Achieving reconciliation maturity will be easier in the decade of data

It’s taken some time for the finance world to put its trust in cloud-based software. Back in 2013, when we first launched Duco, we were advised that a cloud-based system would never get off the ground in finance, writes Christian Nentwich, CEO at Duco.

Financial professionals feel more comfortable entrusting their data to the cloud.

Compliance was a key concern and trust in tech generally was low; legacy on-premise software had often failed to deliver on its reliability and efficiency promises – proving costly, and difficult to integrate or update. Thankfully, the last decade has seen an extensive digital transformation in the middle and back-office departments of our financial institutions. While solutions providers like us were working hard to gain the trust of their users, public cloud providers were also working hard to prove their safety and security. We arrive in 2020 to a finance sector that is, for the most part, ready and willing to try new tech; one where we check the integrity of one billion data records every four business days – compared to one billion over 20 months just six years ago – and where financial professionals feel more comfortable entrusting their data to the cloud.

Reconciliation lags behind

We believe that the increase in data running through our own systems is a clear sign that 2020 will be the year software-as-a-service (SaaS) data management truly comes of age. For banking institutions, we like to think this heralds the beginning of a ‘decade of data’ which will allow them to leave time-consuming and error prone processes behind once and for all. However, there are still some obstacles to overcome – and this is particularly true when it comes to reconciliation.

Reconciliation is a mission-critical process. It helps firms eliminate operational risk, identify potentially fraudulent transactions and avoid errors that could lead to costly penalties. Despite this, many firms still rely on disconnected systems, manual processing in spreadsheets, or inflexible legacy technology to get the job done. The question is why? Why is reconciliation so hard to automate and what can forward-thinking firms do about it? These are questions we have set out to answer over the last year and which led to the development of our Reconciliation Maturity Model (RMM).

The first three stages of maturity

The RMM breaks progression towards full automation into five key stages. It’s designed to help firms benchmark their progress and identify their best next steps. The first three stages represent the status quo for a vast majority of financial firms: stage 1 – Manual, using spreadsheets, macros and home-grown applications; stage 2 – Hybrid, using point solutions developed specifically for different tasks and spreadsheets for those that fall outside their scope; stage 3 – Centralised, with the reconciliation function located in one or more low cost areas. This third stage allows information to be shared, best practice agreed and economies of scale realised. Unfortunately, it usually leaves around 20% of reconciliation work to be carried out manually, so firms will still struggle with auditability and cost-efficiency – and this is where progress often seems to stall.

Achieving full automation

Stage 4 of the RMM is Automated, with all reconciliations consolidated on one automated system used enterprise-wide. Our research tells us that, unlike progression through the first three stages, moving on to this fourth stage is not a natural jump. In many cases it demands a fundamental rethink of how operations are structured. Obstacles to getting there include poor data quality and a lack of standardisation, combined with increasing complexity in the kinds of derivatives and associated data being handled. However, it’s here that SaaS could be a game changer: having the right tech and reliable support will empower more firms to take the leap – and reap the rewards of doing so. By putting all reconciliation data in one place, in one system and in a standardised format, firms can dramatically cut costs, improve efficiency and eliminate risk.

Get ready to embrace machine learning

As we enter this brand-new decade, we hope the RMM will help more firms take the next step towards reconciliation maturity and get closer to the ‘holy grail’ of reconciliation, which we have defined as stage 5 – Machine Learning Enhanced. This is where machine learning technology can be used to continuously improve the reconciliation process, up to a point where the it automatically spots and corrects errors across internal systems, and flags trends in data quality. Ultimately this will result in eliminating the need for the majority of internal reconciliations.

This optimal future isn’t all that far away. Machine learning is already being employed within the Duco system and offers exciting benefits for current users, which will only improve with time. Also, as more people sign up to use the system, this creates a flywheel effect as the machine learning algorithms that start to improve with the amount of training data they can access.

Our recommendation to firms would therefore be: get ready to take advantage of machine-learning, and a future where reconciliation can be minimised. The best way to do this is by choosing a SaaS system, where the amount of training data is ever-increasing. Machine learning algorithms on installed systems are only able to train on a limited set of data. By positioning themselves as ‘stage 4’ organisations, firms will not only see immediate benefits, they will also be ready to take advantage of new functionality when it arrives and to lead the way during this decade of data.