A key skill in improving our investment process is the ability to look a mistake squarely in the face and figure out what went wrong. We will start with the ugly, Conn’s (CONN, Financial). CONN is down 55% since we wrote about it. At Punch Card, if we get something wrong, we go back into the data and we figure out why what happened happened, and why we got it wrong.
As a refresher, CONN was a specialty retailer of appliances, furniture, mattresses and electronics with 89 locations in Texas and the Southwest. Significantly, at that time, CONN financed 77% of customer purchases through its proprietary subprime credit portfolio.
Our original CONN thesis was fairly straightforward:
- Conn’s brick-and-mortar retail business has very few competitive advantages. As we have written about previously, retail is a brutal business with no switching costs and where every innovation is easily copied.
- The subprime lending business is not much better. Like the goods CONN sells, loans are a pure commodity. Even more disturbingly, over and over again subprime lenders have shown a distressing tendency to water down underwriting standards in a quest for short term profits and market share.
- However, by combining these two unattractive businesses, we felt that CONN had created a differentiated business model and a defensible niche. We didn’t discuss it in the original post but one of the models we had in our head was the Berkshire (BRK.A, Financial)(BRK.B, Financial) business, Clayton Homes. Clayton Homes is a mobile home distributor. Clayton’s biggest competitive advantage is customer financing. Berkshire’s AAA rating and low borrowing allows them to lend to customers at very low rates. Clayton targets a similar customer segment as CONN.
- Further, at the time of post, recent difficulties had driven the stock price down. Specifically, the subprime lending arm had run into problems due to lax underwriting. Given the market’s past experience with deterioration in subprime credit, the stock reaction was severe: The price fell from $79 at the start of 2014 to $16.02 by the time of our post. We believed that the market had simply overreacted to moderately bad news.
- Thus, CONN was a good company with a temporarily low stock price.
WHAT HAPPENED SINCE OUR POST?
Since we wrote our post, there has been a steady stream of bad news. Here are some of the highlights (lowlights):
- In March 2015, CONN moved to an originate to securitize model. This model came under severe scrutiny as a chief cause the Great Recession. Companies that securitize wash their hands of credit risk and have limited “skin in the game.” As a result, they have less incentive to maintain high underwriting standards.
- Earnings continued to deteriorate. In Q3 2015, CONN reported its first quarterly GAAP loss since 2012. 2016 was no better. Net income for the year was about half of 2015. The most recently reported quarter, the period ending 4/30/16, was also in a loss.
- Management turned into a revolving door. The CEO was replace in September 2015. The board of directors was shuffled in March of 2016. The CFO, who had been on the job for less than two years, was replaced in June 2016.
- Perhaps most disturbingly, despite continued assurances to the contrary, credit performance has still not hit bottom. Net charge-offs for the quarter ending 4/30/16 was 13.8% compared to 8.9% at the time of our post.
Source: Company filings via Sentieo.com
- The company continues to play accounting games. Most egregiously the percentage of re-aged loans continues to increase. The percent of loans re-aged jumped from 13.1% to 14.8%. The farce of “extend and pretend” continues.[1] They also play another game, account combinations, which had not been disclosed previously. An account combination is the practice of combining a newly originated loan with an existing loan to the same customer rather than tracking the performance separately. The cumulative impact is similar to the impact of re-aging such that when the company reports that say 94% of loans originated in 2015 were paid off, 10-15% were actually combined into a new loan.
Source: Company filings via Sentieo.com
All in all, CONN find itself stuck in a downward spiral it is struggling to get out of.
The company reports retail revenue when a sale is completed. Credit losses are recognized over time. This mismatch works great when the company is growing rapidly. The retail segment reports rapid growth while it takes much longer to recognize the losses. Further, the company uses the re-aging game to slow down the recognition of losses even more.
CONN is now dealing with the other side of this dynamic. As the company stops providing easy credit and retail growth slows, the company must report slowing retail sales at the same time as quickly escalating loan losses held over from the easy credit period. Compounding the problem, management refuses to truly clean up its lending practices for fear of further reducing sales.
WHERE DID WE GO WRONG?
In our original post, we wrote that we were “confident that Conn’s see the error of their ways and is taking steps to tighten up the underwriting process.” We believed a few bad loans would wash out of the business and leave a healthy, growing retailer. We have been severely punished for putting this much faith in management. Most businesses reconsider their lending practices when bad loans skyrocket, but not Conn’s. Instead, CONN is loaning more than ever. The loan balance increased by 14% to $1,559 million as of April 30, 2016.
Despite repeated assurances that the pig was passing through the python, it is clear that the company has still not hit bottom.
We feel this analyst’s pain (Q1 2017 Earnings Call):
Hi. Good afternoon. So wanted to start with maybe a bigger picture question, just on credit and today’s announcement. I guess it’s mostly for Norm. Having followed Conn’s for a while, and if we look at, I think, issues in your credit segment really started to come to light maybe a few years ago and since that point, there’s been a number of announcements from Conn’s and initiatives to tighten up credit. And seemingly those initiatives have not worked as well as they were intended to do and then we get today’s announcement, where once again we’re restructuring, for a lack of a better term, the credit business.
My questions is, Norm, as you look back, still being relatively new to the Company, where did the prior initiatives over the past one, two, three years fall short? And then what’s really the difference today as we look at what you’ll be doing to fix the credit business versus what you had done previously?
LESSONS LEARNED
What can we learn from this mistake?
1. Be very skeptical of turaround stories. As Buffett said in the 1979 letter, “Both our operating and investment experience cause us to conclude that “turnarounds” seldom turn, and that the same energies and talent are much better employed in a good business purchased at a fair price than in a poor business purchased at a bargain price.” Success depends on leadership, a good plan and luck. Turnarounds based on the quality of debt seem particularly difficult for some reason. The number of subprime lenders that have gone down over the years is astounding. We should have held off on concluding the turnaround would be successful until we saw more concrete indications that the credit quality had improved.
2. Don’t trust management. We should have looked more skeptically at management. We thought that a huge drop in the stock price would be sufficient to get their attention and make the hard decisions. We were wrong. Management seems unwilling to dramatically tighten lending requirements at the risk of reducing sales. The biggest reg flag was the company’s continuing reliance on “extend and pretend” lending. We certainly noted the practice in our write-up but we did not treat it as automatic disqualification. We should have.
3. Lending businesses are difficult to analyze. Even a cursory glance at our Track record makes it clear that we have a hard time analyzing lending business. We have reached the conclusion that one must have an intimate knowledge of the culture and the approach to underwriting before you properly analyze the business.
4. Assign probabilities to a range of outcomes. In the original post we wrote: “ There are two narratives at work with regard to CONN. The first is that CONN is a slow-growth big box retailer with a sleazy subprime financing arm with loose underwriting requirements that artificially pumps up sales. This collapse in the loan portfolio was simply inevitable. The second view is that CONN is and remains a fine company and the market greatly over-reacts to any bad news related to subprime.” With the benefit of hindsight, its clear that the first narrative was correct. We should have assessed the probability of each of these outcomes occurring.
[1] Indeed, the company went so far as to loosen the qualifications for re-aging. “Excessively restrictive re-aging policies cause both severity and rate of charge-off to increase. We have modified our re-aging policies to make them somewhat less restrictive, with the most significant change to allow re-aging more frequently.”
