Google+ Followers

Thursday, June 13, 2013

There is no Evidence that Unsecured Recoveries in Chapter 11 are Diminishing (Part 3: Pervasive Errors in Tabulation of Recoveries)

This is the third in a series of blog posts related to a commission appointed by the American Bankruptcy Institute which has been presented with arguments for reform of the Bankruptcy Code to benefit unsecured creditors at the expense of secured creditors.  In the first, I explained that the advocates of reform have relied heavily on a law review article written by a student named Andrew A. Wood under the tutelage of Professor Lynn LoPucki at UCLA Law School, which (1) presents data that purport to show a substantial decline in unsecured creditors' recoveries in recent years and (2) attributes the purported decline to a purported increased use of second lien debt in capital structures.   I then outlined the issues I have discovered with Wood's article.  In the second post, I gave a detailed refutation of Wood's claim that second lien debt had negatively impacted unsecured creditor recoveries.  Now, in this post, I will show numerous, large misstatements in Wood's article concerning the size of unsecured recoveries in the cases he studied.

Pervasive Errors in Tabulation of Recoveries. 

The Wood article and the database it relies on contain several material misstatements and omissions of recovery data.  In several cases that purportedly had low unsecured creditor recoveries, I found glaring errors in the data that substantially understated their actual recoveries. Many of the errors stem from not capturing a plan amendment that changed recoveries, or not being aware of the confirmation of a competing plan that altered them.

For example, here is what Wood says about the Six Flags case: “in the Six Flags bankruptcy, the General Unsecured [again, his capitalized terms] group was expected to recover between 31 and 42 cents on the dollar against Six Flags International [sic[1]], but only 3 cents against one subsidiary and 100 cents against another.”  (Wood, p. 436).  For those recoveries, he cites the company’s June 13, 2009 disclosure statement.  Six Flags, however, did not emerge until a year later.  In between, there were further amendments to the plan and disclosure statement, an intense valuation dispute, and, finally, a settlement in March and April 2010 under which a group of noteholders at holding company Six Flags Inc. agreed to fund payment in full, in cash, to all creditors at other estates (including an agreed amount in respect of prepetition and postpetition interest to the noteholders at the operating subsidiaries, giving those unsecured creditors a recovery of around 110 cents on the dollar). Wood does not appear to have been aware of those material developments. 

The Journal Register case is another example of missing a major development in the case that increased unsecured recoveries.  Wood reports the recovery by general unsecureds to have been 9 cents, based on the plan’s distribution of $2 million across a pool of $27 million in claims. However, as Judge Gropper’s publicly available confirmation opinion stated, the plan also provided an additional $6.6 million distribution to trade creditors, that was a “gift” from the secured creditors out of their own recovery to induce acceptance of the plan.  Wood simply ignores that larger amount, which more or less quadrupled the unsecureds’ recovery.

He makes the same mistake about Pliant, evidently unaware that the plan confirmed was not the one originally filed by the debtor, but the later one that gave Apollo control of the company.  In contrast to the “0.5 cents” recovery he shows, which comes from the debtor’s unconfirmed plan, the confirmed plan gave general unsecureds (including the 2d lien which was classified as unsecured) 17.5 cents on the dollar.

The Smurfit case is also a striking example of the huge inaccuracies in the article.  Wood lists the recoveries to “General Unsecureds” in the Smurfit case as “0-100%”.  In the text of the article, he states “When computing the average, I took the midpoint values for any of the cases that had a range of recoveries for a given class.” That would mean he included Smurfit recoveries at 50% in computing his average for the entire sample.  But,  if he had simply reported what is contained in the court-approved disclosure statement (March 28, 2010), he would not have needed to “take the midpoint”; instead, he would have seen the following in the table of recovery estimates for the US debtors :
Estimated Allowed Claims
Estimated Recovery
General Unsecured Claims (SSCE)
$2.8-3.1 billion
1D, 6C-14C   
SSCC and Non-
Debtors (United States)
$11.2 million
3C, 4E, 5C     
General Unsecured
Corrugated and
$4 million
Convenience Claims
$25-30 million
The “billion” in the first row is not a typo.  The author’s “0-100” range has as its endpoints two classes that are less than 1% of the pool of unsecured claims. Worse, the “0%” only applied to creditors of non-operating debtors!  By taking the midpoint of that range, he completely ignores the substantially higher recovery of 99% of the  unsecured claims in the case, 64-71%, and materially understates unsecureds’ recoveries.

He makes a similar error in the data for Building Materials Holding Corp.  The amended Disclosure Statement filed July 27, 2009, applying to the confirmed plan, shows recoveries for unsecureds to be 55.25% and no recovery for equity.  Wood, however, records recoveries for “Senior Unsecured” as “100 cents” and “General Unsecured” as “12.1 cents” and gives old equity 36% of the equity of the reorganized debtor.  I have no idea where he got the numbers he uses. To the extent he averaged “100” and “12.1” using equal weights, he got to almost the same number as the disclosure statement, but I cannot tell if that is what he did or just a coincidence.

The Luminent Mortgage distribution is also listed incorrectly.  The senior unsecured lenders received approximately twice as much value as he reports, owing to turn-over provisions in the company’s subordinated debt that the article completely ignores.  In that tiny liquidating 11, which involved only $13 million in assets, secured claims, which represented about 1/8th of the debt in the case, were held only by an affiliate and received a smaller distribution than unsecureds, which the plan valued at zero.
In the Hayes Lemmerz chapter 22, Wood simply failed to look hard enough for recovery data.  He footnotes that the financial data was not presented in the Disclosure Statement, which may be true, but in fact the information can be found in a Plan Presentation Exhibit (docket number 824), which shows recoveries for unsecured creditors of approximately 0-5% for several different classes (noteholder, PBGC, general, etc).  Significantly, relative to his thesis, the secured class receives no better dividend.  The value simply wasn’t there for anybody.

Those are the cases where Wood portrays recoveries as less than the 77% figure ascribed to the earlier era.  But turning to the cases with higher recoveries, I found still more material errors. 

Regarding the R.H. Donnelly case, he lists the recovery as 100% for “General Unsecured” and puts nothing in the column under “Senior Unsecured”.  But that plan identified only about $19 million of “General Unsecured”, and separately classifies over $9 billion of unsecured notes that he does not mention.  They recovered from 6 to 88 cents on the dollar depending on which debtor they had claims on.

Regarding Neenah Enterprises, there is a similar omission of the recovery to senior unsecured notes, which was stated in the disclosure statement to be 80%.

Regarding NTK, he lists as a 100% recovery for “General Unsecured” and “0.5-2” for “Senior Unsecured”.  A practitioner can tell at a glance that he has to have one of those numbers wrong.  No plan of reorganization containing such a large disparity is ever going to get confirmed.   In fact, upon inspecting the disclosure statement, one learns that the notes that got that pittance of a recovery had claims only against the holdco.  But there were also senior unsecured notes with claims against operating companies that had recoveries between 24 and 66 cents on the dollar – all of which his article totally omits.

Finally, he depicts Charter Communications as a 100% recovery for general unsecured creditors, with nothing in any column related to senior or subordinated unsecured debt.  But that well-known case had billions of claims based on unsecured notes issued at various levels in the corporate structure with recoveries ranging from 0-100% depending on their structural ranking -- all of which his table completely ignores.[1] At the debtor CCI, for instance, the final (May 7, 2009) disclosure statement reveals that the pool of “General Unsecured Claims” (class A-3) received 100% - but only had $1,019,317 in allowed claims, while a class of notes issued by that debtor (class A-4) that was owed more than $497 million received only 19.4%.[2]  At the same time, the unsecured notes in Class H-4, with over $2.5 billion in claims against a different debtor, received over 100% due to postpetition interest, so the results differ widely from the simple “100%” figure Wood reports.

[1]           Strangely, Charter is one of the few cases where the UCLA database actually gives a figure for unsecured creditor recoveries, although it is less than half of Wood ‘s 100%.  Wood does not explain where the 100% figure comes from.
[2]               The class also got 3.9% from claims against a holdco debtor, according to the plan.  The trustee for those notes subsequently stated in its January 10, 2013 petition for certiorari to the Supreme Court to overturn the confirmation order that the class’s recovery was actually 32.7%.

[1]           There was no debtor in the case named “Six Flags International”; I believe Wood was referring to the holding company, Six Flags, Inc.