COVID-19 Broke Your Risk Models. This is How External Data Can Fix Them.
The previous decade has been one of impressive growth and massive profits for financial services providers. As the global economy bounced back from the recession, new regulations, improved technology, and a friendlier landscape for lenders meant that financing and loans were easier than ever to come by. Over ten years later, it took only three months to undo most of that progress. The new post-COVID-19 reality has rendered most of our old financial assumptions useless.
Companies that were financially stable and paid their loans back on time every month suddenly find themselves shuttered; borrowers with sterling credit may now be high default risks due to factors completely outside their control. For lenders, this new reality means that the data and models they used to predict risk are completely broken.
How can you identify a great loan opportunity when the data you have doesn’t account for factors like whether a company is facing extended furloughs, is closed temporarily, or even if it is in an area that has a high incidence of coronavirus cases? Unfortunately for lenders, pre-COVID-19 data is no longer useful in a post-COVID-19 world. How, then, can we fix it? What avenues are there to redefine our models so that we can continue lending while mitigating our risk?
How Explorium is helping our customers adapt with COVID-19 data
Instead of continuing to rely on models and data that are already outdated in our new reality, Explorium is helping our customers adapt and react to the new landscape. Especially in the lending sector, we understand that gaining as much relevant information as possible is key to making the right decisions. To do so, we’re connecting our customers’ data to an ever-growing list of COVID-19 related sources that are relevant today, to help them make better decisions.
One of our customers, a global small business lender, suddenly found that its risk models were invalidated by the outbreak of COVID-19. Companies with perfect financial records were forced to shut their doors, and entire regions were put under lockdown, placing even the strongest organizations at risk of default. To better understand their risk, our customer needed to view the problem with new eyes — and more relevant data.
We knew that our customer needed better visibility not just into the companies themselves, but into the factors that were now affecting their finances and default risk. As such, we approached the problem from three sides — the broader, industry scale; the company level; and the “crisis” level.
Our first task was to provide better context about the situation, so we looked at industry signals, including:
- Stock market indices for the industries borrowers operate in
- Government signals regarding essential industries, operating restrictions, and business closures
- Web conversations around particular industries, as well as the sentiment surrounding them
Next, we focused on the company level by enriching their data with:
- Internet signals that included web traffic to particular domains, social media, and other conversations surrounding companies
- Domain information including how many times a company is mentioned, as well as the context
- Alternative financial data
Although we knew that this may be enough in normal times, the current crisis requires another level of visibility. So, we also connected our customer to a variety of COVID-19-related datasets. This gave them new, intriguing avenues to augment their predictive capability, including:
- Understanding the direction and magnitude of COVID-19 spread across countries
- Geospatial information about which regions were being affected
- Restrictions and closures at a national level, such as shelter-in-place orders, work-from-home regulations, and even full lockdowns
- Economic data surrounding the pandemic, such as unemployment numbers, hiring and firing data, and limits on business operations
- Data on restrictions being loosened, which can indicate a rebound in the economy
This new, enhanced model completely redefined how our customer modeled risk. Instead of focusing almost exclusively on internal data such as company financials and historic data, our customer now has a new perspective on which to measure default risk. A perspective empowered by data-driven risk management.
They can now account not just for a company’s record, but how their geographic location, their governments’ actions, and even how people’s sentiments about their industry and the pandemic itself could impact their ability to do business. Even in these unpredictable times, our customer can still provide an essential service without putting themselves at risk of making bad loans or being stuck with the bill on loans they can’t recoup, and data-driven risk management is helping them do that.
Prepare for any event with data-driven risk management
It’s impossible to predict crises such as the one we’re facing, and companies’ entire outlook and risk models may be thrown out the window without a moment’s notice. However, this doesn’t mean lenders and financial services providers should be passive about it.
Continuing to rely on business-as-usual methods in abnormal situations is a sure way of ensuring a collapse. On the other hand, seeking out the best ways to bolster your models and find clarity in the most unpredictable of times can help you not just survive, but position your organization to thrive. The answer to today’s uncertainty is to find the best data to complement your existing datasets and information. We’re happy to be able to support our customers with a data-driven risk management solution to do just that.