to be intractable problems. Sometimes these problems seem quite simple on the surface, but when you peel off all the layers, they are far more complex. And of course, complex problems have significant and sometimes unintended consequences.

With that as an introduction, let me begin with a particular use case; reconciliation. In financial services post-trade infrastructure, reconciliation is evident everywhere. There are reconciliations of trade positions and activities, settlement positions and activities, reference data, deal terms, etc. Reconciliation occurs across various areas of the firm, e.g. operations, legal, risk, billing, regulatory reporting. Reconciliation also occurs between enterprises, e.g. between dealers and CCP’s. Some organizations have over one thousand people reconciling their data.

Reconciliation is not just a labor or technology cost. Errors or unresolved exceptions have significant implications. Here are a few of the challenges financial firms face:

1. Missed derivative trades that have had funding, capital, and fee implications, particularly during periods of market turmoil.
2. Settlement reconciliations that have required many hours of research and investigation with significant client issues.
3. Fraud undiscovered until weeks after the event.
4. Inaccurate client reporting leading to client complaints and lost business.

Over the years, I have seen many reconciliation solutions. For example, firms have outsourced reconciliation to groups who specialize in processing. They have purchased new reconciliation tools which include many significant capabilities, e.g. workflow management, exception processing, data normalization, etc. Firms have also created enterprise data groups who focus on these issues from a business and technology perspective; creating data standards across the enterprise and for the industry. While there has been progress, our feedback from clients suggests that reconciliation still remains a fundamental problem.

Why is reconciliation still a challenge? There are a number of factors:

1. Scale — the ability to perform bilateral and multilateral reconciliations on vast amounts of data both within an enterprise and across organizations. Most reconciliation solutions are local; limited to specific use cases.
2. Timeliness — the need to reconcile data and resolve exceptions in real-time. Many reconciliations still occur the day after an event. The longer the lag between the initiation of the event and the reconciliation of its data, the greater the cost and risk to the firm.
3. Aggregation — the ability to seamlessly obtain and aggregate data from multiple parts of the organization; both vertical product silos and horizontal utilities across the organization. For example, trade data is used across the enterprise and organizations. The more localized the reconciliation solution, the greater the risk of data integrity issues across the enterprise. This often negatively impacts “back-end” functions such as regulatory reporting.
4. Lineage — front to back visibility of changes to data for processes. Without lineage, the research and resolution of exception items requires substantial investigative work. Firms spend significant time auditing their processes across the organization.
5. Security — protection and confidentiality of data. Data being reconciled is often critical for the firm and cannot be shared with various areas within a firm or across organizations. There are many incidents of data leakage that occur today.

Now that we have covered some of the challenges, let’s discuss solutions within the context of Distributed Ledger Technology (DLT). Before covering the specific capabilities that DLT provides, it is helpful to think about reconciliation in a broader context. Reconciliation is about collecting and maintaining critical data within an enterprise or between parties that must be accurate, secure, confidential, complete, immutable, timely, auditable, and accessible. Many solutions satisfy some of these requirements and are specific to the organization. DLT offers a more comprehensive solution for the enterprise and is extendible to multi-parties.

Let’s look at DLT solutions and how they can meet the requirements above. I will reference the capabilities of PeerNova’s Cuneiform® platform. Cuneiform leverages both blockchain and “big data” technology; critical for addressing reconciliation challenges.

The following are key features (using business jargon, rather than the traditional blockchain terminology):

1. Exception Processing — A number of capabilities are required for real-time exception processing. For Cuneiform, these are all managed by the user. These capabilities include the ability to seamlessly ingest data from any source in any format real-time, to normalize and canonicalize the data to allow for comparison, to allow the user to quickly set up rules for comparison of data, to provide real-time reporting for exceptions, and to utilize process automation tools to propose corrections and fix data where appropriate.
2. Event Lineage — Cuneiform provides the ability to view the end-to-end lineage for a transaction from its source to its final destination. In order for this to happen, Cuneiform cryptographically hashes all ingested data. It uses a rules engine managed by the user to link the cryptographically hashed data. This capability provides instantaneous transparency into the entire transaction life cycle while also providing full auditability and security of data.
3. Insights & Reporting — Cuneiform offers customizable exception reports and streaming analytics to highlight rule breaks and provide a real-time view of the current exceptions. This allows users to quickly rectify outstanding issues and reduce operational risk.
4. Multiparty workflow — Reconciliation across an enterprise is valuable, but multiparty workflows are where the benefits of DLT are maximized. Cuneiform allows users to define their workflows, allocate roles and set rules for the various actors.
5. Golden source — The real-time ingestion of databanks, reconciliation, and exception processing are not enough to create a golden source of data. There are a number of other key capabilities that are required from a platform. They include user defined rules and workflows to identify and correct anomalies with the data, the ability to secure information at the data element level of a transaction (private vs public blockchain), and the real-time distribution of golden source data to other areas of the enterprise or across organizations.

I hope this has helped shed new light on an old problem. While there is lots of publicity regarding DLT’s potential to eliminate reconciliations, it also is an extraordinary effective tool for the journey towards a “golden source” of data, both for an enterprise and for multi-parties. The observations in this blog are not theoretical but the result of hard work with our clients to develop and implement the PeerNova Cuneiform Platform. We think that DLT offers a new paradigm to addressing reconciliation challenges and a great opportunity to leverage a firm’s key asset; the integrity of its data.