Smarter regulation through data science

Historical analysis shows that several economic and financial crises could have been reduced in severity, or even avoided, if people had been able to collect detailed data and unlock its narratives through advanced analytics. Today, due to recent technological innovations in methods to analyze data, we have entered a time when we have access to data on virtually everything happening around us. It is the time of data.

Leading up to the mortgage crisis and Great Recession, oversight was lacking at every stage of the mortgage life cycle. While detailed data was being collected on nearly all loan-servicing activities, the data was not being used to effectively oversee servicer actions and improve loan performance. Questions remain as to why this was the case, but since house prices were shooting up and credit losses were low, the mortgage industry did not feel the urgency to invest in technologies and innovative analytics to improve oversight. Had the data been analyzed appropriately, signs of the impending crisis could have been identified and mitigated much earlier.

Another (and much-publicized) example comes from the Madoff investment scandal. A decade before the scandal broke, analyst Harry Markopolos investigated Bernie Madoff's results using mathematical finance models, and quickly concluded that these results were fraudulent. Despite Markopolos' best efforts, his models were shunned by the SEC. Smarter regulation, using data and structural models, could have saved a lot of later heartache.

Today, a looming predicament surrounds the housing market. Due to the U.S. political climate, regulations that protect consumers are expected to be reduced or even eliminated through upcoming policy shifts. However, in areas such as the housing market, these protections are necessary to prevent a repeat of the recent crisis. How can we thus effectively manage oversight with limited resources? 

The Solution: Data Science

Fortunately, recent advancements in technology and data analytics now allow for industry monitoring to occur without allocating significant resources. There's an enormous opportunity to swap regulated activities, which are already at risk of being cut, for smarter data oversight.

Data analytics can be applied to monitor banking and lenders in a more cost-effective and efficient way, enabling regulators to analyze available data, pinpoint areas of concern and "course-correct" using data insights. This ultimately creates a smarter way to regulate.

The regulatory agency, the regulatory process and the regulated industry can all benefit from an analytical framework that oversees regulation.

This framework includes the following:

• Enhancing efficiencies for regulatory agencies: Data-based regulation streamlines time, costs and processes. Smart analytics methods leverage cheap and secure computing infrastructures; readily available quality-assured data; and a burgeoning, well-trained workforce of data scientists. Actionable data inferences can be available almost as soon as data is updated, and shared easily through web portals.

• Ensuring a fair and transparent regulatory landscape: Successful oversight of multiple banks and lenders requires fairness. This can be achieved through transparent and objective analytics techniques. While more appealing to many data scientists, overly complex and opaque black-box models often fail as regulatory tools since practitioners are reluctant to adopt them. The analytics frameworks adopted by Fannie Mae and Freddie Mac for their servicer scorecards provide fair oversight by recognizing each servicer's unique loan portfolio. Yet these scorecards gained traction only because they were also straightforward, transparent and objective, so servicers could thus recognize their fairness.

• Improving real-time regulatory process oversight and feedback loop: Smart oversight analytics incorporate the ability to pinpoint where a regulated entity needs to improve performance. Smart tools not only monitor regulatory processes but also provide feedback to improve performance. Lenders often work with loan-level data. Providing tools that not only provide aggregate inferences but also enable a deeper detailed view of an individual loan allows lenders to focus resources on underserviced areas, leading to improved efficiency.

We are in the time of data, and every day there is more of it. By leveraging the power of this data, we allow it to help us build smarter oversight and regulation processes that can be less costly, less polarizing and more secure for consumers as well as markets that depend on oversight for their stability.

The key is knowing what data is available and how to use it. Only then may data be truly harnessed as a powerful solution to benefit the nation, our government and our citizens at large.