There is lots of talk about digital in banking from people such as Chris Skinner in his insightful blog The Finanser. Much of this discussion however is focused on the front office relationship with external customers, specifically about how to build a bank that is responsive to customers in the digital world. I believe that this thinking is too narrow, it is my belief that you cannot have a digital bank without a Digital Back Office. This needs to provide the important services to support the internal management of the bank, as well as the operational support for external customers.
To fully explain the concept of the Digital Back Office and how profoundly different it is to what exists today, it is necessary to understand:
What does “digital” mean?
What does digital mean for a banking back office?
Why should we care?
What does digital mean?
The digital revolution has changed the world. This poses new challenges, but also creates new opportunities. The digital revolution means many things to many people, so to gain a perspective we can contrast the digital revolution with the industrial revolution.
A key driver of the industrial revolution was the economies of scale from the use of mass machine production. To reduce the cost of production, it was necessary to invest significant capital into machines and factories and to apply the concept of Adam Smith’s division of labour. The appropriate business model was to centralise production and to create hierarchical organisational structures to optimise the division of labour.
The necessary pre-conditions driving the digital revolution are advances in computing, specifically 3 observed laws being:
Moore’s Law - the rapid increase in computer processing power
Gilder’s Law - the rapid increase in bandwidth of communication systems, in particular the internet
Metcalfe’s Law - the exponential increase in value of a communication system (or network) as the number of users connect to that system
As with the industrial revolution, the consequences of these advances are a profound shift in how people work, socialise and interact. These advances have democratised processing power and bandwidth allowing anyone with a computer or smartphone to start, or run an internet business which, just a few years ago would have required a large centralised data centre.
Within large organisations too, employees have changed working practices to be more flexible both in content and structure, and have created an environment where networks can develop rapidly outside of organisational hierarchies. We all have the ability to participate in these networks, not necessarily using them to engage with centralised authority, but rather to contribute our opinions as part of a larger disparate group that can become the source of authority. Many people and organisations still fundamentally misunderstand the digital revolution. It is not just about technology and communication channels, it is a cultural transformation, about democracy and empowerment.
This cultural transformation is not good news for existing organisations designed around centralisation and rigid organisational hierarchies. Large organisations that have built structures, systems and processes for an industrial world will need to change in order to succeed in the digital world. To operate effectively in the digital world organisations will need to move from traditional command and control structures to one which harnesses the power of networks, both within and beyond the boundaries of the organisation.
This means a digital organisation should focus its attention on adapting to these new type of behaviours and culture rather than focusing on on the specific technology or communication channels. A key theme should therefore be how to break down centralised structures into smaller units that benefit from democracy and empowerment. Far from being a return to decentralised independent units, the digital world uses federation, a system that allows business units to operate autonomously, but under the guidance of a set of principles and values of a larger organisation, almost like nodes within a network.
What does digital mean for a banking back office?
Whilst the digital revolution affects most industries it is particularly relevant within banking given that banking products are not physical in nature, but are represented as data in banking systems combined with a commitment to provide some service connected to that data. The ongoing service attached to the product is important as it requires the operational processes within the bank to operate smoothly and efficiently as part of the product itself. The part of the bank that provide these operational processes is the “back office”.
When talking about a digital strategy in banking there is almost universal acceptance it should be customer-centric. This requires changing from pushing products to understanding the customers needs and providing the customer with a seamless service and rich user experience in the appropriate time, place and channel.
However this customer-focused approach is too narrowly focused on the front office customer-facing view. Not many people are looking at extending this to include the back office perspective which is critical for seamless access to the operational processes and for quality management information.
In the back office, the concept of digital is currently at best limited to exposing existing data within individual functions onto tablet and smartphone devices. Critically, their status as cost centres means that back office functions struggle to create business cases for investment in digital technology.
In the traditional back office operating model, the front office send data feeds to the back office. This data is then processed in large centralised systems which are used to provide a service to the organisation, for example calculating P&L, risk numbers or checking compliance against rules. The intention of these large centralised platforms was to generate efficiencies through economies of scale, processing this data in a single platform as well as providing a single source of control across many disparate businesses. Control in particular has been given renewed focus following weaknesses highlighted by Libor rigging and other scandals.
Centralised Back Office
The back office builds large centralised platforms to cope with the data volumes and functionality needed to support every type of product across all business lines. This requires large systems that often cannot cope with the variety of functionality required across all businesses, resulting in processes containing manual workarounds. This causes significant cost in both technology and operational support. The data in these platforms, whilst centralised, is often aggregated and inconsistent with the original data from the front office. This necessitates further adjustments usually on aggregated data. For this reason, the effectiveness of the control monitoring is severely compromised. These large platforms have also become inflexible and unable to change at the pace required by all of the front office businesses.
The financial crisis showed that banks back office operations were unable to provide management and regulators with timely and accurate information needed to make critical decisions. Despite significant investment and renewed management attention, banks have been unable to make significant progress on this issue. This requires a radically different model, the "Digital Back Office", that effectively inverts this pyramid and creates a back office structure which is a better cultural fit and takes advantage of the technology advances of the digital world.
Digital Back Office
First, the roles traditionally performed by the back office functions are separated into two parts:
The data management and calculations, and
The reporting of aggregated views, exception management and control/risk management.
Using a federated approach, the functionality to support the data management and calculations is implemented within the systems for each business unit, but in accordance with the standards and rules set by the central function.
The back office still has responsibility for:
Monitoring essential controls through exceptions to the rules generated by each business unit systems, and
On-demand aggregation of near-real-time data at the most granular transaction level across multiple business units
The advances in data virtualisation technology now make it possible to monitor by exception, but still provide full access to source systems to support control investigations, and to perform high performance aggregation and drill-down into trade/transaction level data.
This approach also changes the nature of the relationship between the front and back office. Front office will now provide:
The functionality necessary to calculate metrics required for internal and external client reporting
The business logic/rules that scan data and identify internal control exceptions
This means the front office now provides a service to to the back office, as a client for control monitoring, albeit an internal one rather than a regular fee paying one.
Many traditionalists are likely to baulk at the implementation of control logic on business units systems, citing their lack of independence. However, this model is already standard practice in the front office and there are already precedents for regulatory acceptance provided standard release controls are adhered to.
Why should we care?
There are 3 critical concerns to the top of the agenda of the C-suite executives, in particular the CEO COO CIO and CDO, which are:
Cost Savings - ability to generate significant cost savings to increase P&L
Controls - ability to provide executives with a mechanism to meet their fiduciary duties and increased control mandate required by regulators in the post financial crisis world
Flexibility & Agility - provide an infrastructure that is able to respond to a fast changing environment where customers want new products faster than before and demands from the new regulatory regime
If a significant portion of the back office functionality (data management and calculations) is federated back to the front office, keeping a smaller more nimble and empowered central function that is treated as an important customer, it can help to address these concerns in the following ways:
Cost - large centralised platforms supporting the back office dominate the cost base of most banks. By moving a significant portion of the implementation into each specific business unit, the cost savings would be significant. My practical experience, having been responsible for developing a large back office platform supporting a risk function, is that these front office platforms are more than capable of taking on this additional responsibility. In fact, many front office departments have already developed a “shadow capability” allowing them to validate and respond to the increased scrutiny from control functions, auditors and regulators. Making this functionality “official” by integrating with controlled systems and processes is unlikely to add significantly to the cost, while the savings from dismantling the inefficient and often incorrect back office platforms and processes is very large.
Control - this federated model is in keeping with the shift in behaviours suited to a digital world. Control is about culture and behaviour and a federated model places accountability in the hands of the individual business units, as opposed to being a mandate from senior executives through a command and control structure. The tools made available from advances in technology however still ensure that suitable monitoring can take place independently. Most importantly, the people in the central back office functions move from being reporters and data handlers, to being empowered risk and control decision-makers.
Flexibility & Agility - the most compelling case for this federated structure is that it breaks down the large centralised back office platforms into smaller parts that can be implemented within front office platforms. This gives significantly more flexibility to these businesses to respond to changes without being impacted by changes in other businesses. Whilst there is still functionality required for centralised reporting and control monitoring, this is mainly focused around aggregation and workflow management of control exceptions.
This federated approach has been successfully tried in a number of areas, however most banks have restricted this to small highly targeted areas (typically “quant” libraries which are centrally maintained but implemented on federated trading risk platforms). This is an opportunity for a fundamental shift in organisational design which can provide them with a mechanism to operate an efficient front-to-back service capability which is cost effective, embeds a culture of accountability and control and is agile enough to respond in the digital world.