Strategic Credit Reporting

I thank the people at Credit Slips for asking me to guest blog this week.  It has been a great sounding board.  I want to close by highlighting a part of Charging Ahead about which I am not at all sure: my praise of the American credit reporting system.  As I explain there, you can say that our credit reporting system works well compared to systems in most other countries because it is private, centralized, voluntary and technologically advanced.

But as lending decisions become ever more standardized, I wonder if the incentives to game the process are starting to undermine it.  In particular, increasingly the hot-button disputes about credit reporting are not the inevitable problems about the hassle of correcting unintentional errors.  Rather, what we now see is a set of problems that relate to whether the system is sufficiently complete.  For example, Capitol One and Target want to keep the information about their credit limits private, because disclosing that information would make it easier for other lenders to poach their customers.  Credit card issuers in general have pressed the national bureaus to ignore available information about bankruptcy discharges.  As far as I can tell, most of the strategies move in one way -- keeping "positive" information out of the system to make credit scores lower.

The natural response of many consumer groups, of course, is to try to force more information into the system.  But there are serious costs to an unbounded drive to gather all information into the credit reporting system.  In general, the added information is going to make it easier for lenders to identify and serve riskier sections of the market.  There also are serious privacy concerns, and not just for people of Euro-centric sensibilities.  Here, as in insurance rate-setting, do we really want a system with complete information?  Should the interest rate on my mortgage depend on the grocery store at which I shop?  The items I buy when I go there?

On the flip side, the recent growth of fringe lenders suggests that increased access to this information presents some real value to the lower middle class.  If the credit bureaus had as much information about those people as they have about the swath of the middle class that has a dozen credit cards and two mortgages per household, the rates charge in the markets in which those people borrow well might fall significantly from the levels that are so easily challenged as "unconscionable."

I don't yet have a good answer to this.  Banking groups oppose regulatory intervention, claiming that smaller lenders will drop out of the system.  And our voluntary system obviously would be undermined if a significant group of lenders dropped out.  Still, I can't accept the idea that lenders' incentives are properly aligned with those of borrowers, so I question whether lenders should be allowed to decide what and how much to report.  Farewell!

More on Filing Rates

Following up on my post from Monday, Charles Tabb has posted a new paper on SSRN, Consumer Bankruptcy Filings: Trends and Indicators.  He uses A.O. data through the first half of 2006 and suggests, consistently with my post, that the long-run filing drop will not be as substantial as many seem to think.

He also looks at the Chapter 7/Chapter 13 mix and notes that the Chapter 13 filing rate is not as high as you would expect given the stated motivation of BAPCPA.  Bill Whitford's paper in the Illinois symposium offers some good reasons why you would expect the Chapter 13 filings to be low, and I think Charles's paper buttresses that.  Looking at the weekly Lundquist data to which I referred on Monday, I have a similar take on the Chapter 13 filings.  Although the share of filings has been quite high since BAPCPA, it has been steadily trending downward throughout 2006, getting closer and closer to the pre-BAPCPA filing share.  And the high filing share plainly is an artifact of the preternaturally low level of post-BAPCPA Chapter 7 filings, because the number of Chapter 13 filings after BAPCPA has been much below pre-BAPCPA levels.

Finally, the last part of Charles's paper provides a useful catalog that shows how bankruptcy filing rates correlate roughly with several indicators of consumer indebtedness, including such things as total borrowings and credit card delinquencies.  He clearly is on the right track there.  In Charging Ahead I present some detailed data on credit card borrowing and consumer credit, with some multivariate regressions finding that credit card borrowing is significantly related to bankruptcy filings, even when you account for broader borrowing trends.

Interesting paper worth the read!

Sam's Bank

As I come to the end of Pietra Rivoli's engaging book on the Travels of a T-Shirt in the Global Economy, I learn that the principal reason that clothing imports into this country have been restricted so successfully is that apparel manufacturers, concentrated in specific congressional districts, have much more influence over Congress than Wal-Mart.  So, on the seemingly unrelated question whether Sam Walton's children can acquire a bank, I guess it should be no surprise that the protests of the independent community bankers, the regulated constituency of federal banking regulators, can derail Wal-Mart's application.

But this is a more interesting story than it first appears.  For one thing, there is a substantial "level playing field" argument that supports the regulators' position.  If Wal-Mart can ask pointedly "Why should Target get a bank if we can't have one," CitiGroup can just as well ask why Wal-Mart should be able to own banks with no regulatory supervision while CitiGroup must endure the red tape of supervision by the Fed.  There also is a considerable public-choice narrative.  If the exception for ILCs that would permit Wal-Mart to buy a bank without federal supervision rests only on the power of Utah legislators in Washington, then who needs it?

The intriguing side of the problem, however, relates (of course) to payment systems.  {You might wonder what this has to do with consumer credit, but they shouldn't have asked me to guest blog if they didn't want me to write about payment systems.}  The market for consumer payment systems in our country is dominated by a pair of national networks, whose market shares have grown rapidly over the past 30 years.  During those thirty years, the price of the product -- which is at its core a sophisticated information processing service -- has remained stable even as Moore's Law has halved the cost of information processing time after time after time after time.

For obvious reasons, it is enormously difficult to challenge Visa and MasterCard.  We might better ask (as I do in Charging Ahead) how Visa and MasterCard ever succeeded in establishing their networks in the first instance!  If we were to look for a challenger, and if we look past the possibilities of Google and PayPal (who essentially piggyback on Visa/MC), Wal-Mart certainly would be the most formidable competitor.  Wal-Mart has a network of almost 4000 locations in the United States, with tens of millions of devoted customers.  Wal-Mart is highly skilled at designing products to meet the desires of mainstream American consumers.  A payment system designed by Wal-Mart, accepted at all of its stores, could penetrate the consumer consciousness more effectively than any product since the credit card.

And if the purpose of Sam-Pay was to lower the costs of payments -- cost-cutting being Wal-Mart's core competency -- then it presumably would shift spending from credit cards, which would slow the financial distress associated with credit card use.  To be sure, there is always the possibility that Wal-Mart could follow the lead of Target and transform itself into a consumer-credit operation with an in-house retailing arm.  Wal-Mart's efforts to open full-service banks in Mexico show that this is at least a possibility.  But, my prediction is that Wal-Mart will focus on cost, as it has with the check-cashing services and money orders that are offered in its stores in most states.

In the end, I think we can rely on the lobbying power of the independent bankers to keep Wal-Mart out of the consumer credit business.  So putting the question squarely -- should we craft a regulatory exception for banks specializing in consumer payments? -- count me in!

One Call Too Many?

If most bankruptcy is induced by external factors -- divorce, health problems, and job loss being the most commonly mentioned -- we still don't really know why people call lawyers when they do.  Is it too many calls from a collection agent?  Or perhaps a collection lawsuit is filed.  I suspect that most families use the legal system only when they are already involved in it.  This question of course can be addressed through surveys, but I am considering a project designed to shed some light on this question using quantitative data about bankruptcy filings.

Weekly bankruptcy filings over the last several years reveal several patterns.  For example, at the end of each year, Chapter 7 filings fall steeply during December but rise shortly after the first of the year.  Total filings fall sharply after the first week of the year and then increase steadily through the first quarter (until April 15).  Chapter 13 filings, by contrast, are more evenly distributed throughout the year.  Notably, both Chapter 7 and Chapter 13 filings show a monthly peak.

This led me to wonder what would cause bankruptcy filings to surge on a monthly basis.  In Texas where I live the obvious answer is foreclosures.  Because all foreclosures in Texas happen on the first Tuesday of the month, it might be possible to isolate the share of bankruptcy filings motivated by foreclosure avoidance.  Georgia has a similar statute, so I plan to collect the number of Chapter 7 and Chapter 13 filings by individuals in Texas and Georgia on each date from January 1, 2004 through December 31, 2006.  The statistical analysis might be tricky, especially if foreclosure-motivated filings are a small share of filings.  And I don't see any easy way to account for differences in state foreclosure law or practice.  Still, a discernible rise in the last few days before the foreclosure date might quantify a share of filings attributable to foreclosures.

Looking forward, what would it tell us about bankruptcy filings if we know how many were filed to protect homes?  Also, how can we quantify bankruptcy filings that might be attributable to other causes?  Ultimately, I would be interested in trying to isolate the filings caused by informal collection practices -- people trying to escape what they perceive as harassment.  The policy initiative I would like to explore is the idea that borrowers would benefit if lenders were forced to initiate formal collection procedures more quickly.  When I interviewed collection attorneys several years back, one of the things I learned is just how much information collection calls can produce.  People are willing to give out bank account numbers and places of employment that enable the formal collection actions to proceed.  If the caller can persuade the debtor to make even a single $10 payment, the collector then has access to the acccount information from that check.  It is not clear how much of this activity is efficient.  More fundamentally, as I argue elsewhere, procedures designed to push individuals into bankruptcy more rapidly might be beneficial.

Our Indebted Military

Payday borrowing by the military has been a hot topic since an August 2006 DoD report estimated that as many as 17% of military personnel use payday loans.  But the spin surrounding this report has missed some troubling points.

First, I do not find it useful to accuse payday lenders of "targeting" military families by responding to demographics.  Military personnel always have been ideal customers for check-cashers.  They have relatively low salaries, and thus are persistently cash-poor.  They are unlikely to be laid off or to have their payroll checks late or dishonored.  So, it should surprise nobody that check-cashing stores, and the payday lending stores that have grown out of them, appear near military bases.  We might just as well say that bars "target" college students because they proliferate near the campuses that house their most profitable customers.  If there is a link between financial distress and payday lending, it affects both civilian and military populations.  And, I haven't seen any evidence that military personnel are more susceptible to cognitive biases than the immigrants and other low-income civilians who routinely use these products.

Second, the report does not acknowledge a more significant disparity between civilian and military personnel -- that the credit reporting system disadvantages those who do not own homes.  Because the credit bureaus collect very little information about non-mortgage expenses, they assign lower credit scores to families that make regular rent and utility payments, but not mortgage payments.  {Some make similar claims about payday transactions, but access to this data would only identify payday loan borrowers as potential customers for other subprime and fringe products.}  Unfortunately, the unstable location of military families makes them more likely to rent and less likely to own than similarly situated civilian families.  So military families as a group will have a harder time gaining access to intermediate and long term credit products.  In general, then, they will be more likely to use short-term products like payday loans.

What can be done?  The FTC's plan to increase the reporting of rent and utility information is a start, but DoD's responses are less useful.  The military would take away the product that the military personnel are using without either addressing the conditions that make the product attractive or facilitating a more reasonably priced alternative.  While a well-designed rollover ban would limit reliance on payday loans for intermediate credit needs, the 36% rate caps will make the national chains inaccessible to military families.  If enacted, we could expect to see those families depending more heavily on subprime credit cards, pawn shops, rent to own providers, and unlicensed payday lenders, all which in the long run will be worse for those families than the prohibited payday loans.

I understand that the military (as a large employer) has a real need to control the spending and borrowing activities of its employees.  But financial distress is not limited to the military.  Congress should view the military as an illustration of a general problem, not as a special case of activity that in other settings is benign.  As Jim Hawkins and I explain in a working paper, the policy issues are complex and deserve thoughtful consideration.

On the Immovability of Bankruptcy Filings

Count me in the group that is skeptical about the role of the legal system in influencing overall bankruptcy filing rates. Sure, I can see that the raw filing rates in the United States are a lot higher than they are in other countries. But when you account for factors like level of indebtedness, use of credit cards, and general economic conditions, the apparently large differences between the United States, on the one hand, and the UK and Japan, on the other, seem to disappear. The data suggest that the overwhelming majority of bankruptcy filings are inevitable, and that the principal effect of legal changes is to accelerate or defer the time of filing.

So I have been watching with interest the trends in filing rates since BAPCPA. The conventional view, of course, is that the filings are much lower than they were before BAPCPA and that the mid-2006 plateau in filings suggests that filings have stabilized at a level less than half of the pre-BAPCPA level.

But I’m not convinced. It is true, of course, that the filings for 2006 are much lower than they were in 2004 and 2005. Currently they are about 12,000 per week, as compared to 37,000 per week a year ago and 32,000 a week two years ago. It also is true that the 2006 filings seemed to plateau from around the beginning of April to the middle of August, a fact that might suggest stabilization. But neither of those facts tells us ANYTHING about filing trends. During 2004 and 2005, filings per week declined steadily for much of the middle of the year (weeks 10-28), the same period during which filings reached an apparent plateau in 2006. Thus, if we consider annualized filing trends, the mid-year plateau in 2006 in fact might reflect a push back towards the pre-BAPCPA filing level. To illustrate, the figure below shows the 2004 filings (not affected by the passage of BAPCPA), the 2006 filings, and the difference between 2006-2004.  {Apologies for my lack of graphics expertise.}  The trend line superimposed over the difference line suggests that in the first eight months of this year the relative increase in filings has eliminated about one third of the difference between pre-BACPCPA and post-BAPCPA filing levels.

For me, the hot issue in the consumer credit literature right now is learning what motivates individuals to file at the TIMES that they file. So the passage of BAPCPA provides a natural experiment to see how the statute affects filing dates. There are two obvious filing trends connected with the passage of the statute. First, the “early filing” effect: a LOT of people filed before BAPCPA who otherwise would have filed later. That effect should depress filing rates after BAPCPA until that effect plays out. Second, the “deferral” effect: the provisions that make filing more costly, more bureaucratic, and more humiliating should defer filings until people are deeper in distress. That effect should depress filings initially but ultimately fade away as well.

What is most provocative about the data is the long period over which those effects have played out. I would not have expected pre-BAPCPA early filers to have filed a full ten months early. But if we discard that explanation, we have to think that the deferral effect operates over a similarly extended period, so that the steady upward trend in filing rates reflects the period during which  the deferred filings are slowly rising to their “normal” level. If we have not yet reached that level, BAPCPA is deferring some filings more than ten months.