Are Bankruptcy Filings Increasing Less?

A few days ago, I posted on the U.S. bankruptcy filing figures through October 2007 and projected that there will be 820,000 and 846,000 U.S. bankruptcy filings in the 2007 calendar year. In this post, I want to take a more extended look at the long-term trend going forward.

The data suggest that bankruptcy filings are not increasing at as great a rate as they were previously. That is not to say that bankruptcy filings are decreasing or even leveling off. Rather, the rate of increase is declining. This analysis will not necessarily hold true if conditions change, and some will argue that conditions already have changed because of the subprime lending crisis. All we can do, however, is analyze the historical data. Put differently, I have to make the same disclaimer as the rushed voice at the end of any brokerage commercial: past performance is no guarantee of future returns.

With bills pending in Congress to amend the Bankruptcy Code to help homeowners struggling with out of control mortgage debt, some have questioned whether the bankruptcy system could absorb the hundreds of thousands of new bankruptcies that might develop. The analysis here helps provide some information for that debate. Although one would predict that bankruptcies will continue to increase, we should not expect bankruptcies to continue to rise at the rate they have been for the past twenty months. Looking beyond my analysis here, the total number of annual U.S. bankruptcies is a little more than half of what it was before the 2005 law. There seems to plenty of room in the system to absorb more bankruptcies.

Why do I believe bankruptcy filings will continue to rise but not at the same high rate as in the past (again assuming no dramatically changed conditions)? Later in the post, I have a statistical forecast, but we can begin with just a simple look at the data. The first table to the right/above gives the monthly data for total U.S. bankruptcy filings since January 2006. Pushing the monthly data back before that date would give a misleading picture because the October 2005 changes in U.S. bankruptcy law caused a massive surge in filings and then a near total dearth of filings in the few months thereafter. Arguably, the January 2006 and perhaps the February 2006 data still reflect this abnormal statistical blip, but starting in January gives us at least one complete calendar year.

Looking at the total number of filings each month can be deceiving. Each month, of course, has a different number of days and, more significantly, a different number of business days during which we can expect most bankruptcy filings to occur. The monthly filing figures can fluctuate dramatically just because of the vagaries of the calendar. The second column of the table shows the number of total U.S. bankruptcy filings per business day of each month. It is the bankruptcy filings per business day that I will use in the remainder of this post.

Simply looking at the data shows that the monthly increases are leveling off. The dramatic monthly increases of the first months of 2006 are behind us. Although bankruptcy filings continue to rise, the increases each month are now modest. October 2007 represented a 3.8% increase from September 2007, which in turn represented a 6.1% increase from August 2007. These are not small monthly increases, but they are smaller than the 17.6% increase from January 2007 to February 2007 figures and the 7.4 increase from February 2007 to March 2007.

The graph to the right/above shows a statistical forecast for monthly U.S. bankruptcy filing rates based on the January 2006 - October 2007 data. The solid line to the left shows the actual number of filings per day for each month from January 2006 to October 2007. The maroon dotted line to the right is a forecast based on this trend line out to January 2009. The dotted lines show 95% confidence interval bands for the forecast period. That is, based on the statistical technique I used (structural vector autoregression) and the data I had, I would predict 95% of the actual observations to fall within those bands. The model does allow for some seasonality in the historical data, namely the tendency of January and February to be high filing months and May and June to be low filing months. Again, it is worth repeating that statistical models can only model historical data. This model is a forecast based on what we saw during the past twenty months.

Looking at the forecast, we see that it predicts a continued increase in bankruptcy filing rates, but not at the same rate as for the past twenty months. The forecast also predicts total U.S. bankruptcy filings for calendar year 2008 of 1.19 million (with a 95% confidence interval of 0.96 and 1.43 million filings).
Thus, annual U.S. bankruptcy filings may again rise above the one million mark, although with the filings this calendar year expected to come in around 840,000 it should come as no surprise.

Comments

[Doug]

Thanks for your analysis. And this is a suggestion instead of a comment. The 2005 Bankruptcy Law seems to have purposely discouraged folks from using bankruptcy. A number of legislative proposals floating in response to the housing crisis may have the opposite effect of encouraging bankruptcy filings. It would be helpful if you and your expert colleagues could take your current projection and show alternatives based on scenarios that tie to this legislation. These may not carry the 95% accuracy -- but they would still be useful for folks trying to anticipate 'what might happen'. Thanks

[Bob Lawless]

Doug, thanks for your comment. There is no question that the 2005 bankruptcy law discouraged people from filing bankruptcy. The proposed changes for home mortgages would give people more options in chapter 13, so they would give people reasons to file bankruptcy that are not currently present.

As to prediction, it's really not possible to make a statistical forecast about a piece of legislation that might happen in the future. We would have to know what effect this particular legislation has had in the past. By definition, we cannot know the past effects of a future piece of legislation. I have an article called, "The Paradox of Consumer Credit" that finds, not surprisingly, over the long term household debt levels are the primary determinant of bankruptcy filing rates rather than legal change. The article can be found at vol. 2007, p. 341 of the University of Illinois Law Review.

[Troy L Gleason]

Mr. Lawless, What percentage of the increase for the year will be in chapter 7's v chapter 13's? Do you think they will increase in the same proportion or do you think the much of the increase will be in the form of chapter 7's.

[Steven G. Klesner]

I always enjoy reading your updates about filing rates. While I wonder how useful it is to look at these past two years to predict future filings, I think your analysis is useful for showing how things might go absent some major macroeconomic factors that might have a big effect on filings.
There are several factors coming together that suggest to me a surge in filings will occur in the next year or so. If the rapid growth in filings in the 1980s and 1990s is attributable primarily to the expansion of credit card lending and low rates of saving in our society, then perhaps the filings in the past several years have been retarded by consumers' ability to repeatedly borrow against home equity and stretch out the payments at lower interest rates. If the housing market remains depressed (or values return to historical levels) for some period of time, then the bottom may drop out (or has already done so) very suddenly, leading to a surge in bankruptcy filings, not so much to avoid foreclosure but because other consumer debt problems cannot be deferred by borrowing against the home. I would not be surprised to see rates of filing that significantly exceed pre-BAPCPA levels, as it seems that consumer borrowing against home equity to avoid bankruptcy was something happening mostly from about 1995 to 2007, when the housing values were spiking. There may be a huge number of consumers who have been overburdened for some time and have avoided bankruptcy in this way, but now have no other option.