Bankruptcy: Why the Surprise?

Eliminate Surprise Bankruptcies With Comprehensive Risk Assessments

Most companies shouldn’t get blindsided by a customer filing for bankruptcy. The signs of a customer’s financial distress – locations closing, orders slowing down, revenue dropping, debt piling up, lawsuits getting filed – can start many months or years before a company’s debt becomes insurmountable and inevitably leads to bills going unpaid. Yet a customer having unpaid bills is usually a good leading indicator that bankruptcy is possible – and it’s one that creditors tend to over-rely on: The majority of all bankruptcies, for both public and private companies, show a slowdown of payments prior to a bankruptcy filing.  

Because sometimes – despite showing many of the other signals of financial distress – a customer continues to make timely payments right up until the day it files for bankruptcy. Perhaps it’s because it has an automated payment system, needs to maintain its prompt pay discounts, or is intentionally trying to avoid setting off red flags with creditors. Whatever the reason, the payment patterns of the troubled customer can mask its true financial condition, creating what we call the “cloaking effect.” The end result in such cases is that creditor companies feel blindsided by the bankruptcy filing.

These “surprise” bankruptcies hardly represent the bulk of bad debt losses at most companies, but when they occur, the financial impact can be large because the lack of forewarning prevents any pre-filing risk mitigation. Such surprises can be eliminated by using all the data – including both payment and financial data – that Dun & Bradstreet provides on businesses. But for those companies that tend to over-rely on payment data to trigger account reviews or for those that rely just on financial statement data, getting hit with a surprise bankruptcy can be a painful wake-up call.

Case Study – NewPage Corporation

Remember NewPage Corporation: a bankruptcy filing from 2011 where the company paid most obligations relatively promptly, right up to the day they filed. According to Dun & Bradstreet’s data, of the 616 reported trade lines, 94% of their obligations were being retired in less than 30 days past due, resulting in a D&B PAYDEX® Score of 72. This equates to 12 days beyond terms – generally deemed “acceptable” by most credit departments. Looking at the previous eight quarters of performance (refer to the chart below), the PAYDEX Score remained fairly consistent.

However, other scores – its D&B® Failure Score (then called the Financial Stress Score) and its Supplier Evaluation Risk (SER) Rating – were both sky high leading up to the bankruptcy filing. A company caught off guard by NewPage’s bankruptcy filing was likely relying too heavily on missed payments to trigger account reviews and wasn’t incorporating multiple types of data to get a complete risk picture. By looking at both the Failure Score and the SER Rating, no company would have missed NewPage’s financial distress. But companies that rely too much on a single risk signal lack a complete picture and will many times miss other signs of risk that are present.

Period D&B® Delinquency Predictor Score PAYDEX® D&B® Failure Score Supplier Evaulation Risk Rating
Q3 2011 100 72 1 9
Q2 2011 100 71 1 9
Q1 2011 100 71 1 9
Q4 2010 100 71 3 9
Q3 2010 99 70 5 8
Q2 2010 99 70 5 8
Q1 2010 96 68 1 9
Q4 2009 100 68 1 9
Q3 2009 95 70 2 9
  Key: Score is 1 - 100, where 100 is "good" and 1 is "bad" Key: Score is 1 -100; 80 - 100 indicates a low risk of late payment; 50 - 79 indicates medium risk; 0 -49 indicates high risk Key: Score is 1 - 100, where 100 is "good" and 1 is "bad" Key: Score is 1 - 9, where 1 is "good" and 9 is "bad"

How to Avoid the Cloaking Effect

Credit teams are tasked with maximizing revenue from customers of all risk levels – that’s the heart of risk management. Just because a customer begins to show the signs of financial distress doesn’t mean you start rejecting their orders altogether. There are a number of ways to mitigate risk: You can ask for a deposit, place additional terms and conditions on the order, or sometimes just talk to the customer to better understand their situation. However, you need the right data to decide upon the best approach. It’s important to make sure your company’s credit review process is comprehensive enough to catch any potential instances of the cloaking effect. Generally speaking, there are a few areas where the cloaking effect appears to mask the inherent risk:

  1. Do I Review?
  2. What Do I Review?
  3. How Do I Review?

Question One: Do I Review?

If you’re only reviewing accounts that are past due or have exceeded their credit limits, you’re not getting the complete picture. You’re unlikely to notice potentially high-risk customers who continue to pay on time and remain under their credit limit.

It’s important to receive alerts and review data based on more than just on payment status. In the NewPage case, reviewing the D&B Failure Score would have easily pinpointed the risk. You can also identify customers within your portfolio who might be cloaking-effect candidates by creating a filter that segments good-paying customers who have other signs of high risk, such as a D&B® Delinquency Predictor Score less than 90, a PAYDEX Score more than 70, and a Failure Score of less than 35. By continuously reviewing the situation and all available resources that augment your decisioning process, the reviewed account could be captured within this filter for months before ultimately failing.

Using the archived pre-bankruptcy data elements for NewPage Corporation in the chart above, verify that your credit process would have identified this company for review. If it wouldn’t have, you might want to consider revising your policy such that it can capture this type of credit risk.

Question Two: What Do I Review?

What should you review to assess creditworthiness? If you’re applying too much emphasis on payment-related data, including internal payment history, trade experiences, and trade references, again, you’re missing a piece of the pie.

None of those payment data sources is singularly or collectively sufficient to perform a comprehensive credit analysis. In the NewPage Corporation case, all of that payment data would have been perceived as positive or resulted in positive comments/points/scores – and none would have revealed the underlying financial stress. Almost everyone would have stated they were being paid according to terms, with only slight variances. This historical view and over-emphasis on payments would lead to the same conclusion in other cases – as long as a company’s payments are on time, its line of credit would be extended and the risk of failure would be missed.

Again, the goal is to look at all available data elements. Incorporating the D&B Failure Score into your process will provide you with a more complete view of risk.

Question Three: How Do I Review?

When using automated scorecards in the credit decisioning process, there’s a tendency for credit practitioners to place too much weight on historical payment performance. Back in 2011, we researched and gathered about 100 active scorecards from credit teams that used their own weightings and elements to drive the decisioning outcomes. (Refer to the chart below.) As you can see, the sample set of scorecards weight payment performance heavily relative to the other elements, leaving their companies vulnerable to “automated cloaking.” Because of this relative over-weighting, even if the other elements signified high risk, they wouldn’t generate enough negative points to offset the points received from a positive payment performance. The result? No red flag for review.

Aggregated Scorecard From Use Case Samples (Element | Weight)

  • Payment-Based Scores (D&B® Delinquency Score & PAYDEX) | 44%
  • Internal Company Payment Data (% Past Due, % > 90, Total Past Due, etc.) | 25%
  • Financial-Based Scores and Elements (D&B® Failure Score, D&B® Rating, Financial Statement Data) | 23%
  • All Other Elements (Years in Business, Employees, Public Filings, etc.) | 8%

But simply placing higher weight on the D&B Failure Score would allow most companies to easily spot the risk of “cloaked” bankruptcies.

Additionally, it’s best to periodically reexamine your scorecards and weights. Often, using a negative attribute for lower D&B Failure Scores (i.e., 32 or lower) may provide the desired outcome. If you don’t want to change your scorecard, you can instead create an exception rule to capture these high-risk “cloaking” accounts.

(By the way, don’t reallocate weights for missing elements in scorecards. In situations where data elements are not present, allow the scorecard to remain true by representing only the points awarded for present elements.)

A Complete Risk Assessment Will Eliminate the Cloaking Effect

Don’t misunderstand: There is still significant value in customer payment patterns. Slow payments or increasingly slower payments remain a good leading indicator of concern, but it’s not enough to just look at payment history. As with NewPage Corporation, some companies pay their bills on time and within terms, regardless of their overall financial health. To avoid being surprised, the cloaking effect of prompt payment patterns must be considered in the overall equation – whether analyzing or assessing risk manually or when developing the weighted elements of an automated decisioning system.

NewPage Corporation is not an isolated case of the cloaking effect. Even since their filing, there have been additional examples of other notable surprise bankruptcies, such as travel firm Thomas Cook. Unless credit managers counter the cloaking effect by balancing their risk assessment with both payment and financial data elements, companies are likely to continue encountering surprise bankruptcies. This is why Dun & Bradstreet has Failure and Delinquency scores: used in conjunction with one another they can help maximize your risk assessment and eliminate the cloaking effect.