Future of Finance

CONTACT US FOR FURTHER INFORMATION 📞 07725 160903

The Data Opportunity in Asset Management Part 1

The Data Opportunity in Asset Management Part 1

An asset management industry under margin pressure is turning to data as a source of reduced costs and increased revenues. More efficient handling of data is expected to raise productivity, improve investment product design, tighten and expand distribution, make compliance easier and enable firms to comply meaningfully with ESG mandates. So a lot is being asked of data. The question is how soon rising expectations can be fulfilled, and who will do the work of clearing the obstacles to achievement of these goals.

Click here to see event information on January 19

SUMMARY


In asset management, data has ceased to be an interest of the front office only. It is now seen as a way to cut costs and risks – especially compliance costs and risks – in the middle and back office too. 


Alpha Data Solutions has put the potential cost savings from more efficient internal handling of data by the top 400 asset managers at US$1 billion a year.
In addition to internal needs for higher quality and more timely data, the data demands of clients and regulators are also exerting pressure on asset managers to improve their data management.


Asset managers also see data management as a tool to improve investment performance and increase revenue by deepening relationships with clients.
The first objective is to create a single consistent set of data that different functions within a asset management firm (portfolio management, risk management, sales, operations and so on) can use for different purposes. 


The secondary objective is to extend the benefits of consistent, accessible data to interactions between asset managers and third parties, such as clients, potential clients, fund distributors and custodians. This will make exchanges between all members of the eco-system more efficient. 


Modern techniques for achieving this eschew aggregation of account-based data set in a data warehouse, in favour of capturing rich data about transactions from their origins in the front office, and updating it and making it available continuously. 


Instead of monolithic, centralised databases, the data will be stored in and accessed from the Cloud and be available as time series rather than daily snapshots. This approach owes something to the distributed ledgers pioneered by blockchain technologies.


Global asset managers and global custodians are developing open data management platforms which aim to capture, host and make consistent flows of data host between competing asset managers, service providers and data vendors using Cloud technologies.


Asset managers that remain suspicious of sharing even non-proprietorial data with competitors are exploring alternative methods of achieving the same objective, such as purchasing packages of independent services that can be integrated. 


The obstacles that must be cleared within as well as between asset management firms include multiple data management systems and silos and a variety of Books of Records and data Masters.


This fragmentation inhibit innovation, not only because data sets are hard to reconcile and so difficult to sue for more than one purpose, but because changes in one system have knock-on effects on other systems.


Another obstacle is cultural. Constituencies within asset management firms are reluctant to cede control of data in case it renders their role redundant. Likewise, data vendors are concerned that scalable data platforms will erode their revenues and profits. 


However, the benefits to asset managers of more efficient data management are multiplying. They include the transfer of the regulatory and litigation risk that data is inaccurate; investment product development; better distribution; increased sales; lower costs of data purchase from vendors; and improved investment performance.


Although data standards could accelerate progress towards more efficient data management, confidence that standards can be agreed and adopted is low. Instead, there is an expectation that different parties will align their data models as the benefits of exchanging data efficiently become obvious.


Doing nothing about data is not an option for any asset manager. Managers that overcome inefficiencies in data management, and seize the benefits of more efficient data management, will achieve better investment performance and higher profitability. Those that do not risk being driven out of business.


                                                                                                            FULL REVIEW

Asset managers have always relied on price and economic data to make portfolio investment decisions. Now, under pressure from cost transparency, the shift to passive investing and rising levels of regulatory interest, they are looking at data to improve efficiency at every level, from portfolio management and execution in the front office, through risk management and financing in the middle office, to settlement and custody in the back office.   


Inefficient data management is expensive and risky for asset managers


There is certainly room for improvement. Even sophisticated asset managers often struggle to substantiate something as basic to their business as the value of the assets they manage. This creates disjunctions in client reporting, which creates reputational and sometimes regulatory issues.   


Alpha Data Solutions reckons that internal work on standardising data is costing the top 400 asset managers US$1 billion a year in avoidable expenses. 


The increased and increasingly detailed reporting demands set by regulators – such as Forms ADV, PF, Annex IV, CPO-PQR and OTC derivative reporting – increase data costs and risks, in the sense that reports prepared for different regulatory purposes have to be reconciled to achieve consistency.   


Asset managers also expect better data management to cut the costs of Know Your Client (KYC), Anti Money Laundering (AML), countering the financing of terrorism (CFT) and sanctions screening checks. 


They believe it can help to automate other compliance obligations as well, including their responsibility to treat customers fairly and not mis-sell them products, and meet the growing pressure to conform to Environmental, Social and Governance (ESG) standards. 


Clients need data from their asset managers for their own purposes

Clients of asset managers also need data, most obviously on investment performance. Here, embarrassing gaps and inconsistencies continue to arise, not least because the portfolio managers and heads of operations operate off different sets of data to those used by sales-people and relationship managers. 


But clients need more than performance data to report to the underlying beneficial owners and to their own regulatory bodies, including ESG data. The growing commercial as well as regulatory pressure for greater transparency in the techniques asset managers use – and the costs those techniques impose on their clients – is another driver of investment in better data management. 


So asset managers are under pressure to supply their customers with more (and more detailed) information, and to improve its accuracy. 


They also need to cut the cost of producing and reconciling multiple regulatory reports, and better data management is a means of achieving that.
But there is also a growing conviction in the asset management industry that data can confer competitive advantage. The increasing digitisation of customer interactions – much accelerated by the pandemic – has expanded the volume of data available. 


Data can lift sales by an improved understanding of clients 


In theory, it can be matched with other client data to understand which customers are buying which products and provide forward-looking intelligence about customers that could enable firms to optimise sales opportunities. 


In fact, the Customer Book of Record (CBOR, or information about what clients – generally distributors, rather than underlying investors – own and subscribe, and how profitable they are, in terms of flows, holdings and revenues) is now taking its place alongside the Accounting Book of Record (ABOR, or fund accounting data) and the Investment Book of Record (IBOR, or settled positions and cash balances) in the lexicon of asset management abbreviations. 


The CBOR represents a shift in the asset management industry from seeing data purely as a source of management information to the much-mentioned “client-centric” view. 


It is a significant change, in the sense that knowledge about the client alters the sales and relationship management processes from selling product to addressing client needs and preferences. Being able to prove ESG impact on a client portfolio, for example, is almost entirely a matter of better data management. 


Better data management begins in the front office, not the middle or back 


Equally importantly, data management (as opposed to market data) is no longer seen as largely an operational discipline. Better data management is now seen as beginning in that part of the firm from which all asset management activities ultimately stem: the front office. 


That is where portfolio managers, dealers and risk managers are consuming and producing data in their pursuit of risk-adjusted alpha, and driving the market data purchasing and the financing, collateralisation, settlement, custody, fund accounting and transfer agency activities which also rely on efficient data flows. 


The vision is to build data management platforms that can capture and harmonise the data used and produced by all these different functions within an asset management firm and make any part of it available for use by any part of the firm, in a form that is not only accurate (or at least reliable) but fully consistent and reconcilable. 


Fragmented and duplicative data has defeated previous attempts to manage it


But achieving that end-state means overcoming the legacy of a past in which every part of an asset management firm pursued its own best-of-breed procurement strategy, purchased systems and services that suited the needs of individuals as well as individual departments, used bespoke messaging protocols and ran separate, siloed stores of data. 


One result is massive duplication, even within a single firm, of position, transactional and client data. This necessitates repeated reconciliations with external service providers as well as internal business divisions, but still achieves limited integration of the different data sets. 


An important consequence is to slow down innovation and investment in new services, because every alteration to one system or data set had knock-on effects on several other systems. 


Enterprise data management (EDM) platforms, which were designed to capture, validate, make consistent and then distribute within asset management firms a single set of consolidated data drawn from dealing desks, operations, risk management, customer information and accounting and financing functions, struggled to overcome the complexity. 


New technologies such as blockchain and the Cloud make data management easier


A better approach is simplification. Blockchain technology, for example, allows data to be shared between multiple users without duplication on different systems, repeated reconciliations and asynchronous messaging. 


By using Cloud and native Cloud services, asset managers can tinker with applications without having to re-design their entire technology infrastructure.


A new breed of data platform is transforming the management of data in asset management


Data platforms using these technologies, and conventional technologies, are now being built by global custodian banks. 


The platforms access the ABORs, IBORs and CBORs of asset management clients and the flows of data from their various service providers – namely, data vendors, index vendors, fund accountants, transfer agents and other custodian banks. 


Though these data management platforms are controlled by global custodian banks and/or large asset management groups, this is not deterring other global custodians and asset managers from working with them.


Though some are concerned about conflicts of interest, most asset managers take the view that they can piggy-back off the innovations of rivals. Likewise, most global custodians are content to make use of platforms run by rivals if their buy-side clients are enthusiastic. 


Indeed, the initial aim of the platforms is to create data sets that are sufficiently aligned to cut data handling costs within asset management firms. 


The next step is to extend the benefits of data consistency to interactions between asset managers and third parties, such as clients, potential clients and fund distributors, to make exchanges between all members of the eco-system more efficient. 


Not every data source is enthusiastic about sharing data


This will take time because information, as the old saying has it, is power, and each member of the eco-system is fearful of losing control of their portion of the data. 


This is especially true of third-party data vendors, which see data platforms as a threat to a highly profitable model built on selling views of data at high margins.


Even though the emergent class of data platforms is focused on managing data rather than creating or selling it (the global custodian banks in particular have abandoned the ambition they had a few years ago of re-packaging and selling the data they create in the normal course of business) they will be in a strong position to cut the prices charged by data vendors as they add scale.


But the reluctance to cede control is true inside firms, as well as between them and between them and their data vendors. This is why conventional outsourcing has not progressed further in the asset management industry. 


Data management platform are not another version of outsourcing


Outsourcing service providers such as global custodian banks that provide middle and back office services are trapped in an outmoded account-based view of asset management, in which it is good enough most of the time to deliver a snapshot of stocks and flows of data once a day, and that snapshot has no value to the front office. 


The data warehouse model of outsourcing that developed at the turn of the century accepted that there would always be multiple applications and overlapping and inconsistent but siloed data stores within asset management firms. It sought to solve this by placing all the data in one place and trying to make it consistent enough to apply to various purposes. 


Futuristic data management services start with individual transactions


Now, a new breed of technology vendor is trying to capture rich and detailed data at the level of the individual internal event or transaction, extract positions from it, and use event or transactional data as the foundation of a data management service capable of supporting multiple functions within an asset management firm.


It is an approach inspired by the distributed ledger of blockchain technologies. In some manifestations, it relies on tokenisation of underlying data, allowing it to be used without needing to aggregate it. 


In theory, this not only dispenses with ABOR, IBOR, CBOR and the other use-specific data sets but makes possible the provision of multiple views of the same data, intra-day, off a single initial view. The transactional data can be enriched later from external sources, such as data vendors. 


The data management platform which recognises that transactions begin in the front office proceeds from the same foundation and with the same end in mind: multiple views, not multiple databases. 


This is why the benefits of data management platforms do not stop at the division of labour achieved by conventional outsourcing (that is to say, the opportunity to concentrate on what the firm is actually good at) but offer more compelling gains.


The benefits of better data management accrue in product development and distribution


They include the transfer of the financial risk of data quality from the asset manager to the data management platform provider. This is a material benefit in an environment where a re-balancing error in an indexed fund cost a major fund manager US$105 million in compensation to investors.


Data management platforms also present asset managers with an opportunity to expand and differentiate their investment products, distribute them more successfully, and deepen their relationships with their clients. 


Data can also improve asset managers’ understanding of markets and market sentiment, and their ability to construct and then scale and automate portfolios that reflect that understanding. It can enable them to offer mass market products with the same level of customisation as they offer to their institutional clients.


But moving from the present (which is characterised by multiple Books of Record such as ABOR, IBOR and CBOR) to the future (in which there is only one Book of Record created from rich transaction histories) nevertheless entails confronting data sets that are more in the nature of disconnected heaps of information than searchable and well-connected databases. 


The sheer range of jargon and abbreviations that apply to databases in the asset management industry shows how new data management services are not being set up in greenfield sites. 


In addition to the ABOR, the IBOR and the CBOR, experts refer to the Trading Book of Record (TBOR, consisting of positions as adjusted by activity during the trading day), the Performance Book of Record (PBOR, which shows the performance of funds and portfolios against selected benchmarks) and the Universal Book of Record (UBOR, a self-reconciling summation of all data sets that can act as a single source for multiple views – something not yet extant). 


Then there is the Product Master (portfolios by asset class, investment strategy, investment product and investment vehicle), the Security Master (reference data, security identification numbers, corporate actions), the Entity Master (counterparties, clients and service providers) and the Risk Master (open trades and collateral needs). 


Each of these has to be understood by the data architects, in terms of its role in the workflows of an individual firm as a whole, because each asset management firm means something different by terms such as IBOR and Security Master. 


The various modes of operations of third-party data vendors, such as the suppliers of market data, indexes and risk management products, also have to be understood in relation to the individual asset manager, because each firm uses them differently. 


Indeed, different parts of each firm see the same data in different ways and use it in different ways at different times. 


Operations take an accounts-based view of the data, because they must reconcile it with custodians, while portfolio managers take a view based on investment strategies and sub-portfolios. Extracting the different views required by the various parties from a single set of data is not a simple task.


Fortunately, few of the internal databases, and all the services purchased from vendors, are not a source of competitive advantage or differentiation to an asset manager but merely a condition of being in the asset management business at all. 


If they can be reduced to a commoditised purchase by a platform instead, the price can fall. That is what data vendors fear, but it generates an immediate gain for platform users.


Data standards would accelerate progress but alignment of data models is easier


The larger gains from better data management will take longer to achieve, because of the legacy of multiple and ageing systems and disjointed legacy databases. Progress could be accelerated if data standards could be agreed and were quickly and widely adopted, but standards consistently prove elusive. 


An IBOR Standards Working Group (ISWG), for example, succeeded in devising standards for the IBOR alone. Unfortunately, adoption proved harder than design. It did not help that the centralise storage and processing technologies available were not adequate to the task of continuous, live extraction of data. 


Blockchain-based data designs – which are well-adapted to storing transaction histories and extracting position data from them – and Cloud storage can overcome these technical obstacles.


In fact, Cloud storage allows data management platforms to create different views of the same data not just across different parts of the same asset management firm but across different asset management firms.


That is because different firms can hold and blend sharable, non-proprietary data in the Cloud. They can use it to develop apps for in-house use, or to augment internal data, once they are convinced the data is consistent, reliable and well-defined. 


Indeed, the benefits of drawing on well-defined but collective data sets could spark the evolution of an alternative to standards, as firms develop the habit of aligning their data models in the Cloud simply because they find it advantageous in their own business. 


LEIs are a standard that is proving useful in organising data 


One standard that is proving useful to asset managers is Legal Entity Identifiers (LEIs). Although even LEIs are being adopted slowly and patchily, managers use them for counter-party risk assessment and management, identifying the same clients in different data sets and for regulatory reports in which they are, for example, required to prove that they have identified beneficial owners. 


LEIs also contribute to unravelling the ownership structures, not just of counterparties and service providers but of the companies in which funds invest. Matching LEIs to client account structures and legal agreements is also valuable in maintaining contractual discipline. 


Some asset managers might choose to assemble their own data packages


Eventually, however, LEIs are likely to become part of digital identity services asset managers buy from third parties. In this sense, they may become one of many data services purchased from third parties in the same way as asset managers buy market data today, for example. 


Purchasing data streams might even develop into a separate evolutionary path for asset managers to that offered by the new class of data management platforms. Instead of building everything in-house, or relying on a platform, asset managers might achieve greater agility by assembling packages of independent data services that can be integrated.


That might appeal to asset managers reluctant to embrace a data vendor model or a data management platform run by a competitor or a global custodian bank they do not use. But what is clear is that no manager can afford to do nothing about data. 


Managers that fail to tackle the problems caused by fragmented and duplicative data, let alone fail to seize the opportunities presented by clearing them away, are at risk of being displaced by new entrants armed with superior technology. 


In other words, the crucial questions about the data opportunity in asset management are not about whether but how improved data management will be accomplished – how long it will take, how much it will cost, who will do the work and which managers adapt and which do not. 


Adaptation is crucial because, once the data is clean and consistent, artificial intelligence (AI) and machine learning (ML) can not only effect significant cost reductions by raising levels of automation, but enhance portfolio investment and risk management decisions by offering superior predictions of their likely outcomes. 


In other words, data-driven asset managers will start to outperform, in terms of profitability and investment performance.


Questions to be addressed at the next Data in Asset Management discussion
1. Is more efficient data management delivering on its promise of lower costs?
2. How much progress have open data management platforms made?
3. Are data vendors changing their behaviour?
4. What new use-cases are being found for more efficient data management?
5. Are data standards irrelevant to progress in data management?
6. Is regulatory reporting data likely to be made available for commercial use?
7. What progress have asset managers made on cost disclosure?
8. Can tokenisation and fragmentation tools help in data management?

1. A so-called DataNet, which allows data to be shared while keeping it secure and anonymous by using universal identifiers and federated learning rather than direct access will also facilitate data sharing. See https://www.futureoffinance.biz/b/the-promise-and-the-power-of-digital-identity-meeting-summary-

If you would like to participate as a panellist please contact Wendy Gallagher at wendy.gallagher@futureoffinance.biz
If you would like to participate in the audience please let us know below or contact Wendy Gallagher on the email above
If you would like to participate as a sponsor please contact Valerie Bassigny on valerie.bassigny@futureoffinance.biz