(CONN) Conn’s lending practices

(This is not a short writeup where I say what I truly think.  I don’t like doing such writeups because I don’t want to get sued.)

In a nutshell, I think that it is worth taking a deep dive into Conn’s lending practices.  In my opinion, the 10-Ks and 10-Qs seem to ignore some important ‘nuances’ of Conn’s lending practices.

The SEC has been asking Conn’s some very good questions.  This is probably because somebody shorting the stock has really done their homework and has been giving pointers to the SEC staff.

In one of the CORRESP filings, Conn’s provides an example of its loans:

The following provides a summary of the impact of the origination of a single $2,000 installment loan with a 12-month promotional interest-free program contract addendum. The analysis assumes the underlying retail installment loan contract provides for an annual interest rate of 21%, a term of 24 months and an origination date of February 1. For presentation purposes, the analysis also assumes that the customer makes monthly payments when due and repays the remaining principal balance in full in month 12.

conns-lending-example

Here’s my way of looking at it.  The customer has a 24-month loan at 21%.  The customer has an option to repay the loan early without interest.

A very important feature about this loan structure is the balloon payment.  If the customer wishes to repay the loan early, he/she has to come up with a large balloon payment at the end of the 12-month promotional period.

In practice, I seriously doubt that salespeople would explain such a loan as a 24-month loan with an option to prepay early.  If Conn’s salespeople want to maximize their commissions, they would explain that this is a 12-month interest-free loan with a scheduled monthly payment of $102.77.  They will trick the customer into thinking that they need to make 12 payments of $102.77.  In practice, this seems to be the case for some customers.  Looking at reviews of the company online, some customers try to pay their loan off early but do not understand that they need to make a larger-than-normal payment for the final payment.  They get frustrated when they mistakenly think that they have paid off their loan.

The prepayment option is a bad idea

This is just my opinion, but I think that the early prepayment option is a terrible idea.  When making subprime loans, lenders need their responsible borrowers to subsidize their deadbeats.  If the responsible borrowers pay no interest, there will not be enough interest income to cover the losses on delinquent loans.

As well, customers often don’t understand the value of the free option and don’t see the value in it.  It doesn’t make a lot of sense to sell something that the customer won’t overpay for.

What might be happening

I believe that Conn’s sales staff is pushing the no-interest loans onto both creditworthy and risky customers.  The company’s SEC filings show that fewer buyers are paying in cash (or with a non-Conn’s credit card) and that more buyers are choosing Conn’s “no interest option” loans.  This suggests that many buyers who would otherwise pay cash are taking advantage of no-interest financing.

By giving loans to cash customers, Conn’s is likely increasing the average FICO score of its loan portfolio as well as reducing the delinquency rates.  However, I believe that this practice decreases the risk/reward profile of Conn’s loan portfolio.  While these loans are low-risk, the returns may not be attractive if prepayments are extremely high.  The only way for Conn’s to make money on these loans is if creditworthy borrowers accidentally miss a payment or intentionally choose not to make their balloon payment.

For risky borrowers, the “no interest” loans behave a lot like longer-term subprime loans.  However, the returns are reduced by the prepayment option.

Overall, it is possible that Conn’s is lending poorly to both creditworthy and risky customers.

EDIT (9/23/2014): I don’t think that Conn’s is pushing loans onto customers with high FICO scores.

Loan dilution

It may or may not be the case that the quality of the loan book is being obscured in a few ways:

  1. Lending to customers who would otherwise pay cash.  This can cause delinquency rates, % of the portfolio re-aged, etc. to be misleading.
  2. Bad loans being “re-aged” and restructured/modified.  After modification, these loans are no longer considered to be delinquent.  I could not figure out if such behaviour changes the origination year of such loans.  Doing so would change the interpretation of the static loss data presented in Conn’s 10-Ks and 10-Qs.
  3. Beginning in May 2014, the company changed the re-aging policy: “customers may re-age their account once every four months which compares to once every six months previously”.  This may or may not have an effect on reported 60-day delinquency rates.  Re-aging accounts more frequently may reduce the number of loans considered to be delinquent.
  4. The loan portfolio is rapidly growing in size.  This is because Conn’s has dramatically increased the amount of loans it has made to customers, is in the process of opening new stores, and is getting into product categories with higher price tags (furniture).

Hypothetically speaking, suppose that bad loans were being modified to lower delinquency rates.  Suppose that the origination date of these loans were reset.  Static loss data would appear to be good during the initial years as bad loans are pushed into future origination vintages.  Past vintages will look great when bad loans are removed.  Later vintages would appear awful unless the bad loans were diluted with good loans from creditworthy borrowers.

Investor transparency

When I first skimmed the 10-K, I mistakenly thought that the “short-term, no interest credit programs” were for loans of 12 months or less.  If I had simply took everything written in the 10-K as gospel, I would have lots of… ‘misconceptions’… about Conn’s lending practices.

Similarly, my opinion is that the company’s investor presentations foster misconceptions.

The company’s current underwriting standards

I have not performed on-the-ground due diligence.  I have not walked into a store to talk to the staff about whether the company:

  • Asks customers for a utility bill.
  • Asks customers for proof of income.
  • Calls employers to verify employment, time on the job, and income.
  • Checks that the customer’s phone number is real.
  • Asks for a landline if the customer puts a mobile number on the application.
  • Asks for references.

Riskier lending?

The company has stated that it intends on moving into areas with higher price tags (e.g. furniture).  This may increase the riskiness of their loans to consumers because the debt burden is higher when the monthly payments represent a higher portion of the consumer’s disposable income.

Capitalized expenditures

You should do your own research as to whether or not the overall capex spend makes sense.

The general idea is this:

  1. Retailers usually spend most of their capex on stores, warehouses, and distribution centers.  Some retailers also capitalize the development of in-house software.
  2. You can make an estimate as to how much was spent on #1.
  3. Something unusual may be going on if #1 does not represent the majority of the capex.

Comparing FY2014 to FY2014, the company ended the year with an additional 11 stores by opening 14 new stores.  Net capex totaled $52.083M.

As far as the cost of new stores go, the 10-K states:

Our capital expenditures for future new store projects should primarily be for our tenant improvements to the property leased (including any new distribution centers and warehouses), the cost of which is estimated to be between $1.0 million and $1.5 million per store (before tenant improvement allowances), and for our existing store remodels, estimated to range between $500,000 and $1,000,000 per store remodel, depending on store size.

The FY2013 10-K states:

Our capital expenditures for future new store projects should primarily be for our tenant improvements to the property leased (including any new distribution centers and warehouses), the cost of which is estimated to be between $1.0 million and $1.5 million per store (before landlord contributions), and for our existing store remodels, estimated to range between $400,000 and $750,000 per store remodel, depending on store size.

The September 2014 investor presentation states that new (future) stores will have a “Build‐out Cost, Net of Tenant Allowance” of $500k.

I’ll leave it to you to figure out whether or not the capex numbers make sense.

Regulatory risk

In theory, the CFPB may consider the marketing of “no interest” loans and the monthly payments on such loans to be misleading.

What I’m doing with my money

I currently have no position in CONN.  In the future, I may very well have a short position in the stock (common shares and/or put options).

*EDIT (9/25/2014): I bought a few CONN puts.

Market cap: $1B
Short interest: 30.4%
Interest to borrow shares (Interactive Brokers):  1.5%

Links

Skirting Ability to Repay Catches Up with Conn’s – This excellent blog covers companies that serve the unbanked and underbanked customers.

Greenberg: Conn’s Crunched by Credit – TheStreet.com article

VIC writeup #1 (short)

VIC writeup #2 (long) – Personally, I would never use FICO scores to measure underwriting quality.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s