To gain an advantage in today’s hyper-competitive world, asset managers must not only evolve their operations, but they must also discover new ways to leverage them. Managers must be operationally adept; particularly in a business that grows more costly and complex, they need an infrastructure that will equip them to better meet client needs, satisfy regulators and compete more effectively.
But the days when operating systems were considered just the plumbing behind the real business of asset management are long gone. Sophisticated managers now understand that operations lie at the heart of a firm’s value chain, knitting together the back, middle and front offices, and providing the critical infrastructure for everything an investment organisation does.1
Traditionally, an investment management firm’s middle office has been cast in the shadow of the front and back offices. While consensus is building that the middle office is largely an untapped source of value, it’s often viewed as a cost centre supporting mundane, ancillary activities, sandwiched between its more important counterparts.2
Mundane or not, the demands now being placed on the middle office are greater than ever – increased regulatory requirements, daily performance returns, closely monitored enterprise risk, and new client reporting challenges are putting unprecedented pressure on the middle office. Adding to these functional requirements, the middle office must also support an increasingly complex set of financial instruments, an increasing flow of unstructured data, and a heightened demand for data governance and quality.
In this paper, we consider the above factors, how they are influencing the new middle office and attempt to answer the following questions:
The problem with traditional approaches
Suboptimal and potentially dangerous data practices
Data is the lifeblood of all investment management firms. Many firms, however, find it difficult to unlock the value of the data that flows through their organisation. This challenge is often rooted in the middle office, which is betwixt and between the relatively straightforward functions of the back and front offices. In contrast, middle-office functions can be nebulous and hard to pin down, making it particularly challenging to put effective systems and processes in place.3
The middle office is theoretically the guardian of a firm’s important data and is where much of its potential value is harnessed. The reality, though, is that many firms continue to rely on traditional operating models where reporting and analysis are based on front- or back-office systems. This model inevitably leads to the implementation of multiple reporting tools, each based on its own underlying application. The result is that different report sets are generated and delivered using disparate presentation approaches which results in varying interpretations of the same underlying data. Furthermore, reporting is often generated directly from various vendor- or service provider- owned systems that are likely to interpret the same data differently.
Business users are understandably frustrated. Common concerns include:
1. Questionable data quality/accuracy from the organisation’s core applications
2. Poorly presented data from the application
3. Excessive time to run reports or extract data from the application
4. Insufficient history stored by the application
Given these aggravations, it comes as no surprise that business users often resort to maintaining their own separate data stores. This can extend to users relying on their own data visualisation and reporting tools to create their own presentation layer. The most common example is business users running readily downloadable Business Intelligence (BI) tools, such as Tableau or QlikView, against their personally maintained Excel spreadsheets or SQL Server databases and analytical tools.
This approach causes inconsistency and inefficiency within firms, leading to costly and time-consuming daily reconciliations between systems. Relying on business users to maintain stand-alone spreadsheets and databases makes a bad problem worse, and can be detrimental to the health of an organisation. In addition to increased costs, the firm is exposed to risk associated with a lack of data governance. Also, user efforts are likely duplicated since there is no centralised group responsible for managing data sets. Each of these barriers prevents the firm from being able to maximise the value of its data.
Why is change so difficult?
How do investment firms break free of this model? The answer is to centralise data management, data stewardship, and data governance and to report data to consumers through an intuitive user interface.
This is easier said than done. You can’t buy a single “middle-office system.” There isn’t one. The middle office often shares usage of core systems designed for the front and back offices, with supporting, function-specific applications as required. For example, a compliance team will likely leverage the pre- and post-trade guideline monitoring capabilities of a front-office Order Management System (OMS). The portfolio performance team likely will use elements of the performance measurement and attribution systems purchased primarily for portfolio managers. An enterprise risk team will use elements of the risk systems implemented for front-office risk attribution and analysis. In some cases, even the investment accounting function of an asset manager or service provider may be leveraging an application that was built for back-office (fund accounting) purposes.
Even if there were a single middle-office system, a firm probably wouldn’t want it. Trying to change even one component of a single, integrated system can drain resources and be costly. Taking a more modular approach to building a middle-office services platform allows asset managers to be more flexible over time. When a platform uses a few different systems, firms can swap out components, as they get stale or too costly and can create different business workflows between components to support different use cases.
The new middle office
Derive value from data
Given the traditional model described above, it is clear that firms that want to better leverage their middle office will need to do more than simply upgrade a single legacy application. Any serious effort to unlock the value of the data coursing through the middle office will require an Enterprise Data Management Platform consisting of:
Become guardians of the data
Who has responsibility for owning and managing the data management platform? While IT departments can support the technology, it’s also necessary to define ownership for the daily governance and stewardship of the organisation’s data. While the front and back offices are important users of, and/or contributors to data, the guardian of the organisation’s biggest asset lies squarely in the middle office.
The level of success that the new middle office has supporting data stewardship, data governance, and data quality management will directly affect how well the organisation executes the rest of its operations. The impact – and risk – should not be underestimated.
Centralise data management and reporting
Under a new centralised model, data flows from source systems into an enterprise data management platform where it is validated, cleansed, mastered, and then distributed as required.
Distribution includes feeding front-office applications with Investment Book of Record (IBOR) data upon which portfolio managers will base their daily investment decisions. IBOR is a snapshot of a firm’s positions and cash, updated continuously based on market events and investments. Its counterpart, the Accounting Book of Record (ABOR), is what a firm believes it held at a specific point in time (typically end of day or monthly) that is updated according to market events and investments once per period. In addition to informing investment decisions, data distribution supports the daily maintenance of a data management platform.
In the previous section, we highlighted the inefficiencies associated with multiple reporting tools running against siloed applications in the traditional application-based operational model. In a centralised reporting environment, data received from vendor applications, service providers and external sources is processed and categorised by the data management platform. The data store/ warehouse then supports the various reporting requirements of the middle office, (eg, performance, enterprise risk and client reporting).
Centralised data management does not benefit only the middle office. The front office can use a data management platform to visualise data in ways not supported by traditional front-office applications. For example, instead of relying on two separate report sets from two different risk management systems (eg, supporting short- and long-term risk respectively), a data management platform can easily meld those two data sets into a single report, which analyses both short- and long-term risk for the portfolios managed.
The concept of centralised data stores is not new, but traditionally they have been fed from multiple application sources, making it challenging to maintain data consistently. The differentiator in the new operational model is that it uses the data management platform and centralised data stewardship and governance to cleanse, categorise and channel mastered data into the data store/warehouse.
The provision of value by service providers
The ability to extract maximum value from the new middle office will be dependent on the entire data management platform working in concert:
The data needs to move from the underlying systems to the data store/warehouse The quality of the data needs to be verified
The reporting and data visualisation tools need to call and present the data
One problem for firms is that implementing a true data management platform can be an expensive and risky undertaking. System evaluation, selection and procurement are a significant undertaking. Defining and implementing the required data validations, transformations, and the mastering process is a costly and time-consuming exercise that will require significant attention and resources on an ongoing basis.
In response to these challenges, we are seeing a trend towards Data as a Service (DaaS) offerings from service providers. To alleviate the need for clients to develop and implement their own data management platforms, some service providers are now making their own data management platforms available for their clients to use.
The service providers that are introducing such client-configurable data management platforms should support the entire data management platform rather than just a single set of data (ie, positional data but not transactional data), a certain period of time (ie, current year but without historical data) or a portion of the client’s book of business (ie, Private Equity data but not the separate account data) – ensuring consistency and efficiency of the final solution. Alternatively,
a client can manage data storage components of the data management platform on its own, but outsource integration and support to a third party, or vice versa. Partnering with the right provider is essential for the latter, piecemeal implementation, because asset management firms need a provider whose data management platform is both source and destination agnostic.
To complete the picture, service providers can provide standard, client-configurable data visualisation and reporting packages that sit on top of an asset manager’s proprietary data store/warehouse. This integration, using best practices, can create a larger value-added service and has a greater return on investment compared to a data management project, which can carry a hefty price tag.
In any of the scenarios listed above, the asset manager should always retain ownership of their data. Whether a firm’s entire data management platform or only specific components are outsourced, the manager should ensure it has access to the data in its most raw form, not simply through systematic reporting and data visualisation facilitated by the service provider. This can be accomplished through direct database access, APIs or via contractual workflows that produce portable data sets that the manager receives. Some service providers will even create client-specific data stores where the manager has direct read/write access to database and has the ability to extend the data set.6 An arrangement like this opens many avenues to a technically savvy manager while providing full transparency and portability of the firm’s data, even though the data store itself is not physically located at the manager’s site.
The service provider enterprise data management solution
The data management platform required for the new middle office can be provided as an extension of the suite of options provided by service providers.
Asset managers who already outsource send their trade data, for example, to service providers to update their investment book of record and to support downstream processing. This data can now also be routed to the service provider’s data management platform.
The asset manager helps define the data that will be supported and the processes for how that data will be distributed to downstream applications. The range of data available, and the frequency of update and maintenance, is configurable. The service provider provides the technical implementation and support. The result is an asset manager-configurable outsourced data management solution.
The value proposition to the manager is significant. No longer does the manager need to support multiple reporting tools, which access disparate vendor provided systems, proprietary databases or data warehouses. The data management platform provides one-stop shopping and supports all reporting needs, using the manager’s preferred reporting/data visualisation tool. There is cost savings as the result of reduced support and maintenance and, importantly, this approach provides data consistency.
Reporting sourced from different systems generally leads to reports that show conflicting results. A centralised data management platform, however, generates reports from a central data store or a “single version of the truth.” Thus, over time, the asset manager can achieve consistent, reliable, trusted reporting across the organisation. The time frame within which benefits are realised is generally dependent on the complexity of data, the volume of data, the number of source systems, and the number of downstream applications that need to consume the data.
Asset managers’ use of alternative investments and other new and evolving instrument types have increased exponentially over the past decade. The related analysis conducted by asset managers generates large amounts of unstructured data, (eg, documents, PDFs, graphs, etc.), which need to be sifted and analysed to support decision-making.
While most organisations focus on initiatives dealing with structured data, they must not lose sight of the opportunity to derive significant business value from properly harnessing the power of unstructured data. IDC has estimated that such unstructured content already accounts for an astounding 90 per cent of all digital data, much of which is held across a variety of different data stores, in different locations and in varying formats.4 The data management platform of the new middle office will be designed to support unstructured data. This goes beyond optical character recognition, to the next set of applications that can process and analyse large amounts of unstructured data and produce predictive results.
Given the sheer potential size of a data warehouse, significant opportunities exist for artificial intelligence, machine learning, predictive and prescriptive analysis. In the SEI paper, “The Upside of Disruption,” we explored the idea of “Watsonisation” and how cognitive computing is transforming how things are discovered, interpreted, decided and accomplished and its potential impact on the investment management industry.5 By leveraging cognitive technologies, such as IBM’s Watson or Microsoft Azure’s Cognitive Services platform, asset managers can establish a competitive edge, even as alpha becomes more elusive. In terms of portfolio management, the front offices of asset managers BlackRock and Bridgewater Associates, for example, are trending towards use of cognitive algorithms for predictive analysis of unstructured data to inform investment strategy.
While there are a few asset managers leading the pack in this regard, the vast majority simply do not have the clean and appropriately governed data necessary to benefit from predictive analytics yet, let alone the in-house skill set to determine what questions need to be answered and how those answers can be predicted. The new middle office is changing that, and the benefit of the new data management model has the potential to increase exponentially as cognitive learning technologies mature and gain further use in risk management and operational optimisation.
The service provider reporting suite
In the new enterprise data management platform where all of the data needs to be presented, a robust reporting and visualisation tool is essential.
Service providers are making use of the latest visualisation and reporting tools to produce a suite of charts, diagrams and reports based on their enterprise data stores. While the base suite of reports and visualisations can be seen as “standard,” the latest tools are highly configurable. Asset managers will be able to take the standard service provider report set and customise this to satisfy their unique requirements without extensive development efforts. Some providers also support a bring-your-own-application (BYOA) approach for reporting and visualisation. This approach allows the manager to choose a tool that meets their specific needs or a tool that already has an established presence within their firm. These providers are providing an API service layer to simplify the integration between applications.
The DaaS option
If we consider the aforementioned model, the possibility of providing DaaS becomes very real. In its most simple form, DaaS employs cloud-based data storage. However, DaaS, as delivered by service providers, offer a one-stop shop for asset managers who may pick and choose components to add onto their current in-house platform. If the manager is already a client of the provider, they may choose to extend their current list of services they are getting from that provider by adding data management services.
Many clients have been hesitant to adopt cloud-based tools over security concerns. The fact is, a considerable number of vendor applications and service provider solutions already are supported under an application service provider (ASP) framework that is essentially using cloud-based storage at some level. Firms would be hard-pressed to find a vendor application or service provider who has not moved, or is not exploring the option of moving, to the cloud, whether private or public.
Due to perceived security and regulatory risks and lack of a compelling economic rationale, many banking and other financial institutions historically have not embraced public cloud services provided by firms such as Amazon Web Services and Microsoft Azure. According to a Wall Street Journal article, however, “the pressure to cut infrastructure costs and increase flexibility, paired with more security and compliance services from the cloud vendors, has boosted banks’ willingness to explore the technology.”6 Researchers at Deutsche Bank were told by some bank IT executives that they could go from “zero use of the cloud compute or IaaS (Infrastructure as a Service) model today to 20% to 30% (mostly for dev/test workloads) within 3 years,” an extraordinary growth trajectory.7
The future of large-scale data management platform solutions will require significant data storage capabilities, for which cloud-based computing is a natural solution.
Benefits of the new middle office
Implementing the new middle office can be challenging, but the improved tools, insight and productivity that accompany it bring tangible benefits, including greater operational efficiency, improved transparency, and reduced business and investment risk. These alone should be enough to convince most investment firms to consider at least the new middle-office approach as they face unprecedented levels of competition and regulation. There are nevertheless several additional benefits that are worth highlighting.
Insulation from vendor risk
Bloomberg’s acquisition of Barclays BRAIS and FactSet’s acquisition of BISAM (as well as industry rumblings of some legacy investment accounting systems to end soon) have highlighted the risks associated with vendor loss. Because individual vendor applications provide unique functionality as well as support for associated reporting, clients are increasingly dependent upon the individual applications that support their best-of-breed application model. In addition, the management of the data used by vendor systems is controlled by the vendors, (e.g., the usage of vendor-supplied security master and market data). Finally, and perhaps most important, the data created and saved by vendor applications is often housed only on that vendor application, which means that the organisation’s historical data is sitting outside of the organisation on a system that may not be in existence 18 months from now.
With a traditional application-based operational model, the asset manager is dependent on vendors and, thus, sensitive to vendor loss. The new middle-office model, however, can insulate the manager from vendor risk. With a centralised data management platform, data maintenance and storage is not the responsibility of individual vendors. Rather, data sourced from each of the asset manager’s vendors is cleansed and categorised by the data management platform provider. The data management platform becomes the central data repository for the organisation. If a vendor is lost, there is no need to initiate a project to extract all the historical data from that vendor’s system and remap it to the replacement application.
Creating an abstract data layer, essentially a general data set built to be leveraged rather than for a specific-use case, allows firms to move reporting functions from underlying systems to the data management platform layer. Aside from the benefits of having data in one place, a good data management solution also alleviates the reporting burden of other systems. Back- or front-office systems rarely have robust reporting capabilities and, particularly where multiple systems are used, users typically will have different experiences. Moving reporting to the data management layer allows for consistent reporting across data sourced from all underlying systems. Additionally, this move also gives the manager more flexibility to select which systems they use. Once a data management platform is in place, upgrading or changing systems becomes much easier. All that is required for the new system is a onetime integration and mapping exercise to the data management platform.
Less customisation required
While there are legitimate reasons for bespoke reports and processes, customisation is expensive, time consuming, and resource draining. With required data fully defined and populated in the data management platform, the need for the report customisation is minimised. The traditional application- based operational model forced reporting tools to link data sets from different sources and applications together, which often led to slow running reports with complex logic. The data management platform, as supported by the new middle office, presents all of the data needed by the enterprise in one place. End- user visualisation and reporting becomes an exercise in slicing and dicing data that is readily available.
The new middle office recognises this issue and presents an environment that allows the business user access to the data they need from a governed, supported environment. By providing the reporting and visualisation toolkits that business users are already employing, the new middle office offers a solution in an enterprise-supported secure framework that ensures consistency and governance of the end results.
Restoring user trust in data
One of the greatest benefits of the new middle office is its potential to restore business user trust in data. If the business users do not trust the quality of the data and, in turn, the systems presenting that data, they will not use those systems or the outputs of those systems; instead, they will create their own work-arounds and subsystems. The new middle office governs the data used by the organisation and restores business user trust in that data. The middle office also needs to make sure all users and potential users of the data understand the current state and the future road map of the data.
The future belongs to those that safeguard data
Investment management firms face unprecedented challenges, including the exponential growth of data volumes, diminishing opportunities to generate alpha, and a looming arms race in artificial intelligence and machine learning. Asset managers are looking for competitive advantages wherever possible, and it is in this context that the new middle office is stepping out from the shadows to assume the role of guardian of a firm’s most valuable asset: data. By using a data management platform to proactively steward the organisation’s data and support consistent and governed reporting, asset managers can harness value throughout the organisation.
To generate the most value from data, the new middle office transforms structured and unstructured data from a variety of sources into valuable information and actionable knowledge. It does this by leveraging advanced reporting and analytics technologies, replacing disjointed application-based models, ungoverned databases, and siloed reporting tools, dashboards and portals. The value of this shift is apparent in improved investment decision-making, risk mitigation and increased efficiency. The new middle office will support the demands of increased regulatory reporting, facilitate greater demand for investor transparency, insight into front-office strategies through advanced data quality, availability, and predictive analytics.
For asset managers looking to further leverage their partnerships with their service providers, or for those that do not want to invest in a data management programme, service providers are offering DaaS to their clients with access to the full suite of data management functions. This environment encompasses a secure data architecture, including data control, governance, quality management and stewardship, as well as advanced reporting and analytics through visualisation tools and reporting.
While the traditional middle office may be described as mundane, its next evolution is anything but. Serving as the key conduit between operations and investments, the new middle office represents the future of asset management where cutting-edge technologies enable the things that have always mattered the most: optimising the client experience, cultivating relationships, delivering optimal risk- adjusted returns, and running a successful business.
1 “Evolving in the New Operational Frontier,” SEI, December 2014.
2 “Middle Office: A hidden source of competitive advantage,” SEI, May 2016.
3 “Middle Office: A hidden source of competitive advantage,” SEI, May 2016.
4 Jaikumar Vijayan, “Solving the Unstructured Data Challenge,” CIO from IDG, 25 June 2015.
5 “The Upside of Disruption: Why the future of asset management depends on innovation,” SEI, June 2016.
6 Steven Norton, “Big Banks Starting to Embrace Public Cloud, Deutsche Bank Says,” The Wall Street Journal, 9 June 2016.
7 Brandon Butler, “Deutsche Bank: Nearly a third of finance workloads could hit cloud in 3 years,” Network World, 14 June 2016.
Information provided by SEI Investments Distribution Co; SEI Institutional Transfer Agent, Inc; SEI Private Trust Company, a federally chartered limited purpose savings association; SEI Trust Company; SEI Investments Global Fund Services; SEI Global Services, Inc.; SEI Investments–Global Fund Services Limited; SEI Investments– Depositary & Custodial Services(Ireland) Limited; and SEI Investments Global (Cayman) Limited, which are wholly owned subsidiaries of SEI Investments Company. The Investment Manager Services division is an internal business unit of SEI Investments Company.