Skip to main content

What next for US credit-card debt?

By taking care of customers, lenders can help reduce the severity of the looming economic crisis and build deeper relationships that will endure through the cycle.

Americans are experiencing firsthand the immediate humanitarian impacts of the COVID-19 pandemic. As of April 27, 2020, WHO reported nearly three million confirmed cases worldwide, with the United States in an unenviable leading position, with nearly one million. But the longer-term impact on US economic health is only beginning to come into focus.

Despite the US Congress–approved $2 trillion stimulus package, which included $350 billion in funding to help small businesses stay afloat over the next few months, the average US consumer will receive more modest assistance. Households in which adults earn less than $100,000 per year will receive, on average, around $3,000 to combat the effects of rising unemployment and economic hardship. By contrast, the average credit-card debt of households currently stands at around $8,400. 1

The latest estimates from US Department of the Treasury Secretary Mnuchin predict the country’s unemployment rate rising as high as 20 percent in the coming year, while other forecasts include a worst-case scenario of more than 30 percent, 2 well in excess of the 10 percent seen immediately after the recession in 2009. Inevitably, those worst affected will also be those least able to afford a shock to their monthly finances. 3 Consequently, we expect consumers’ ability to service their credit-card debts to drop markedly with the rise in unemployment.

The decline in consumers’ ability to pay off their credit-card debt will coincide with an emerging shift in lenders’ attitudes and practices. Whether directed by federal or state authorities or under prudent self-governance, lenders are rapidly scaling back outbound collection-calling activities, especially in the areas worst affected by the spread of COVID-19, such as New York. Furthermore, those lenders without a robust work-from-home infrastructure and a digital omnichannel strategy will find it increasingly difficult to contact customers through traditional channels.

Unless the accounting rules regarding how losses must be recognized are amended, all signs point to a wave of losses crashing over the US retail-lending landscape over the next nine to 36 months. In the short term (three to six months), according to loss forecasts by McKinsey, the current stock of medium- and late-stage delinquent accounts is likely to charge off almost in its entirety, leading to losses of $15 billion to $25 billion. Longer term, the impact of an up to 20 percent unemployment rate will render those customers already struggling to manage their revolving debt largely incapable to do so and will spark an extended period of elevated losses, which could total an excess of $130 billion over the next two years. Lenders urgently need to assist their customers.

Fortunately, lenders can take actions today that will mitigate the impact of the various forces on consumer debt. These actions to “change the bank” need to be taken by a plan-ahead team in parallel with a nerve-center team solving for the immediate demands to “run the bank” by prioritizing good customer assistance and fair outcomes for those most in need. The effects will be felt differently according to the response of different lenders in the market. Specifically, the operational sophistication, speed, and agility of lenders will directly affect their ability to contact customers and keep them engaged to maintain good financial habits at a time when public messaging suggests customers should expect broad relief from servicing their debt.

Four factors will determine the future of US consumer debt

To model the impacts of COVID-19 on credit-card losses in the United States, we anticipate the future of US consumer debt will be directly affected by four factors.

1. Collection policies

We expect state or federal policies or self-regulation to curtail outbound collection activity significantly over the next few months.

There is mounting pressure from government agencies to restrict outbound collection activities. We have seen several examples, including the following, in recent days:

  • Countrywide restrictions. Indian and Greek governments have recommended the suspension of all installment loan payments for three months. The US government has announced an immediate suspension of collection on federal student loans until September 30, 2020.
  • State-based restrictions. New York State has implemented a 30-day halt of collection on medical and student debt.
  • Product-based restrictions. We have seen the suspension of collection and foreclosure activity on all government-backed student loans and mortgages.

We have also witnessed proactive moves, including the following, from many lenders to self-regulate the adverse effects of collection activities on their customers:

  • Self-imposed moratoria. Bank of America, JPMorgan Chase, Wells Fargo, and others have placed moratoria on mortgage collections in California. American Express, Apple Card, Capital One, and others have offered suspensions of loan payments.
  • Limitations on specific activities. There have been limitations on activities such as wage garnishments and offsets. There have also been expansions of payment-relief programs (for example, by Ally Bank and others).

2. Lender capabilities

In the immediate aftermath of COVID-19, many lenders are already operating only at partial capacity and may struggle to make outbound calls:

  • In the immediate aftermath of COVID-19 infections and physical distancing, many lenders find themselves short-staffed simply because of staff taking extended time off to care for children, friends, relatives, and other loved ones. In the last two weeks of March 2020, we witnessed, on average, more than 50 percent staff absences across collection-operations centers.
  • Many centralized call centers have effectively shuttered and been replaced with various forms of work-from-home measures (with varying levels of success). For example, operations dependent on offshore facilities in India—where contact centers have been closed since March 24, 2020—have been severely affected. Not all US lenders have telephony services that will function remotely.
  • With limitations on outbound calling, lenders must rely more heavily on digital channels and self-serve, but many do not have mature or well-orchestrated digital channels or self-serve platforms.

3. Customer ability and willingness to pay

As the economic consequences of physical distancing and prolonged widespread isolation emerge, it is clear that consumers’ appetite to service their debts will be directly affected both by the health of the economy and by the widespread messaging regarding their responsibilities as borrowers. Scenarios from our published macroeconomic forecasting help set a range for potential paths for credit-card delinquency. We consider two scenarios:

  • Virus contained (scenario A3). This scenario forecasts GDP recovery in the United States to precrisis level by the end of the fourth quarter of 2020 and a short-term six- to nine-month pulse of elevated delinquencies and losses, with aftereffects subsiding by early 2022.
  • Muted recovery (scenario A1). In this scenario, US GDP does not recover until the end of the first quarter of 2023. There is an extended 24-month period of initially high delinquencies, with gradual decline after lending rules inevitably tighten by late 2022.

Multiple actions already announced in the media may reduce the likelihood that customers will be willing to pay down their credit-card debt:

  • Payment relief, including deferred payments, may break good habits for customers regularly making payments on revolving balances and increase the likelihood delinquent accounts will age to loss.
  • Any signaling on the prioritization of debt payments may lead to secured loan payments (for example, for auto and real estate) being made first, at the expense of unsecured credit-card payments.
  • Suspension of credit-bureau reporting eliminates a significant and lasting consequence for customers choosing not to pay off debt.

The scale of unemployment across the United States will inevitably prove to be the biggest driver of customers’ ability to pay. However, lessons from the last recession also suggest that customers will make deliberate choices to prioritize servicing certain types of debt over others. In the last recession, customers prioritized unsecured debt obligations and chose to maintain credit-card and auto-loan payments (since they needed a vehicle to drive to work) in preference to making payments on secured debt, such as mortgages, which subsequently deepened the collapse of the housing market. 4 Under current stay-at-home restrictions, we would expect this payment hierarchy to be reversed—customers may prioritize credit-card over auto-loan payments if they cannot leave home but need to make essential purchases, especially online.

4. Accounting policies

How aging delinquencies will be recognized will have a direct impact on loss write-offs and provisions.

We assume current accounting policies will apply, with credit-card charge-offs to be written down after 180 days. In the United States, bank regulatory-accounting requirements state that creditors must charge off closed-end installment loans after 120 days of delinquency, while open-end revolving credit accounts (such as credit cards) must be charged off after 180 days. Provisioning rules may change, but losses will still likely need to be declared. 5

However, we also expect that rules for taking provisions on bad debt may be relaxed. Early guidance from the Federal Reserve Board, FDIC, and Office of the Comptroller of the Currency suggests banks can ignore higher-capital requirements under current expected credit losses. That concession may extend to provisioning for losses in consumer debt. With the prospect of rapidly rising delinquencies, the net result may still be an increase in total provisions, especially if we see the wholesale aging of accounts that are already in late delinquency.

The four factors could yield outcomes of varying severity

The outcomes of those factors will directly determine the resulting economic landscape for consumers and lenders. Several scenarios exist for the future of the retail-lending landscape, driven by a combination of collection policies, lender capabilities to engage their customers, and the willingness and ability of customers to pay. Accounting policies will be important, but only in determining the way in which losses are recognized (and potentially steering lender capacity to provide customer assistance to those at greatest risk).

Depending on the shape of the recession forecast, we see two quite different patterns of credit-card losses. These have lasting consequences for longer-term losses, which, in the worst case, we expect will peak at almost three times the annual losses previously recorded in 2019 (Exhibit 1).

We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

In a best-case scenario, we expect the following:

  • Regulators favor self-regulation of customer-assistance activities, encouraging lenders to be more lenient for those customers unable to make minimum-due payments (that is, not declaring them delinquent) and providing incentives for customers already in deeper delinquency to find sustainable solutions to resolve their outstanding debt.
  • Lenders respond quickly to the shift to digital channels and create a more engaging customer experience, leading to better long-term engagement and financial wellness.
  • National lockdowns can be enforced through late summer 2020, and the number of COVID-19 reinfections declines sharply, enough for economic recovery to begin by the third quarter of 2020 and GDP to return to precrisis levels by end of the fourth quarter of 2020. Even under this scenario, lenders will experience a wave of charge-offs and extended delinquencies through 2021 that will require new solutions to provide high-quality customer assistance.
  • Accounting rules are extended either to enable delayed charge-offs or to consider any moratorium to have “stopped the clock” to prevent further aging of delinquent accounts.

Conversely, in a worst-case scenario, our expectations are as follows:

  • There is an effective three-month moratorium on all outbound collection activity. A combination of regulatory pressure and risk aversion leads to a widespread moratorium on all outbound collection activity for up to three months. Any lender violating this industry code of conduct suffers lasting reputational harm.
  • Lenders attempt to switch to largely digital contact strategies, but they struggle. Lenders suspend outbound collection activities for all but the most severely distressed customers (for example, those already engaged in hardship programs). All other activities rely on switching to customer-assistance practices, including payment alerts, self-serve functionalities where possible, and self-guided financial education.
  • The economy experiences an extended U-shaped recession lasting through 2021. Without enforcement of physical distancing for an extended duration, reinfections keep the economy spluttering into 2021, yielding a recession more closely approximating an extended U-shaped recovery curve. Losses are at least as bad as those in the last recession, with a 20 percent unemployment rate and credit-card charge-off losses of up to $100 billion per year. Not only are many customers unable to pay, but signals from the federal government relieve distressed customers from feeling their responsibility to pay back debt.
  • Credit-card debt charges off after 180 days. The legal requirement to charge off revolving credit-card debt at 180 days past due remains, although the requirements to take provisions for losses may be relaxed.

For each extreme, the key drivers of outcome appear to be the ability and willingness of customers to pay, given economic disruption and financial-stimulus support from the federal government, and the ability of lenders to adapt their operating models to ensure that customers remain engaged through the coming months. Exhibit 2 illustrates likely outcomes.

We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

In scenario A, restrictions on the ability of lenders to engage with their customers leads to a rapid acceleration in losses, even if the economic shock of COVID-19 is relatively short lived and a federal stimulus package softens the impact for businesses and their employees. The degree and rate of adoption of a truly digital approach to customer engagement will play a major role in separating well- from underperforming financial institutions.

Scenario B reflects the smallest impact from the COVID-19 crisis. The economic impact on customers’ ability and willingness to pay is relatively short lived, and lenders enact programs to prevent accounts, largely, from becoming severely delinquent. The institution of regular and clear communications with customers enables lenders to maintain good habits through widespread restructuring of debts.

Conversely, scenario C reflects the greatest overall losses, with likely long-term structural changes in the lending value chain. Specifically, a shortfall in financial support for customers facing widespread, unprecedented unemployment; an extended humanitarian crisis; and a strictly enforced moratorium on outbound collection activities lead to extensive, sustained losses. The rate of account closure is large enough that it creates a significant segment of the population who find it difficult to obtain credit again in the future readily.

Finally, scenario D reflects widespread economic hardship and large losses for lenders from their preexisting stock of accounts in delinquency. Proactive engagement of customers in predelinquency avoids a majority of accounts becoming delinquent, although high unemployment rates still generate elevated delinquencies for an extended period.

The time to act is now

In each scenario, we expect material differences in lenders’ success in engaging with customers, providing meaningful customer assistance, and maintaining good habits. To maximize successful outcomes for customers and minimize economic losses, lenders must act now to implement actions and strategies in collections that will have impact over three time horizons. Furthermore, the focus of those activities will need to shift through time, focusing initially on the aging of stock delinquencies and alerting predelinquent customers to the benefits of good habits but subsequently diverting efforts to serving best the swelling population of early-delinquent customers and potentially creating in-house recovery facilities to support account holders more effectively after charge-off.

Horizon one: The immediate future

Keep core operations functioning to provide immediate customer assistance, take care of employees, and avoid reputational harm.

Many credit-card lenders have already witnessed a sharp increase in the volume of inbound inquiries from concerned customers and have acted quickly to reallocate capacity from predominantly outbound and customer-facing (such as branch) roles to maximize customer assistance. In particular, we recommend the following actions:

  • Focus on the safety of employees. Ensure safe call-center environments (for example, reconfigure floor layouts to enable safe distancing, thoroughly and regularly sanitize facilities, and separate teams across floors) and divide collectors across multiple shifts.
  • Ensure the availability of customer-assistance teams. Optimize customer-assistance capacity across product portfolios, repurposing capacity from other operational functions and scaling remote-working capabilities, to serve customers in the best manner possible.
  • Provide immediate payment-relief options. Offer customers options for payment relief, such as simple payment deferrals, payment holidays, fee and interest waivers, loan modifications, and hardship programs. In addition, offer early, regular, proactive communications to reassure customers about their debt at a time of great uncertainty and confusion.

Horizon two: The next few months

Upgrade critical capabilities for the resumption of activities under the new normal.

Many credit-card lenders possess a thorough playbook for the improvement capabilities needed in the event of an economic downturn, but the onset of the COVID-19 pandemic was far swifter than most had anticipated. Accordingly, accelerated investments need to be made now in critical infrastructure that improve customer communications, simplify transactions, and distribute sound financial education. Specifically, we recommend the following actions:

  • Improve digital communications. Launch or upgrade digital communication channels for customer assistance (for example, develop an omnichannel communication engine to drive synchronized and personalized contact via email, text, website interstitials, and in-app pop-ups). In addition, launch a dedicated self-serve capability, with digital offers and financial education, for customers in delinquency.
  • Expand the use of advanced analytics. Expand the suite of analytic models and data used for customer segmentation, contact, and treatment strategies (such as geospatial- and occupation-based indicators of distress and models of contact-channel preference).
  • Upgrade frontline capabilities. Whether or not the future of customer assistance entails distribution via a virtual contact center rather than the traditional centralized model, the needs of customers over the ensuing months will require more specialized responses from staff with greater expertise at helping them achieve the best outcomes.

As the flow of inbound enquiries from customers inevitably begins to wane, lenders will want to shift the allocation of their customer-assistance capacity and create specialized assistance teams trained to help newly delinquent customers with new solutions as the effects of economic hardship intensify.

Horizon three: The longer term

Rethink the collection operating model, offer more creative solutions, and rebalance portfolio exposure.

Inevitably, the imminent economic downturn will cause many lenders to redraw their historical credit boxes while reevaluating broader credit exposure. The need for physical distancing, the closure of many offshore call centers, and the heterogenous geographic development of the impacts of COVID-19 will likely cause many lenders to rethink their collection operating models and strategies entirely. In particular, we encourage lenders to remain vigilant about which customers (in different stages of delinquency and hardship) need the most urgent assistance and begin redesigning several core pillars of their approach to customer assistance, including the following:

  • Upgrading to risk-segmentation models. Include microeconomic indicators, customer behaviors and contact preferences, and susceptibility to economic shocks in models. Such models can be helpful not just in customer assistance but in new originations, credit-line management, targeted predelinquency communications, and recoveries.
  • Enhancing customer engagement through more creative solutions. Echoing the current economic stimulus plan that favors small business, consumers will also need assistance in finding solutions to cash-flow challenges. In particular, there is an opportunity for lenders to be innovative with new products that help customers find greater liquidity. Based on upgrades to underwriting, products could include personal installment loans designed for certain types of purchases (for example, e-commerce), more flexible credit lines, and low-interest consolidation loans. Lessons can be learned from China and Europe, where such innovation has already begun, including programs to incentivize sustained good behaviors and manage deliberate retention of the growing number of charged-off accounts actively.
  • Managing the capacity and acquisition of core capabilities. Few business-continuity plans were created with the expectation of operational disruption on the level that the COVID-19 crisis is causing. Similarly, future operations require futuristic capabilities, such as dynamic risk modeling, engagement with virtual assistants, and more extensive customer self-serve facilities.

By many measures, we are entering unprecedented economic conditions while still trying to quantify the likely impact of the greatest humanitarian crisis in a generation. First and foremost, lenders’ priorities remain their customers and their employees. Resolving the immediate challenges of how to maintain effective operations will enable lenders to provide authentic and meaningful assistance—functions that are essential for the well-being of customers—at this time.

With the prospect of credit-card losses increasing to almost three times those in 2019, lenders also face the prospect of their own economic crisis. But the economic outcome of the COVID-19 pandemic can be mitigated, not only by the actions of the government, but by proactive steps taken by lenders to transform their operations and strategies.

There is no magic bullet to increase suddenly borrowers’ ability and willingness to pay. We recommend urgent investment in core capabilities that maintain engagement with customers; create tailored, affordable payment relief plans; and give borrowers much greater control and the ability to self-serve with minimal friction. Investing in a new level of assistance and cooperation among customers and lenders will not only reduce the severity of the looming economic crisis but build deeper and longer-lasting relationships that will endure through the cycle.

About the author(s)

Matt Higginson is a partner in McKinsey’s Boston office, Frédéric Jacques is a partner in the Montréal office, Nick Malik is a senior partner in the New York office, and Rohit Singh is an associate partner in the Stamford office.

The authors wish to thank Dave Kaminsky for his contributions to this article.

Related Articles