Since the events of the financial crisis, economists and industry analysts from both the public and private sectors have been working to determine the true cause of those painful financial events. Many have theorized ways to better minimize or even prevent the next catastrophe. Many different approaches have been put forward, and this is a good thing. But one fact has emerged that is proving to be foundational to all of the approaches.
No matter what solutions are agreed to; no matter what laws change; no matter what new regulations are introduced – none of the solutions that are being discussed and implemented will be successful without accurate, timely, and complete data. Data is the lifeblood of our economy. It is the key factor of input into our business functions. It is the key indicator that tells us about the health and wellness of our financial system.
Simply having data, however, is not necessarily the solution. In many cases, more is not better. During the financial crisis, banks and regulators had plenty of data, but the data they had, in many cases, was not actionable. Data was not harmonized. It was not consistent and it was not comparable. It could not easily be aggregated, manipulated or analyzed. To paraphrase the Rime of the Ancient Mariner, “data, data everywhere, nor any drop to drink.” The data that was available to the banks and to the regulators before and during the financial crisis simply did not satisfy the requirements needed to fully understand the interconnectedness and complexities of the financial system. In many cases, data was not collected and captured in a methodology and format that would enable decision makers to effectively utilize this data to determine both individual firm risk as well as holistic, systemic risk. Few analysts or economists would tell you that data was the root cause of all the financial troubles. However, many in the financial industry, academia, and the regulatory community recognize the impact that a lack of harmonized data had in making the crisis worse.
The collapse in the mortgage market was at the center of the financial crisis. For various reasons, the strict criteria for mortgage approvals were eased. The standards were weakened. Individuals were being approved for mortgages that in the past, would never have qualified. The fundamental criteria and operating model of the mortgage business had changed. Mortgage companies, once anchored as risk-driven businesses, became fee-driven businesses – and they approved as many mortgages as possible. And the risk – who carried the mortgage once it closed – was not longer an issue. The mortgages were removed from the balance sheets almost as soon as they closed and were sold to other institutions.
Now here’s where the data part gets interesting ...
As these mortgages were sold to the large banking institutions, the individual loans found their way into pools, which found their way into collateralized instruments, which found their way into a variety of derivative products. And all along the way, bits and pieces of the fundamental data about these instruments disappeared. Packing and re-packing, slicing and dicing, then packaging again, created a wedge of abstraction between the original collateral and the newly created investment vehicles. The supply-chain of data was broken.
In essence, the descriptive information about the underlying instruments became so abstracted from its source that it became increasingly difficult, if not impossible, to truly assess the value of the financial instrument in hand. A lack of consistent, normalized data from source to financial instrument made understanding the holistic risk of these instruments nearly impossible as the underlying collateral deteriorated. Financial executives did not have the actionable data they needed to assess and react to changing market conditions.
And then, the unthinkable started to occur – banks started to fail.
As banks began to fail, panic struck. What will be the impact of a bank failure? How will other banks be affected? How extensive will the damage be? And individual banks began to ask: what is our exposure to this impending systemic risk?
One of the critical data requirements for systemic risk analysis is being able to uniquely identify entities and define their hierarchies. Risk managers at a financial institution must comprehend the complex relationships of a legal entity and all its subsidiaries to understand the relationships between parent and sub-entity to fully grasp the systemic relationships and risks. Without this relationship transparency, properly assessing a systemic risk event is very difficult.
On that dreadful Saturday in September 2008, word spread that Lehman Brothers might not find a suitor, and that the unthinkable may happen: Lehman may fail.
Technology and operations teams streamed into their respective banks across the globe. Data crunching began. What was the exposure to Lehman? Billions of dollars were waiting to be cleared. Without a clear view of financial exposures, banks were unsure if they would be transferring money to bankrupt entities.
As banks across the globe poured through data on exposures to Lehman Brothers, the realization was immediate: most financial firms did not know their actual exposure to Lehman Brothers.
There was no clear understanding of the Lehman organizational structure, and there was no clear understanding of which of those entities would be declaring bankruptcy. Was it the U.S. division only, or the U.K. division? Was it Lehman Brothers Holdings? What about the subsidiaries like Neuberger Berman that didn’t even have the name “Lehman” in their title? Risk managers were scouring millions of transactions scattered across numerous trading and accounting systems with multiple identification schemes and disparate data descriptions.
When the banks finally pulled together the information they had on the Lehman entity structure, they encountered the next data hurdle. With deal structures so complex, and the roles and responsibilities that counterparties play on each deal not easily discernable, what deals would be affected by a Lehman bankruptcy? Remember, it was just as bad to not send dollars to legitimate entities as it was to send dollars to bankrupt entities.
Legislators, regulators, and Wall Street decision makers alike did not have actionable data to fully realize the systemic impact of a Lehman Brothers failure.
What Did We Learn?
No matter how all these events occurred, the financial industry’s ‘portfolio’ ended up immersed in excessively complex financial instruments without a clear understanding of exposures and risk.
In January 2013, the Basel Committee on Banking Supervision published a paper entitled “Principles for effective risk data aggregation and risk reporting,” now known as BCBS 239. The paper’s introduction states:
“One of the most significant lessons learned from the global financial crisis that began in 2007 was that banks’ information technology (IT) and data architectures were inadequate to support the broad management of financial risks. Many banks lacked the ability to aggregate risk exposures and identify concentrations quickly and accurately at the bank group level, across business lines and between legal entities. Some banks were unable to manage their risks properly because of weak risk data aggregation capabilities and risk reporting practices. This had severe consequences to the banks themselves and to the stability of the financial system as a whole.”
In short, BCBS 239 is a mandate to banks and regulators alike to get their risk and data infrastructures organized, standardized, and modernized so that they are better positioned to prevent – and, if need be, respond to – the next financial crisis.
All of the major financial institutions will be subjected to this requirement. So you can see, accurate, timely, complete, and harmonized data is no longer a “nice to have” – it is a must have. Our best defense against the events of 2007-2008 happening again is anchored in data. Data is our fire alarm. Even if another crisis occurs, timely, accurate, and harmonized data will be needed to make informed and sound decisions. Without quality data that is actionable, the industry forfeits its ability to respond.
So how do banks address this challenge? How does the industry begin to recognize the importance of data and the need for a sound, methodical approach to managing it?
First and foremost, data must be viewed, across the enterprise, as a strategic enterprise asset. This may seem like a simple concept, but it is a critically important one. The Data Management Association International puts it this way:
“Assets are resources with recognized value under the control of an organization. Enterprise assets help achieve the goals of the enterprise, and therefore need to be thoughtfully managed. The capture and use of such assets are carefully controlled, and investments in these assets are effectively leveraged to achieve enterprise objectives. Data, and the information created from data, are now widely recognized as enterprise assets. We have to break the tendencies to stove-pipe data within business verticals and start to view this enterprise asset as a horizontal utility.”
Tom Peters, author of the best-seller “In Search of Excellence” and numerous other business management books, describes the criticality of data as an asset in this manner: “Organizations that do not understand the overwhelming importance of managing data and information as tangible assets in the new economy, will not survive.” Once a bank understands the importance of data as a critical enterprise asset, it is only natural to understand the importance of managing it methodically and effectively.
A bank that embraces this concept and recognizes the holistic importance of data in its daily operations can gain time, cost, and data quality efficiencies. There are measurable improvements in time-to-market for new products if a financial organization can view harmonized data as a common service across the operations of the institution. Although there are technology hurdles to jump, this is a far broader challenge than technology alone. Making the data and data management part of an organization’s DNA is a challenge of mindset. It is a challenge of culture change and it is a challenge of organizational realignment.
One way to change the mindset is to have banking leaders start to think like executives at data companies. Banks acquire and analyze data in support of their clients, in order to bring them the best possible solutions for their investments. They have a responsibility, therefore, to manage the bank’s data in the most effective and sustainable way. To put it another way: imagine any financial services organization operating without accounting principles, or operating without a technology plan. Data management must be viewed in the same manner – integral to a firm’s every day activity.
The Evolution of Data Management
So once banks recognize data as an asset, how do they begin to harmonize data? There are two basic concepts to achieving this objective.
First, data must be “engineered” according to industry standards with unambiguous shared meaning and standardized identification. The BCBS 239 states in its second principle:
“A bank should establish integrated data taxonomies and architecture across the banking group, which includes information on the characteristics of the data (metadata), as well as use of single identifiers and/or unified naming conventions for data including legal entities, counterparties, customers and accounts.”
Harmonized data is harmonized meaning. Data management is not just about technology. The financial industry has been very good at the technical “movement” of data: acquiring, storing, processing, transforming, and delivering data. And, as the processes and technology evolves, the industry continues to do all of these things faster and cheaper. But collectively, the financial industry has not been that good at understanding the meaning of data. Data management is about the content and standards. It’s about normalizing data so we can better analyze, evaluate, and draw conclusions. With unambiguous shared meaning across our operations we can fully understand the essence of the data. Only then can we be better prepared to aggregate our data, and provide this important perspective to the regulators who are tasked with ensuring our system operates safely.
To see where this is working, one doesn’t have to look farther than Silicon Valley. Firms such as Google, Amazon and Pandora are driving this data philosophy and using it to achieve new innovations every day. They are constantly discovering new ways of looking at their data, modeling it, and linking and relating their data such that it becomes the most important source of insight into their products, services and their customers.
The good news is there is an effort underway across the finance industry today, driven by the industry trade association, Enterprise Data Management Council (EDMC), to establish a harmonized language for financial instruments and entities. FIBO (Financial Industry Business Ontology) is a logical representation of all financial instruments and entities into an ontological model. Engineering data in this fashion will move firms to this “next generation” of data engineering that abstracts data from the physical databases and represents it as real-world objects. By ‘unshackling’ data meaning from the physical database, data engineers and data scientists can create logical knowledge maps that represent how data really relates and behaves. Using FIBO enables flexibility and insight into the complexity of the financial world. Initiatives such as FIBO will be critical as banks move toward better understanding of the risks and opportunities that data presents.
Embracing the responsibility of managing data via best practices, also addressed in BCBS 239, is the other crucial component. The Chief Data Officer has emerged as a new role at many financial services organizations over the past 10 years. If you go back no further than five years, you would have found at most, a handful of CDOs. But today, this role has dramatically expanded. Most major financial institutions have a CDO and in other industries such as insurance, healthcare, and manufacturing, as well as the public sector, CDOs are now more common.
Yet, with all of these new appointments, we still do not have a standard approach to data management. But we are seeing an emergence of best practices. These ensure CDOs approach data management in a methodical and structured way, such that the data is accurate, timely, complete, and a trusted factor of input into the most important and critical business functions.
Best practices help establish and sustain an effective data management program. They identify the capabilities needed to define a strategy, build a business case, formulate a governance structure, define data architecture, and drive data quality. Best practice approaches move financial institutions from “data management on the fly” to data management as an organizational construct. Best practice approaches are about changing the culture of organizations such that data management is not seen as a project, but becomes a daily habit.
We see examples of best practice benefits across many industries today. Medical, accounting and manufacturing are great examples of how best practices have propelled these disciplines. Medical services are driven by established best practices in patient care. Accounting is driven by industry and regulatory best practices to assist firms in satisfying their obligation to manage their internal finances. And buildings are constructed following established and proven best practices regarding how various structures are raised, what materials are used, and how critical building infrastructures are designed. The same benefit can be realized when firms follow best practices in managing their data. By applying consistent and proven best practices to how critical data is sourced, processed, stored, cleansed, delivered and consumed, we realize benefits in overall data quality, flexibility and usefulness of our data across all of our business functions.
In conclusion – it’s all about the data. Data is the raw material. Engineering can make it actionable. Best practices can change industry culture so that business leaders take responsibility for it. We live in an information world. According to IBM, 2.5 quintillion bytes of data are created every day! So much that 90 percent of the data in the world today has been created in the last two years alone. The financial industry has a choice – master this asset to its greatest advantage, or be drowned by it. When it comes to the health and well-being of the financial industry, I choose the former.