Bankruptcies Down 12% in 2014, Forecast Predicts the Same Decline fo...

The final count for 2014 U.S. bankruptcy filings is in, thanks to the data from the ever-helpful Epiq Systems. There was a total of 910,090 total bankruptcies in 2014, a decline of 11.8% from the previous year.

In early June, my prediction was for 2014 bankruptcies to be "just over 900,000," a solid prediction but with half the data already in the books.

My prediction for 2015 bankruptcies is somewhere around 800,000, which would be another decline of 12%. I arrived at that figure in two ways.

First, the bankruptcy filings for the last six months of the last four calendar years have been between 53% - 55% of the filings for the next calendar year. Using the last six months of 2014 as a similar benchmark for 2015 produces an estimate of between 782,000 and 812,000 filings. This method is admittedly crude but also gives a starting point for an estimate.

Second, a mathematical model that uses revolving and nonrevolving consumer debt as well as national personal income does a good job of capturing the variation in bankruptcy filing rates. The reasons have been repeated many times on this blog. As households carry less debt on their balances sheets, bankruptcies go down, but short-term credit crunches can lead to increased filings. Two posts that particularly explain the idea are here and here.

The mathematical model predicts 801,807.2 bankruptcies for 2015, which is right in the middle of our cruder estimate. I don't know who is going to get stuck with that .2 part of a bankruptcy, which illustrates that the point estimate is just that -- an estimate. The 95% confidence interval on the forecast is 559,000 - 1,081,000. Like most any forecast from existing data, the assumption is the dynamics among the data do not change. If, for example, bankruptcy suddenly became dramatically less expensive, the forecast is probably no good.

All things considered, a good estimate for 2015 bankruptcy filings is 800,000.