Batten Down the Hatches: Preparing for the Downturn with a Data-Driven Approach to Debt Collection

They say that the water level drops precipitously in the moments right before a tsunami hits.

For lenders, it feels eerily similar to this right now.

Back in the spring, at the early onset of the COVID pandemic, banks scrambled to prepare for the worst. The Big 6 reserved over $10 billion in Q2 to cover for the inevitable increase in loan defaults. Collection operations moved quickly, at first to suspend outbound campaigns and provide relief programs to debtors. Once things stabilized, the focus shifted to ramping up staffing in preparation for the expected rise in delinquencies that would occur as a result of the devastating impact the virus has had on employment and the economy.

But what has actually happened so far is something that no one saw coming, and certainly not me.

The last few months have been banner ones for collection shops. Many lenders are reporting collection levels at all-time highs, and delinquency rates have dropped dramatically.

What on earth happened?

“Unprecedented” Response

Typically, unemployment rate is the single biggest predictor of credit delinquency rates. Millions of Canadians lost their jobs in March and April, and while many have begun to return to the workforce, unemployment is still at recessionary levels at 10.2% in August.

But unprecedented times require unprecedented measures. The response by the Canadian government and the banks has had a massive impact, as shown in the recent benign risk metrics and strong collection results. This relief has come in two forms:

Payment Deferrals – In concert with the government, the Banks launched comprehensive hardship programs for those needing support. These consisted of payment deferrals of up to 6 months and reductions in interest rates on credit cards. Over 16% of the mortgages in the banks' portfolios has taken one of these deferrals or skip-a-payment offers.

Government Transfers – The Canadian Emergency Response Benefit (CERB) has provided a taxable benefit of $2,000 every 4 weeks for up to 28 weeks to eligible workers who have stopped working or whose work hours have been reduced due to COVID-19. As of September 13th, 8.8 million Canadians had applied and received $78 billion in relief. In addition, the Canada Emergency Business Account (CEBA) program provides interest-free loans of up to $40,000, of which $10,000 is forgiven if the loan is repaid by the end of 2022. As of September 17th, more than 750,000 CEBA loans amounting to $30 billion have been handed out to small businesses.

The result has been an 11% increase in disposable income in 2Q, as the government transfers more than offset the loss in wages during this period. This, combined with a 14% decline in household spending during the shutdown, has created a “savings spike”. And it is this increase in disposable income the past few months that has led to the increase in collections and the decrease in delinquencies.

The Calm Before the Storm

But while it is heartening to see Canadians being fiscally responsible during this time, it is also very likely the calm before the storm. All indications are that this is only a temporary effect:

·      The CERB program is ending September 26th. As this wage subsidy ends, the full effect of double-digit unemployment will start to be felt.

·      Household debt remains at record levels. Canadians still hold $2.3 trillion in household debt, of which $1.5 trillion is mortgages and $800 billion is consumer debt.

There is also still a lot of uncertainty in the market as to what the long-term impact of COVID will be on the economy and credit conditions. As I think about this, I have three overarching questions:

·      Labour Market & The Economy – “Will it be a V-shaped or a W-shaped recovery?”

·      Stimulus Unwinding – “How much credit risk is being masked by CERB and payment deferrals?”

·      The Threat of a “Second Wave” – What would be the impact of another shutdown?”

I don’t profess to have a crystal ball, but what I do know is that a data-driven approach can help you navigate through this period of uncertainty and prepare your collections business for the downturn.

A Data-Driven Approach to Collections

We are in the midst of a digital revolution, and data is the lifeblood of this new digital economy. Things like AI, automation, and next-gen analytics are being leveraged to re-engineer our workforce and re-define our core industries. Almost every action now leaves a digital footprint and generates data that can be used to build new and innovative products and experiences.

Yet, most collection operations are still stuck in the analog world. Particularly in an outsourced environment, collection operations have not fully invested in digital capabilities and have not been very sophisticated in their use of advanced analytics.

Why is this? Well, historically there have been structural barriers that have been impediments to a data-driven approach:

Lean Operating Model - Intense competition for business has led to margin compression in recent years for outsourced collections operations. Historically, data and analytic capabilities would require significant capital investment, a difficult proposition to bake into winning contract bids. However, with the sharp decrease in cost to store and process data, this is becoming less of an issue over time.

Lack of Data Sharing - Many clients are only willing to share the absolute minimum amount of data to service the account in collections. Even when there is a willingness to share data, there are often technology barriers that require integration development to be prioritized on the client side (and paid for by one or both parties).

Misaligned Incentives - Flat commission compensation schemes are very common, particularly in the later stages of collection or recoveries. This creates incentive to optimize simply on total dollars collected, rather than prioritizing based on the relationship between action and incremental payment lift.

Short-Term Focus - The monetization of a data-driven approach occurs over a long horizon; it takes time to first establish data capabilities and then to measure and adjust strategies. Collections contracts are often designed with monthly or quarterly business shifts, placing the emphasis on urgency over importance.

Fortunately, many of these barriers are dissipating over time, and since they are unique to third party providers, do not limit the capabilities of internal collection teams. There are so many use cases in collections where data and analytics can optimize decision making throughout the customer lifecycle.

A useful framework for identifying points of leverage is what I call the "Collections Conversion Funnel”.  

The Collections Conversion Funnel

Capacity – There is a distinct first mover advantage in collections when a downturn is approaching; it is a high churn labour pool and it is more difficult to staff up quickly when the competition is also hiring. By developing sophisticated forecasting models for delinquency inventories and their correlation to optimal staffing levels, you can avoid a “hiring squeeze” when credit deteriorates.

Segmentation –The most common dimensions for segmenting accounts in collections are days past due, balance or amount owing, and perhaps a risk score like FICO. There can be a lot of juice in leveraging decision sciences to get more granular, such as: conducting experimental design testing for different treatments on test vs. control populations; developing discounted cash-flow models (NPV/LTV) to measure the financial impact of a given action for a particular customer segment; building ML models to predict key outcomes, such as Right Party Contact (RPC) or Cure within 30 days.

Contact – Automated dialer systems generate millions of data elements that can be optimized through advanced analytics to increase RPC rates. Nuanced insights such as the best time of day to contact a debtor can be used to drive dialer strategy; for instance, a retired person or a student might have a higher RPC rate during regular working hours than someone in the workforce. Properly designed tests can also optimize channel and identify opportunities for a digital-only strategy, saving significant operations expenses.

Offers – During a downturn, providing collection agents with more tools and offers is important, as it can help you move up the payment hierarchy and get more Promises-to-Pay (PTP). Having well-designed tests for a range of settlement or hardship offers, combined with NPV models to analyze the trade-off of interest reduction versus payment lift, can increase profitability. It can also provide more flexibility to your customers to help them through the downturn, so you can retain more customers on the other side.

Conversion – Agents are typically incentivized on outcomes, such as dollars collected, rather than on actions, such as negotiation skills or customer empathy. Leveraging data to design a more sophisticated performance management system, with personalized targets for all metrics on the conversion funnel, and combined with coaching feedback loops focused on what can be controlled– i.e. the actions – can produce a massive increase in collector performance.

Getting Started

In a data-driven decisioning system, data is the oxygen, and oxygen must be protected at all costs. Thus, the foundation of any data-driven approach starts with upholding to a high standard of data quality:

Clean – Data lineage should be mapped and documented to avoid “garbage in, garbage out.” Automated data quality checks should be programmed to run on a frequent basis.

Accessible – Data often resides on multiple, disparate systems that do not talk to each other. Investing in a centralized data lake will allow for easier building and deployment of models and segmentation strategies. Building in real-time replication to replace batch jobs will enable better real-time decision making.  

Compliant – Collections operates in highly regulated industry groups; it’s imperative to have a robust control system and testing program for privacy and other applicable regulations. Internal policies on permissible use should also be created for gray areas, such as social media.

Payson Solutions is here to help. Since 2018, we have helped dozens of companies across the globe realize the power of their own data, serving a diverse client base that includes top-tier banks, credit unions, fintechs at both start-up and growth stage, collection agencies, and alternative lenders. We are on a mission to democratize data. Our business was built on the fundamental belief that data democratization can be a massive advantage for any business – and we can help them get there.

Here’s hoping for a light rainstorm, rather than a tsunami; but prepare for the worst, and don’t wait! Get started today.  

Share your thoughts

About the Author

Brent Reynolds

After a long career as an executive in financial services, I started my own company, Payson Solutions, to help companies transform their business. I have a passion for building high-performing teams and leveraging advanced analytics to build amazing customer experiences. If you would like to connect, drop me a line at brent@payson.ca.

View this post
on Medium
Data solutions for your business
See Our Solutions