Conseco: Restructuring An Insurance Giant With A Subprime Lending Arm

Case Study: Conseco, Inc.
Sarah Woo, Stanford University
CONSECO, INC.
RESTRUCTURING AN INSURANCE GIANT WITH A
SUBPRIME LENDING ARM
Working Paper 2/4/08 Page 1
This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Alike 3.0 United States License.
Case Study: Conseco, Inc. Sarah Woo, Stanford University
1 Objectives of Case Study

This case study on Conseco, Inc. (“CNC”) is part of a larger empirical study investigating the
determinants of loss given default (“LGD”) and developing predictive econometric models for
LGD on publicly-traded bonds and loans. The objective is to extend prior work through an
analysis of common LGD drivers contextualized in the intricacies of bankruptcy and
restructuring cases, i.e., bridging the gap between law and finance. Specifically, we hope to
confirm findings in the literature, reveal additional determinants, examine factor interdependence
and causality issues, consider the sources of uncertainty bearing on the level of LGD; and most
significantly, generate new hypotheses to test on cross-sectional data.
2 Overview and Chronology of Events
2.1 The Company
The Conseco Group was one of the largest insurance conglomerates in the United States, with
numerous subsidiaries operating in the business of health insurance, annuity, life insurance and
other insurance products. In 2001, the group had over $5.5 billion of annual premium and asset
accumulation product collections and more than $24 billion of insurance-related investments.1
The group grew to this size primarily through acquisitions – since its inception in 1982, it has
acquired 19 insurance groups.
A pivotal acquisition occurred in the spring of 1998 when CNC, the holding company, purchased
Green Tree Financial Corp (renamed as Conseco Finance (“CFC”)), a consumer finance
company, for $6 billion in stock. CFC was a market leader in subprime manufactured housing
lending. Underlying this acquisition was the vision of building the Conseco franchise into a one-
stop financial stop through cross-selling initiatives between CFC and the insurance operations,
one which was never materialized.
2.2 Organizational Structure
CNC was the top tier holding company for both the insurance and finance businesses. The
insurance business was operated through subsidiaries owned directly and indirectly by CIHC, an
intermediate holding company controlled by CNC. The finance business was operated through
CFC, a wholly-owned subsidiary of CIHC.2
A simplified snapshot of the organizational structure is presented as follows. For more details,
1Conseco, Inc., Annual Report (10-K filing) for the year ended December 31, 2001.
2Re Conseco, Inc., et al, Second Amended Disclosure Statement for Reorganizing Debtors' Joint Plan of
Reorganization Pursuant to Chapter 11 of the United States Bankruptcy Code, pp13.

refer to the complete organizational structure provided in Appendix A.
Figure 1: Organizational Structure of Conseco Group
Source: Extract from Re Conseco, Inc., Disclosure Statement, January 31, 2003. 3
It is important to note the multi-tiered structure of the group for two main reasons. First, it
enabled the use of structural subordination of various debt obligations through the use of
guarantees and pledges of subsidiary stock – an issue affecting the priority of claims. Second, it
implies a mosaic of intercompany liabilities, the treatment of which is unclear prior to
bankruptcy. This is a source of uncertainty in LGD levels and raises the issue as to whether we
should set the prior in the Bayesian LGD model to allow for higher standard deviation where a
company is flagged as having multi-tiered structures.
2.3 Initial Signs of Distress
In 1999, cracks began to surface on several levels. CNC’s stock price had declined around 68%
since the CFC acquisition, leading the company to change an accounting method in CFC (which
investors had criticized for inflating earnings).4 This resulted in a fall of third-quarter earnings by
19%, amidst a dividend cut and warnings by the company its 2000 earnings would be below
estimates. 5 Complicating the situation was the substantial increase in indebtedness – Conseco
3
The shaded boxes represent entities which were not included in the Chapter 11 petition.
4 See Conseco 3rd-Quarter Earnings Fall 19% on Accounting Change, Bloomberg, October 27, 1999; and
Conseco refocuses on its balance sheet and gets an immediate response, New York Times, Dec 1, 1999.
5 Conseco Warns of Below-Expected Earnings in 2000, Cuts Dividend, Bloomberg, November 30, 1999.

incurred substantial debt and trust-preferred obligations primarily to fund CFC's business
following its acquisition. Between 1997 and 1999, the total liabilities of the consolidated entity
almost doubled from $21.8 billion to $43.0 billion. To maintain an investment grade rating, CNC
obtained a $500 million capital infusion from the buyout firm, Thomas H. Lee Group, in return
for convertible preferred stock in late 1999. 6
Things came to a head on March 31, 2000. CNC announced that it was reviewing the value of
CFC's interest-only securities and servicing rights, and that it expected to record a write-down of
around $350 million after taxes. The company also announced that it would explore the sale of
CFC,7 which had turned out to be a severe cash drain. Between August 1999 and February 2000,
CNC injected $1.7 billion into CFC’s business. 8 In April 2000, the company filed its annual report
for 1999 which included restated results for 1999 and it disclosed that it had over-stated its net
income for the first three quarters of 1999. Following these events, the long-time CEO, Stephen
Hilbert, and the CFO, Rollin Dick, resigned by the end of April, and Conseco began its arduous
path towards restructuring.
During the interim, executives from the Thomas H. Lee Group ran the firm and the investment
group pushed for the appointment of Gary Wendt, the former head of GE Capital, as the CEO
of the company in June 2000. Wendt began a series of measures to turn around the company,
mainly through job cuts and sales of non-core assets. For the rest of 2000, the company posted
losses in its quarterly earnings report. However, the prices of the company's shares and publicly-
traded senior notes took a temporary boost in the middle of 2001 when the company posted
operating earnings of $545 million. Nonetheless, the company's problems continued as it faced
increasing defaults on the loans originated by CFC with the slowdown in economy.9 It suffered a
net loss of $419 million in the 2001 fiscal year, though its interest expense amounted to $1,609
million.
The situation was exacerbated by legal problems. CNC was plagued by class action lawsuits filed
by shareholders and TOPRS holders, alleging that the company violated federal securities laws by
making false and misleading statements about the current state and future prospects of CFC
(particularly with respect to the performance of CFC's loan portfolio). Shareholder derivative
suits were also filed against directors and officers for breach of fiduciary duties by, amongst
others, disseminating false statements concerning CFC and engaging in corporate waste by
causing the company to guarantee loans for directors and officers to purchase the company's
stock.
In 2002, CNC suffered one blow after another. CFC exposed CNC to higher than expected
losses in the subprime lending market. As the group's deteriorating financial condition became
well-publicized, there was a fall in sales of insurance products and an increased number of policy
redemptions and lapses, worsening a business which was already hurt by the market slump and
low interest rates. In April 2002, CNC attempted to manage its debt load by undertaking an
6 Supra, n4.
7 Conseco Announces Plan To Sell Its Consumer Finance Business, Business Wire, March 31, 2000.
8 Conseco Finance in Troubled Times, CNN Money, May 26, 2000.
9 Conseco's Wendt Tries to Reassure Investors on Credit Defaults, Bloomberg, November 15, 2000.

exchange offer. Under this exchange offer, qualified holders and accredited investors of the
senior notes were offered a swap for new notes with an extended maturity in exchange for
priority over the original notes. More than half of the holders tendered their notes in this
exercise.
In CNC’s second-quarter report in 2002, it announced a net loss of $4.4 billion which reduced
shareholder equity to $533 million. The reduction in shareholder equity negatively impacted its
debt ratings and the financial strength ratings of the insurance subsidiaries, thereby precluding
them from raising additional capital to ease liquidity problems.
2.4 Events Leading to Default
In August 2002, CNC announced that it had failed to make interest and principal payments on
various senior notes, corporate obligations and trust-preferred securities. The failure to make
such payments constituted an event of default under these instruments, and this triggered cross-
default provisions in other facilities. CNC pursued private restructuring efforts, having received
temporary waivers and forbearances from its creditors. It announced that it had retained Lazard
Freres and CSFB to advise the company with respect to the restructuring, including the sale of
CFC which was its top priority.
At this point, the stock had traded down to around $1, and the NYSE halted trading in Conseco's
common stock. CFC had also defaulted under its warehouse and credit facilities, leading to a
withdrawal of its debt ratings by Moody’s. With heightened uncertainty in the market involving
its potential insolvency, it was no longer able to access the securitization market. This loss of
liquidity severely affected CFC's ability to originate, purchase and sell loans, i.e., to conduct its
core business. 10
Shortly after, Gary Wendt resigned, leading Standard & Poor’s to further downgrade CNC’s
counterparty credit rating from ‘SD’ (selective default) to ‘D’ (default). 11 As the rating agency
explained, ‘the ‘D’ rating reflects Standard & Poor’s view that Mr Wendt’s resignation is a prelude
to an ultimate bankruptcy filing.’ Worse news broke out in November 2002 when Conseco posted
a $1.8 billion third-quarter loss, mostly attributed to write-downs in the value of its investment
portfolio and goodwill.
2.5 Chapter 11 Proceedings
On December 17, 2002, Conseco Inc. (the holding company) (“CNC”) filed a voluntary petition
under Chapter 11 in the Northern District of Illinois (Eastern Division), along with four
subsidiaries, namely CIHC (the intermediate holding company), CFC, Conseco Finance Servicing
Corp., CTIHC, Inc. and Partners Health Group, Inc. Most of the other insurance subsidiaries
10 Supra, n2, pp26.

11 S&P Lowers Conseco Inc. Ratings to 'D', Business Wire, October 4, 2002.

were not party to Conseco's indebtedness and they remained out of the Chapter 11 process.
The companies attempted a pre-packaged bankruptcy. 12 On December 19, 2002, CFC entered
into an Asset Purchase Agreement with CFN, a vehicle of Fortress Investment Group, JC
Flowers and Cerberus, under which CFC would sell substantially all its assets in a section 363
sale. In January 2003, CNC filed its disclosure statement, reporting its liabilities as follows:
Figure 2: Post-Petition Disclosure of Liabilities
Liabilities ($ millions)
Bank Credit Facility 1,537.0
Intercompany Claims 9.0
Guarantees under D&O Facilities 500.0
Senior Secured 93/94 Notes 93.7
Senior Guaranteed Notes ("Exchange Notes") 1,371.0
Senior Notes ("Original Notes") 1,242.0
CIHC Unsecured Claims 140.0
CNC Unsecured Claims 60.0
Trust Preferred Securities/Subordinated Debentures 2,019.0
Total Liabilities 6,971.7
Source: Conseco, Inc.’s Disclosure Statement, January 31, 2003.
*The above table does not include the amount of post-petition interest, default interest and fees, administrative fees,
etc.
In the disclosure statement, CNC proposed a reorganization plan whereby the senior lender
claims (the bank credit facility and guarantees under the Directors & Officers (“D&O”) facilities)
would receive a pro rata share of reinstated debt, with the remainder of their claim being satisfied
with preferred stock and warrants in the restructured company. The holders of the new notes
under the April 2002 exchange offer (the “Exchange Notes”) agreed to make a “gift distribution”
to the holders of the notes which were not tendered (the “Original Notes”) and other junior
classes, including a residual distribution to the common stock. This plan was met with vehement
objection by the trust-preferred securities holders (the “TOPRS”) which insisted on a higher
recovery.
On March 18, 2003, CNC received approval of its Amended disclosure statement – see the
details on proposed recoveries in the disclosure statement and the Amended disclosure
statement in Appendix B. By this point, the common stock had been eliminated from the junior
recovery, with the residual distribution going towards one of the two classes of preferred stock
(the one held by the Thomas H. Lee Group) and the TOPRS. CNC solicited votes from eligible
creditors for approval of the plan and the voting ended on June 6, 2003. There were sufficient
votes from all classes to accept the plan, except for the TOPRS who continued to object to the
12
As the holding company which was heavily reliant on dividends from subsidiaries (especially after losing access to
liquidity facilities), CNC was under too much time pressure to pursue a private workout. Several subsidiaries were
prevented from up-streaming dividends and other distributions to CNC as a result of regulatory intervention
(regulators were preserving capital for policyholders).

plan. 13
Following this, bankruptcy court proceedings to confirm the reorganization plan commenced.
Over the next few months, legal battles were fought as the TOPRS objected, amongst others, to
the valuation of CNC. There was a full hearing conducted in relation to the valuation of the
company. However, prior to the court ruling, the company and senior creditors reached a
compromise with the TOPRS, offering them common stock, warrants and rights to 45% of the
net D&O litigation proceeds (capped at $30 million). Under this plan, all classes of preferred
stock and common stock were extinguished.
In the meantime, on March 14, 2003, the court approved the sale of CFC to CFN subsequent to
an auction process. The rest of CFC assets, carved out from the sale, were liquidated, with the
main warehouse lender, Lehman, obtaining nearly full recovery (albeit adversary proceedings).
The asset-backed certificates issued by CFC suffered a series of downgrades and defaults, given
changes in the servicing fee agreements and extremely low recoveries in relation to guarantee
claims on CFC.
On September 9, 2003, the bankruptcy court confirmed the company’s Amended Joint Plan of
Reorganization, allowing CNC to begin its emergence from bankruptcy. The company emerged
as an insurance pure-play. Its new capital structure incorporating lower leverage, based on a
valuation of $3.8 billion (using “fresh start accounting”) is presented as follows:
Figure 3: Capital Structure of the Restructured Conseco, Inc.
Capital Structure of Conseco, Inc. at Emergence

($ millions)
Bank Loan Tranche A 1,000
Bank Loan Tranche B 300
Total Debt Outstanding 1,300
Preferred Stock 860

Common Stock 1,640
Total Equity 2,500
Source: Conseco, Inc.’s 6th Amended Joint Plan for Reorganization, September 9, 2003.
3 Comparing Recoveries to Historical Estimates

In this section, we examine the recoveries (i.e., the complement of LGD) on CNC’s debt
instruments according to the final plan of reorganization, against historical mean estimates from
13Conseco Inc: TOPrS Panel Voices Plan Distribution Objection, Bankruptcy News, Issue No. 22 (Bankruptcy Creditors'
Service, June 2003)

Standard & Poor’s LossStats Database. 14 A bifurcated approach is taken as we examine both the
ultimate recoveries and prices of the traded debt subsequent to default. 15 The former refers to
the market value of payments and instruments received at the resolution of distress.
Figure 4: Historical Mean Recovery Rates (1988-2003)
Nominal Ultimate Discounted Ultimate Trading Price at

Observations
Recovery Recovery Default
Senior Bank Debt 859 87.3% 78.3% 61.1%
Senior Secured Notes 283 76.0% 64.9% 58.6%
Senior Unsecured Notes 486 59.3% 42.2% 41.5%
Senior Subordinated Notes 390 38.4% 30.4% 32.3%
Subordinated Notes 354 34.8% 29.9% -
Source: S&P’s Research (2004).16
Figure 5: Creditors' Ultimate Recoveries in CNC
Claim Description Plan Treatment Nominal Discounted

Recovery Recovery
Senior Lender Claims Senior Bank Debt New Bank Debt 100.0% 90.9%
&Preferred Stock
93/94 Notes Senior Secured Notes Cash (under Finance 100.0% 90.9%
Debtors' Plan)
Exchange Notes Claim Senior Unsecured Notes 72.0% 65.5%
Original Notes Claim Senior Subordinated Notes 42.0% 38.2%
New Common Stock
CIHC Unsecured Claims Unsecured Debt 100.0% 90.9%
CNC Unsecured Claims Unsecured Debt 22.0% 20.0%
TOPRS Subordinated Notes Add warrants+rights to 1.3% 1.2%
D&O liti proceeds
Preferred Stock Preferred Stock 0.0% 0.0%
Common Stock Common Stock 0.0% 0.0%
Source: Conseco, Inc.’s 6th Final Reorganization Plan.
We observe that recoveries under the senior bank debt, senior secured notes and senior
unsecured notes are much higher than historical mean estimates. The recovery under the senior
subordinated notes is close to the historical mean, while the TOPRS recovery is substantially less
than that for subordinated notes, being around 1%.
14 2003 Recovery Highlights, Standard & Poor’s Research Report, February 6, 2004.
15 Market prices are typically observed within 30 days after the date of default, subject to data availability.
16 The discounted recoveries are arrived at using an assumption of 10% for the discount rate. Setting the level of the
discount rate for the purposes of calculating economic recovery rates or LGD has been a topic of much debate –
see, generally, Ian Maclachlan, Choosing the Discount Factor for Estimating Economic LGD in ‘Recovery Risk: The Next
Challenge in Credit Risk Management’, Ch 16 (Risk Books, 2006).

The higher recoveries achieved by the senior bank debt are consistent with a finding from current
literature that a higher debt cushion leads to higher recoveries.17 The debt cushion refers to the
amount of junior liabilities that a given seniority has below its level. The underlying rationale is
that the size of this cushion is largely determined by the credit capacity and support of the firm.
The larger this is, the more likely there will be assets of value available for distribution to the
more senior tranches in a liquidation or reorganization.
In CNC’s case, the debt cushion for senior bank debt is 71%. According to an empirical study,
where the debt cushion for senior bank debt is at least 50%, the historical mean for discounted
ultimate recovery improves to 89%, with much lower standard deviation for this estimate.18 This
comes extremely close to the actual recovery in this case. We can also use this hypothesis to
explain the relatively higher recovery for the senior unsecured notes in this case.
The issue then turns to whether this hypothesis can be applied in reverse in relation to the
TOPRS – where the proportion of senior debt above is at a high level (71%), does this result in
substantially lower recovery for subordinated notes? An alternative approach is to consider the
depressed recovery for the TOPRS grounded in the premise that trust-preferred securities
constitute a debt-equity hybrid closer to preferred stock. Nonetheless, this argument may not be
tenable since the bankruptcy court did not recharacterize the TOPRS as equity. 19 Another
argument is that the TOPRS recovery would have been 10% higher but for the existence of the
unsecured claims. This raises the question of whether we might be able to incorporate in the
model an adjustment for the extent of general unsecured claims and to test the use of leverage
ratios in cross-sectional data as potential proxies for this phenomenon.
Figure 6: Post-Default Prices of CNC's Traded Instruments
Post-Default Trading Prices

Senior Secured 93/94 Notes $80
Senior Unsecured "Original" Notes $8-13
TOPRS $1-2
Source: Bloomberg.
Turning to post-default prices, we observe that these prices are generally lower than ultimate
recoveries – a finding which is in line with contemporary empirical studies showing that the two
measures of recoveries are weakly correlated, with post-default recoveries being lower. 20 This is,
to much extent, related to the issue of risk-adjusted discounting. Logically, there should be some
financial incentives for holding on and participating in the bankruptcy process.
17 Altman, E. et al, Corporate Financial Distress and Bankruptcy (2006), pp328.

18 Supra, n14.
19 Many courts have held that their general equitable powers under the Bankruptcy Code authorize
recharacterization. See, e.g., Bayer Corp. v. Mascotech, Inc. (In re Autostyle Plastics, Inc.), 269 F.2d 726, 750 (6th Cir. 2001).
20 Brand L. and Bahar R., Recoveries on defaulted bonds tied to seniority rankings (Special report, Standard & Poor’s, 1998).
See also Servigny, A. and Renault, O., Measuring and Managing Credit Risk (McGraw Hill, 2005) at pp123.

A rough calculation of the level of internal rate of return (IRR) produces 18% for the senior
secured notes and 300% for the senior subordinated notes. The high level for the latter suggests
market expectations of the type of ultimate recovery – holders of the senior subordinated notes
are more likely to obtain a combination of various classes of equity in the restructured entity, i.e.,
instruments bearing higher risk than cash (which the holders would get if they had sold upon
default).
Another major observation is that the senior secured notes were trading at a much higher level
than historical estimates of post-default prices, while the senior subordinated notes were trading
at more than 50% lower than historical numbers. One explanation is that market participants had
priced in the subordination of the Original Notes to the Exchange Notes (as reflected in Fitch's
downgrade of the Original Notes from 'B-' to 'CCC+' following the Exchange Offer), but not
the possibility of deviations from absolute priority at that point. Also, it is possible that the
signal-to-noise ratio is low, given the supply-demand situation for defaulted bonds in markets
during the 2002-3 downturn, and we should test for market supply-demand effects on cross-
sectional data.21
4 Effective Leverage and Relative Seniority

To properly model LGD, one needs to accurately factor in the extent of liabilities in the capital
structure (to determine drivers such as debt cushion) and thereby the effective leverage of the
company. However, the effective leverage at default is often under-stated by the book value of
liabilities – an input commonly captured by current LGD models.
CNC provides an instructive example as to certain types of adjustments which we should

consider in the model development process. First, we highlight the extent of contingent liabilities
present in Conseco's case which might have contributed to the divergence of the recoveries of
junior creditors from historical estimates. 22 The company guaranteed bank loans (the “D&O
credit facilities), and made related interest loans, to certain directors, officers and key employees
for the purchase of the company's stocks. As the company’s stock price sagged, the company
made provisions for these liabilities in its financial statements (amounting to $110.2 million in
2001, $150 million in 2000 and $11.9 in 1999). However, these amounts fell far short of the
aggregate amount which, as of the petition date, was around $580 million. Given that these were
senior bank debt ranking above other claims, the failure to properly adjust for the full extent of
these liabilities may lead to misleading predictions.
21 It should be noted that the Original Notes were traded very infrequently after the Chapter 11 petition. The year of
2002 was a year with unprecedented default volume, posting the highest total dollar volume of defaulted debt (in
nominal and real terms) since 1930 - see Hamilton, D. et al, Default & Recovery Rates of Corporate Bond Issuers,
Moody’s Investors Service Report (January 2004).
22 CNC’s contingent liabilities extended beyond its guarantee obligations under the D&O credit facilities. Part of the
CNC unsecured claims derive from securities fraud class action settlements.

Figure 7: CNC's Capital Structure as of the last Annual Report prior to default
Liabilities ($ millions)
Notes payable and commercial paper 4,087.6
Notes payable to subsidiaries (eliminated in consolidation) 353.5
Payable to subsidiaries (eliminated in consolidation) 41.9
Other Liabilities 589.5
Total Liabilities 5,072.5
Trust-Preferred Securities 1,914.5
Source: Conseco, Inc.,’s Annual Report (10-K Filing) as of December 31, 2001. 23
Second, we note the difficulties of estimating the level of liabilities for an entity in a
conglomerate reporting consolidated statements. As the figure below shows, inter-company
obligations constitute a substantial portion of CNC’s liabilities. During Chapter 11 proceedings,
certain obligations were cancelled and the rest were reinstated, with the latter being part of the
unsecured claims. As discussed above in the section on Organization Structure, we may test
whether it is useful to include a variable flagging multi-tiered structures and complex inter-
company obligations.
This leads us to the next issue – why are the CIHC unsecured claims ranking ahead at a relative
seniority of 1 (see the following figure summarizing the relative seniority of claims on CNC)?
This relates to the use of structural subordination to manage a company’s capital structure, while
not being in violation of debt:capital ratios found in covenants of bank credit facilities. In this
case, the creditors of CIHC, the intermediate holding company, are necessarily senior to those of
CNC since CNC holds CIHC stock as the primary asset, and therefore has value to the extent
that CIHC creditors are paid in full and equity value flows upstream.
In current LGD models, the relative seniority variable usually reflects only contractual
subordination. Structural subordination is occasionally considered through the collateral quality
using dummy variables to flag the presence of subsidiary guarantees and pledges of capital stock
of subsidiaries. This method may not adequately reflect the relative seniority of unsecured
obligations – in this case, the ultimate recovery stood at 100%. It would be worth testing whether
the model is more predictive, where the relative seniority variable takes into account priorities
among instruments through their placement in multi-tiered corporate structures.
Figure 8: Relative Seniority of Claims on CNC
23In CNC’s case, we need to do a further adjustment to add back the trust-preferred securities – which was akin to
subordinated notes – to Total Liabilities, since the company reported it below the line, but above the Stockholder’s
Equity section.

Claim Description Estimated Seniority/Collateral Relative

Amt ($ M) Seniority
Intercompany Claims Notes incurred with various subsidiaries 9.0 Reinstated. 1
Bank Credit Facility Credit facility with Bank of America and other 1537.0 Subordination provisions in "Senior Notes" 1
banks indentures; CIHC Guarantees
Guarantees of D&O Guarantees of credit facilities provided by Bank of 500.0 Subordination provisions in "Senior Notes" 1
Facilities America, etc, to directors and officers to purchase indentures); CIHC Guarantees; 1999 D&O facility
CNC's common stock further secured by pledges of stock of CIHC, CFC,
CCM and certain intercompany notes
93/94 Notes Senior notes issued in 1993 and 1994 93.7 Secured by stock of CIHC, CFC, CCM and other 1
subsidiaries and intercompany notes
Senior Notes Senior notes from 1998-2001 re-issued pursuant 1371.0 Subordination provisions in Trust Preferred 2
("Exchange Notes") to an Exchange Offer extending maturity dates in Securities indentures; CIHC Guarantee
return for CIHC guarantees
Senior Notes ("Original Senior notes issued between 1998-2001. 1242.0 Identical to the Exchange Notes except for the lack 3
Notes") of CIHC Guarantee
CIHC Unsecured Claims Unsecured claims incurred by CIHC, the 60.0 Unsecured; structurally senior to CNC obligations 1
intermediate holding company which also
guarantees certain debts of CFC
CNC Unsecured Claims General unsecured claims incurred by CNC 140.0 Unsecured. 4
Trust Preferred Subordinated securities issued by subsidiary 2019.0 Subordination provisions; unsecured. 5
Securities trusts, guaranteed by CNC
Preferred Stock 90,000 Series E Preferred Stock; - 6
2,855,502 Series F Convertible Preferred Stock
held by Thomas H. Lee Group in exchange for
1999 capital infusion
Common Stock 346 million shares (delisted from NYSE effective - 7
Sep 2002)
Source: Conseco’s 2nd Amended Disclosure Statement, March 18, 2003.
5 Deviations from the Absolute Priority Rule (“APR”)

A general doctrine of bankruptcy law is that of absolute priority, which stipulates that creditors
should be compensated for their claims in a certain hierarchical order and the most senior
creditors are entitled to being paid in full before a less senior claim receives anything. However,
deviations from this doctrine are common and well-documented.24 The question then turns to
how we should model the presence and extent of such deviations.
Absolute priority in CNC’s case would mean distributions (after administrative and priority
claims) to the senior bank debt, senior secured notes, CIHC unsecured claims and senior
unsecured Exchange Notes, making them whole. Based on a valuation of $3.8 billion, this would
24 See, for example, LoPucki, L. and Whitford, W., (1990). Bargaining over equity’s share in the bankruptcy reorganization of
large, publicly held companies, Penn. Law Rev. 139 and Weiss, L., Bankruptcy Costs and Violation of Claims Priority in
Bankruptcy & Distressed Restructurings edited by Altman, E. (Beard Books, 1999).

have left less than $200 million for the remaining creditors. This means around 15% recovery for
the Original Notes and the interests of more junior creditors would be extinguished.
Nonetheless, we observe two major deviations from absolute priority in this case – an
observation in line with findings in the literature that firm size (CNC being the third largest
bankruptcy in the United States after Enron and Worldcom) is important in predicting whether
priority of claims will be violated. The larger and more complicated the bankruptcies present
more opportunities for junior creditors to extract concessions from more senior creditors. 25
We can take another angle, viewing these deviations as occurring primarily due to irregularities
leading to litigation potential and uncertainty associated with the firm’s valuation. Since litigating
these potential claims would be time-consuming and costly, the parties compromised with a plan
deviating from absolute priority. This supports Baird & Bernstein’s general theory that the
uncertainty inherent in restructuring a large corporation creates bargaining dynamics that explains
the deviations.26 The valuation uncertainty issue will be analyzed in the Section 6, while we
discuss how the issue of irregularities bears on two major deviations which played out in CNC’s
case.
5.1 Irregularities in Prior Restructuring Mechanisms
The first major deviation occurred where the holders of the Exchange Notes agreed to make a
gift distribution to junior creditors, including common stockholders (as proposed in the
Disclosure Statement). The Exchange Notes were issued under a distressed exchange offer in
April 2002. As an inducement to holders of senior notes (Original Notes) to tender their notes in
exchange for new notes with extended maturities, the latter were issued with a guarantee by
CIHC. This effected a structural subordination of the Original Notes to the Exchange Notes.
When CNC filed for Chapter 11, parties recognized the potential for litigation in relation to the
exchange offer.27 The offer exchange offer was not registered with the Securities Exchange
Commission and was only made to certain classes of investors (“qualified institutional investors”
and “accredited institutional investors” in the United States.28 This could have constituted a
violation of indenture provisions prohibiting the company from impairing the noteholders’ right
to receive payments or affecting their ability to enforce such rights.
In short, prior debt restructuring measures could have been flawed with irregularities, leading the
door open for potential litigation over the priority of Exchange Notes. Against this backdrop, the
Exchange Note holders agreed to re-allocate around $380 million from their distribution under
absolute priority to enhance the recovery of the Original Notes and provide a portion of
25 Supra, Weiss, n24.

26 Baird, D. and Bernstein, D., Absolute Priority, Valuation Uncertainty, and the Reorganization Bargain (2005), John M. Olin
Law & Economics Working Paper No. 259.
27 Re Conseco, Inc., Reorganizing Debtor's Memorandum of Law (June 13, 2003).
28 Conseco, Inc., Press Release in Current Report (Form 8-K), April 12, 2002).

recovery to other junior creditors.
Following this, the question is how we should reflect such irregularities in the modeling process.
One method would be to create a dummy variable for prior restructuring measures, recognizing
the litigiousness of bankruptcy proceedings whereby junior creditors are likely to challenge
restructuring measures undertaken not long from the time of entry into Chapter 11. Another way
to reflect this might be to test a factor for “time in distress”, i.e., based on a hypothesis that more
irregularities are likely to be found where a company had been in desperate straits for a longer
time. CNC would provide good support for a “time in distress” variable – its agency ratings were
first cut to junk status in March 2000 (1.5 years prior to its Chapter 11 petition).
5.2 Irregularities owing to Corporate Governance Issues
The next major deviation emanates from the tussle of the TOPRS with CNC and their senior
creditors. Constituting one of the largest claims in dollar amount, the TOPRS launched a
multitude of objections against the proposed reorganization plan which initially left them with
less than 1% recovery. They raised an attack on the company’s valuation of the company – a
classic move in a junior creditors' holdout which will be discussed in the Section 6. What is more
interesting is their other major attack on the D&O credit facilities.
As discussed above, CNC’s guarantee obligations under D&O credit facilities came about as a
result of its executive compensation plan. In the proposed reorganization plan, CNC proposed to
pay the amounts owed under these facilities to the lending banks and release these officers and
directors from their repayment obligations. 29 This was a major source of contention in the Plan.
The U.S. Trustee and the Securities & Exchange Commission objected to the plan as well,
challenging the right of a Chapter 11 debtor to use a plan of reorganization as a vehicle to obtain
a release for officers and directors.30 CNC later relented on this point and agreed to provide the
TOPRS with 45% of the net D&O litigation proceeds (with a cap of $30 million).
However, the issue did not end at this juncture. The TOPRS also claimed that the guarantees
under D&O facilities should be void and unenforceable because, amongst others, they were in
violation of Regulation U and section 402 of the Sarbanes-Oxley Act.31 The main basis for an
assertion that the facilities violated Regulation U lay with the maximum loan value requirements
of Regulation U, though the chances of successfully pursuing this argument might be low given
that the facilities were structured to comply with Regulation U. The Sarbanes-Oxley argument
was even weaker, given that the Act was effective as of July 2002 and the facilities at stake relate
29 Re Conseco, Inc., et al, Disclosure Statement for Reorganizing Debtors' Joint Plan of Reorganization Pursuant to
Chapter 11 of the United States Bankruptcy Code, January 31, 2003.
30 These objections are generally based on Section 524(e) of the Bankruptcy Code which provides, in relevant part,
that “discharge of a debt of the debtor does not affect the liability of any other entity on, or property of any other
entity, for such debt”. There is also case law where the bankruptcy court was found to lack jurisdiction over the
release of non-debtors – see, for example, Union Carbide 686 F.2d 595 (7th Cir., 1995).
31 Supra, n27.

to loans made in the 1990s and early 2000-1. 32
The TOPRS staged a holdout, blocking the proposed reorganization plan in June 2003 and put
up a strong fight throughout subsequent proceedings. Eventually, the senior lenders under the
D&O facilities compromised and gave up the warrants (part of their initially proposed
distribution) to the TOPRS. 33 Since this re-allocation of distribution from the senior lenders to
the TOPRS was mainly a compromise owing to the litigation potential over the D&O facilities,
the D&O lenders’ pro rata share of new preferred stock (with a liquidation preference of $7
million) was re-allocated to the non-D&O lenders classified under the same class of senior
lenders. 34
While this specific issue is very unlikely to recur now that the Sarbanes-Oxley Act is in force, it
highlights how poor corporate governance may constitute a source of irregularities and potential
litigation in bankruptcy. As such, we would test using cross-sectional data whether model
accuracy improves with the incorporation of factors reflecting corporate governance behavior. 35
It should be noted, however, that empirical studies show that the correlation between corporate
governance practices of U.S. firms and firm value has been considered to be relatively weak. 36
This may influence the type of testing to be undertaken. To illustrate, if the model does not show
significant improvement with the use of corporate governance indices, it may be worth trying
variables reflecting more extreme behavior, e.g., the presence of multiple class action and/or
shareholder derivative suits against the company. Such variable would have flagged poor
corporate governance in CNC as well. Prior to default, the company was plagued with securities
fraud class action suits and shareholder derivative suits alleging corporate waste through its
guarantees of D&O credit facilities and dissemination of false statements regarding CFC’s
performance.
6 Asset Valuation
Valuation is at the heart of the bankruptcy process. The determination of the going concern
value of the company is central to the negotiation and confirmation of the reorganization plan.
32 Id.
33 Note that the strike price of the warrants was set at a higher level than the initial level (where the company
proposed to distribute the warrants to the senior lenders.
34 Re Conseco, Inc., et al, Reorganizing Debtors’ 6th Amended Joint Plan of Reorganization pursuant to Chapter 11
of the United States Bankruptcy Code, September 9, 2003.

35
Besides the D&O facilities, CNC was notorious for its extravagant executive compensation packages. The total
cash compensation for the CEO and the 5 highest-paid executives amounted to $10.4 million and $6.2 million in
2000 and 2001 respectively, even though the company was reporting negative net income in both years. In terms of
board independence, CNC had 4 directors out of 10 who were insiders.
36 Bhagat and B. Black , The uncertain relationship between board composition and firm performance. Bus. Lawyer 55
(1999), pp. 921–963.

In this section, we will examine the major drivers of valuation, or put another way, the main
sources of uncertainty surrounding valuation, drawing themes from CNC’s valuation drama and
the CFC subplot.
CNC, supported by senior creditors, pegged the value of CNC at $3.8 billion; while the TOPRS
committee37 argued for a value of over $5 billion. At $3.8 billion, the TOPRS was a constituency
which was “out of the money”. While the court dismissed a motion by the TOPRS to file an
alternative plan, it allowed the TOPRS to hire a financial adviser to provide an alternative
valuation, 38 and held a valuation hearing. The valuation gap was significant – at $5 billion, the
TOPRS could have achieved at least 25% recovery, even if the absolute priority rule applied (see
the figure below).
On the other hand, CFC was acquired by CFN in a section 363 sale for $1.3 billion. This was a
fairly low number, considering that a conservative valuation of CFC earlier was $2.2 billion to
$3.0 billion. 39 As a result, CNC did not receive any net proceeds from the sale beyond the
extinguishment of CFC's liabilities and the related guarantees of CNC and CIHC.
Figure 9: Impact of Valuation on Recoveries in CNC's Restructuring
Source: Various CNC bankruptcy filings.
The sources of valuation uncertainty are broadly classified into the following categories: 40
37 Re Conseco, Inc., Official Committee of Trust Originated Preferred Debt Holders’ Post-Trial Memorandum of
Law in Opposition to Confirmation of the Holding Company Debtors’ 4th Amended Plan of Reorganization, July
24, 2003.
38 It should be noted that, out of the three members of this committee, two were hedge funds – United Capital
Markets and Oppenheimer Capital.

39 See, for example, Moody's Investors Service research report on Conseco, Inc. (July 2000).
40 This analytical framework is borrowed from a study done by Baird and Bernstein. See, supra, n26 at p30.

i. External Environment: Much is unknown about the prospects for the economy and the
particular industry sector in which the company operates. The uncertainty surrounding
future market performance and expectations compound the difficulties in determining
the cost of capital – the rate at which the market will discount the company’s future cash
flows. This issue is especially interesting in this case since the Conseco Group filed for
Chapter 11 during the downturn and emerged at the cusp of a market rebound.
ii. Firm-Specific Factors: Much is unknown about the company itself – the strength of its
new management team, growth prospects, the quality of its financial information, the
realization of tax loss carryforwards, etc. In particular, we will discuss how the accounting
quality might have elevated the uncertainty in its asset value.
The corollary of valuation uncertainty is “appraisal variance” where parties to the reorganization
negotiations come with different views of the enterprise value. 41 The scene then changes from
one of analytical complexity to organizational complexity where negotiation dynamics, and
occasionally the bankruptcy process itself (e.g., the section 363 sale), drive the valuation. We will
investigate whether it is possible to find proxy factors to reflect the impact of parties’ negotiation
strategies and their relative bargaining power on LGD.
6.1 External Environment: The Downturn
The Conseco Group filed for Chapter 11 during a time of severe economic slowdown and in a
year with an unprecedented peak in default volume. 42 During bankruptcy proceedings in 2003,
the Fed opined that there was still “considerable uncertainty attends the near-term outlook for
the U.S. economy”, with a negative picture coming from production and employment statistics. 43
Indeed unemployment rates peaked in the third quarter of 2003 while CNC was holding its
valuation trial. However, the company emerged at the cusp of the market rebound, with the S&P
500 showing 15% growth between September 2003 and December 2002.
The question is: how did the downturn affect LGD levels? This is a particularly interesting
question given an ongoing debate on the Basel II requirement of downturn LGD in the
regulatory capital formula. There is empirical research showing that LGD generally rise during
the stress period,44 juxtaposed against research that the systematic variable had no effect on LGD
after bond market conditions (supply-demand imbalances) were accounted for.45 Another school
41 Id.
42 Supra, Hamilton, n21.
43 Ferguson, R., Uncertain Times: Economic Challenges Facing the United States and Japan, FRB Speech (June 13,
2003)
44 See, for example, Carey, Mark and Michael Gordy, 2003, Systematic risk in recoveries on defaulted debt, Federal Reserve
Board, Working Paper; and Araten, Michel, Jacobs, Jr., Michael and Peeyush Varshney, Measuring Loss Given Default on
Commercial Loans for the JP Morgan Chase Wholesale Bank: An 18 Year Internal Study, RMA Journal, May 2004.
45 Altman, E., The PD/LGD Link: Empirical Evidence from the Bond Market, Recovery Risk edited by Altman, E. (Risk
Books, 2006).

of thought found that LGD is more heavily influenced by industry-specific factors, rather than
the macro-economic state.46
The filings in the valuation trial clearly show that parties were arguing their view of enterprise
value in the shadow of economic and market indices. The valuation was undertaken using a
comparable company analysis and an actuarial discounted cash flow approach. The first approach
provided much fodder for the TOPRS argued for a higher value over time in light of the rising
stock market in 2003 (see figure below). 47 In the TOPRS’ own words, “[the] change in the price
earnings ratio materially increases [the] valuation of the company because under the comparable
company method, the stock market acts as a benchmark or a measure against which to compare
the company that is being valued”. This may support the empirical finding that the average S&P
500 equity return has a stronger correlation with LGD, compared to other macro-economic
indicators such as GDP growth or the Moody’s All-Corporate Default Rate.48
In relation to the actuarial approach, parties fought vehemently over the appropriate discount
rate.49 The generally accepted method to arrive at the discount rate is the use of WACC analysis,
which depends on the risk-free rate and a risk premium. 50 This suggests the option of testing on
cross-sectional data bond spreads to reflect the impact of risk premium on LGD (a factor
generally overlooked in current LGD modeling). As Figure 11 shows, the composite spread index
for speculative grade issues remained at fairly high levels (albeit off the 2002 peak) in the first
half of 2003. This might have bolstered the arguments by the company and senior creditors
based a high discount rate of 12%, instead of the typical hurdle rate of 10-11% for insurance
companies. 51
46 Acharya, Viral V., Bharath, Sreedhar T. and Anand Srinivasan, Understanding the recovery rates on defaulted securities,
London Business School, Working Paper, 2003.
47 Supra, n37.
48 Jacobs, Michael, Understanding and Predicting Ultimate Loss-Given-Default on Bonds and Loans, Working Draft, September
2007. Available at SSRN: http://ssrn.com/abstract=1015331

49 TOPRS memo – insisted that the discount rate of 12% was artificially high and that the distress premium of 5%
was unwarranted given positive developments in financial markets and CNC’s improved position.
50 See generally Chapter 12 in Scarberry, Klee, Newton & Nickles, Business Reorganization in Bankruptcy: Cases and
Materials (2006).
51 According to an August 2002 White Paper for the American Council of Life Insurers, the typical hurdle rate for
insurance companies in the US is around 10-11%.

Figure 10: S&P 500 Indices (1997-2006)
Source: Bloomberg.
Figure 11: Bond Spread Indices (1998-2004)
Source: Bloomberg.
On the other hand, the specific case of CFC may provide support for the school of thought

focusing on industry-specific factors and suggest that we should develop Bayesian priors relating
to industries which tend to be more vulnerable to downturns. The downturn had an
asymmetrically severe effect on CFC’s subprime lending business. Within the consumer finance
sector, manufactured housing was the worst performer.52 In January 2002, one of the largest
issuer of manufactured housing ABS and CFC's main rival, GreenPoint Credit exited the
business, stating that the decision was driven by “the most severe downturn in the manufactured-
housing business that has ever occurred”.53
The deterioration in the sector hit CFC on all fronts. First, the rise in delinquencies and loss
severity led to loan losses and its costs of servicing increased with the elevated foreclosures.
Second, CFC suffered a material cash burden owing to its provision of credit enhancement to
ABS holders (it provided guarantees to lower-rated tranches and its servicing fees were
subordinated to payments to the tranches). Third, its substantially-weakened financial position of
CFC precluded its access to securitization markets – a source which had been the company's
lifeblood and the lack of which put to doubt the viability of its business model. In this light, it
was not difficult to explain the low bids for CFC in its section 363 sale.
6.2 Firm-Specific Factors: Accounting Quality
In terms of firm-specific factors, we observe an interesting source of valuation uncertainty in

this case – its aggressive accounting treatment of financial statement data. CNC had been
historically aggressive in its accounting practices, applying the purchased method for its
acquisitions. 54 The company would record the discounted value of the profits expected in the
future from the acquired company’s operations, instead of goodwill which had to be amortized
over time.
Subsequently, to avoid charges to earnings where the actual present value of premiums were less
than the present carry value of that asset, CNC shifted its bond portfolio from ‘held-to-maturity’
status to ‘actively managed’ to carry investments at the high prevailing market prices of the 1990s
(a unique practice in the insurance industry).55 This accounting technique eventually backfired
when the economy slowed down. The company reported unrealized losses on its fixed income
portfolio of $781 million for the year of 2001.
The accounting quality of financial statements filed by CFC was worse, ranging from aggressive
treatment to fraud:
i. CFC historically used the gain on sale accounting method for the securitization of its
financial receivables. 56 In 1997 and 1998, prior to Conseco's acquisition, securitization
52 U.S. Fixed Income 2003 Mid-Year Outlook/Review, Nomura Research Report, June 27, 2003.
53 GreenPoint Credit Press Release (January 2002)
54 Abe Briloff Ponders Accounting's Current State, Barron’s, October 11, 1993.
55 Id.
56
In the latest rules representing U.S. implementation of the Basel II Capital Accord, a financial institution has to

gains alone amounted to 60% and 52% of total revenues respectively. Gain on sale
accounting relies on subjective and easily-manipulated assumptions, allowing management
to front-load earnings. Note that CFC's accounting strategy, which focused on its own
assumptions of discount rates, default rates and loss severity, predates the recent FAS 157
which states that “fair value is a market-based measurement, not an entity-specific
measurement”. 57
ii. The risk of over-stating asset values using gain on sale accounting materialized in late
1999 – CFC recognized an impairment charge of $554 million to reduce the book value
of retained interests (mainly interest-only securities and servicing rights). Subsequently in
2000 and 2001, CFC recognized further impairment charges of $516 million and $387
million respectively. After September 1999, CFC switched to the portfolio method
whereby the securitized debt remained on the balance sheet as secured borrowings.
Theoretically, this method should be more conservative in terms of revenue recognition.
However, it introduced another subjective dimension – CFC had more leeway to defer
impairments as long as they can argue that the losses were temporary in nature.
iii. CFC adopted an aggressive stance in its treatment of loss provisions and guarantees.
Between 1997 and 1999, CFC's loss provisions relative to total originations was less than
1%. This ratio stood at 0.16% in 1997. Assuming a typical loss severity of 40%, the
default rate would have been projected at less than 1% - an extremely low number for
subprime exposures! The loss provision eventually increased by 191% in 1999 and 175%
in 2000. Furthermore, in its 1999 annual report, CFC stated in relation to its $1.6 billion
guarantees of finance receivables that the “the likelihood of a significant loss from such
guarantee is remote.” Subsequently in 2003, CFC projected that payments under the
guarantees would exceed the gross cash flows from the retained interests for the next five
years.
iv. The SEC eventually instituted cease-and-desist proceedings against the company for
making materially false and misleading statements about their earnings, overstating their
results by hundreds of millions of dollars. 58 The SEC found that, in 1999, CNC’s CFO
and Treasurer instructed the accounting department to change the historical basis of each
IO security. This allowed the company to avoid taking charges to earnings albeit a
weakening subprime market, and coupled with other improper adjustments, they falsely
increased earnings. 59
deduct gain on sale associated with securitization exposures from Tier 1 capital, to offset the aggressive accounting
treatment that provides an increase in Tier 1 capital at the inception of securitization.
57 Under FAS 157, “[the] use of an entity’s own assumptions about future cash flows is compatible with an estimate
of fair value, as long as there are no contrary data indicating the marketplace participants would use different
assumptions. If such data exist, the entity must adjust its assumptions to incorporate that market information…”
This, however, does little to prevent “mark-to-make-believe” where the industry itself uses over-optimistic
assumptions.
58 Re Conseco, Inc., Order Instituting Cease-and-Desist Proceedings, Making Findings, and Imposing A Cease-And-
Desist Order pursuant to Section 21c of the Securities Exchange Act Of 1934 (see order available at
http://www.sec.gov/litigation/admin/34-49392.htm ).
59 Id.

The valuation trial took place in the context of poor accounting quality of the Conseco Group’s
financial data. To illustrate, the TOPRS argued that an earlier Ernst & Young report valued CNC
at $5.9 billion as of December 2001, asserting that “[the] Debtors failed to offer any quantitative
analysis attempting to justify a $2 billion drop in value in six months (1/01/02 to 6/30/02)”. 60
The retort by the debtor was premised on the uncertainty of further charges to earnings, e.g.,
citing a $900 million writedown in 2002. As a Fitch commentary put it simply, it was unclear
whether policies adopted by the company would “ultimately generate lower loss results, or simply
delay actual loss recognition”.61 As such, this case may support a hypothesis that a factor
reflecting accounting quality should be useful in estimating LGD levels.
6.3 Negotiation Dynamics and Stakeholder Incentives
It is widely recognized that negotiation dynamics play a significant role in determining the
valuation of a reorganized entity.62 Typically, senior creditors argue for a lower valuation, while
junior creditors and equity holders are better served by a relatively high valuation which allows
some distributions to be allocated to these lower classes. This classic position played out in
CNC’s case.
Gilson et al suggest that we can develop empirical proxies to represent the likelihood that
incentives of the senior and junior interests might influence plan value. 63 One possible proxy
factor was the relative size of the creditor’s claim and the presence of vultures gaining a
controlling stake in the debt class. 64 In CNC’s restructuring, the senior lenders and the TOPRS
were finely balanced 1:1 in terms of the relative size of their claims. In such a case, it is possible
that the senior lenders had more bargaining leverage, since negotiations occur in the shadow of a
potential cramdown and the threat of the case being moved into Chapter 7.
In most cases, there is a high likelihood that some form of liquidation imposed if a reasonably
prompt reorganization cannot be achieved and, this risk is exacerbated in CNC’s case where
insurance regulators were ready to appoint receivers for Conseco's subsidiaries if restructuring
was not quickly completed. This might explain why the TOPRS finally caved in and agreed to a
settlement pegged at the lower $3.8 billion valuation level.
As for management’s incentives in the negotiations, the general view is that management veers
towards a low valuation where the plan provides equity incentive compensation to management
priced in conjunction with the valuation, especially if they do not already own a significant share
60 Supra, n37.
61 Fitch Rating Watch, Fitch Research Report (2002).
62 See, generally, Gilson, S. et al, Valuation of Bankrupt Firms, Review of Financial Studies v13 (2000) and Moyer, S.,
Distressed Debt Analysis (J Ross Publishing, 2005) at p336-9.

63 Supra, Gilson, n62 at p66.
64 The Oppenheimer Group had a substantial stake in the TOPRS and was in fact on the TOPRS committee.
Interviews would be conducted for the next draft to explore in greater detail the impact of the participation of
vultures in CNC’s restructuring process.

of the equity. This hypothesis is supported in CNC’s case where the management argued for a
lower valuation. The management was relatively new and the proposed plan included a
management incentive plan giving key executives 10% of equity holdings in the form of
restricted stock. 65 This suggests that we should include variables relating to management’s post-
petition stock incentive plan and management turnover.
Nonetheless, we note that CNC’s management was driven by another key factor which might be
harder to model. Management found it easier to convince the senior lenders to take a
combination of reinstated debt and preferred stock and the noteholders to take common stock,
where a lower valuation provided some upside for their acceptance of riskier instruments. In its
legal memorandum, CNC stated that, if it had established a higher valuation, that would have
affected the strategy to emerge with a lower leverage ratio. 66
Finally, we observe that the uncertainty over the outcome of the valuation generates option value,
and this directly influences the allocation of securities whose value is contingent on a higher than
expected valuation. 67 In the final settlement between the company, its senior lenders and the
TOPRS, the latter were awarded warrants where the payoff was directly tied to the company
attaining a high future value. At an exercise price of $27.60, the TOPRS would be “in the
money” only if the company was indeed worth around $5 billion – the amount they asserted
during proceedings.
6.4 Process: Section 363 Sale
In this section, we would briefly discuss whether the reorganization process itself, specifically the
section 363 sale of CFC, substantially affects valuation. LoPucki and Doherty published a
seminal work in late 2007 which found that recoveries from reorganization cases yielded more
than double the recoveries from section 363 going concern sales and identified various sale
characteristics contributing to the firesale. 68
Many of these characteristics were present in CFC’s case. It was a rushed sale - two days into the
Chapter 11 petition, CFN was chosen as the stalking horse and entered into an asset purchase
agreement with CNC. CFN was protected from “outside” competition though the use of bid
protections. The bid protection in this case consisted of a breakup fee of $30 million
representing 4.0% of the stalking horse price and a requirement for an incremental bid of 5.3%
higher than the stalking horse bid. This was way higher than historical estimates where breakup
fees averaged 2.3% of the stalking horse price and terms of sale requiring the competing bid to,
65 In March 2004, the restructured CNC announced that its top executives made nearly $34 million in awards
primarily from restricted stock granted during Chapter 11 proceedings – see the annual report for the fiscal year
ended December 31, 2003.
66 Supra, n27.
67 Supra, Baird and Bernstein, n26.
68 LoPucki, L. and Doherty, J., Bankruptcy Fire Sales, UCLA Law & Economics Research Paper No. 07/07 (2007),
available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=980585.

on average, be at least 3.7% higher.69 There were also bulk-sale and all-cash requirements which
the Unofficial Committee of Noteholders objected to as procedures which would “chill potential
bidders from submitting bids and maximizing value”. 70 Furthermore, CFN was a DIP lender with
a significant information advantage over other bidders. 71
As expected (given the crucial advantages provided to the stalking horse), CFN won the auction
with its final bid of $970 million in cash, plus assumption of certain liabilities, subject to certain
adjustments and a restructuring of CFC's guarantees and fees in relation to the MH servicing
business. A twist to the story happened, raising the final purchase price. Fannie Mae held an
adequate protection lien and would only waive this in the event that Berkadia (a joint venture
between Berkshire Hathaway and Leucadia National Corp) won the auction.72 After further
negotiations, the purchase price was raised to approximately $1.3 billion and gained the support
of the major constituencies for the sale.
Nonetheless, though CFC’s section 363 sale fit the paradigm described by LoPucki and Doherty,
the process itself probably had less influence on the valuation than the sources of valuation
uncertainty discussed above. Following the order for sale, CFC reported in April 2003 a net loss
of $1.1 billion for the 3 months ended December 31, 2002, resulting in a shareholder's deficit
was $100 million.
7 The Subprime Lending Subplot

An interesting question is: how did the creditors fare in CFC’s sale and liquidation? At the time
of liquidation, CFC's primary debt obligations consisted of: (i) DIP financing;73 (ii) Warehouse
credit facilities mainly owed to Lehman; (iii) Liabilities under guarantees of securities issued
under securitization transactions.
On Lehman’s part, it faced adversary proceedings brought by the Official Committee of

Unsecured Creditors of CFC started adversary proceedings against Lehman arguing, amongst
others, for avoidance for fraudulent or preferential transfer and equitable subordination,
especially in relation to the 2002 transfers and fees incurred.74 The key argument proffered by the
unsecured creditors, however, was substantive consolidation of CFC and two special purpose
entities (SPEs).
69 Id.
70
Note, also, various objections entered by securitization trustee, noteholders committee, etc.
71
Baird, D. and Rasmussen, R., Private Debt and the Missing Lever of Corporate Governance, 154 U. PA.L.Rev. 1209 (2006)
(commenting on the information advantage of a DIP lender who was concurrently a potential purchaser).
72 Daily Deal, Berkshire still on Conseco deal trail (March 12, 2003).
73 This includes the US Bank swingline facility which was rolled into the super-priority DIP Facility Agreement.
74 Statement of Position for the Estimation of Lehman Claims by The Official Committee of Unsecured Creditors,
May 7, 2003. See retort by Lehman in Objection of Lehman ALI Inc., Lehman Brothers Inc., Lehman Commercial
Paper, Inc. and Lehman Brothers Holdings Inc., to Finance Company Debtors’ 2nd Amended Joint Liquidating Plan
of Reorganization Pursuant to Chapter 11 of the U.S. Bankruptcy Code, June 6, 2003.

These SPEs were the direct counterparties with Lehman for its warehouse facilities, though CFC
and CIHC provided guarantees. These SPEs were set up for the sole purpose of being a conduit
between CFC and Lehman, and to ringfence Lehman’s collateral in certain CFC assets transferred
to the SPEs. They had no other counterparties or business purpose. The court ruled in favor of
the substantive consolidation during the hearing, thereby eliminating Lehman's exclusive claim to
the assets of the SPEs.
In spite of this setback, Lehman as the senior secured lender recovered in full the principal and
interest under the warehouse and residual facilities (i.e., a recovery rate close to 100%), though it
was given a haircut adjustment for other claims in relation to additional fees incurred under
forbearance agreements and other arrangements undertaken in 2002. This followed from the fact
that the purchase price in the section 363 sale was primarily set to ensure that the senior secured
liabilities would be paid off. In the Disclosure Statement, it was stated that the purchase price
payable by CFN would be the sum of the Lehman Secured Debt and the DIP facility amount
and approximately $159 million. 75
On the other hand, the holders of CFC’s asset-backed securities fared worse. While the
securitization trusts were not brought into bankruptcy and bankruptcy-remoteness issues were
not at issue, the securities holders were adversely affected in three major ways:
i. Many tranches faced downgrades and defaults in light of the performance deterioration
of the manufacturing housing loans asset pool and the decline of over-collateralization
beyond target levels. These losses were exacerbated by CFC’s bankruptcy filing which led
to changes in servicing practices. Since the suspension of CFC’s lending business, CFC
was forced to liquidate repossessed units through wholesale channels rather than retail
channels. In addition, CFC discontinued its financing and assumption programs to reduce
servicing costs, causing repossessions to spike as loans previously eligible for this
program were then taken as repossessions. This led to a flood of repossessed units in the
wholesale market, further elevating loss severity rates.
ii. There was a reduction in the amount of excess spread available to cover losses. Excess
spread had been reduced since the court issued an interim order increasing the amount
and priority of the servicing fee.76 The fee rose by more than 100% from 50 bps to 125
bps in 2003 and 115 bps in 2004. Prior to this order, the servicing fee was to be paid as
an expense prior to distributions to the tranches, so the order essentially rewrote the cash
waterfall.
iii. CFC provided $2.3 billion of guarantees for subordinated tranches, known as B2
certificates, of which $600 million had fallen due at the time of its liquidation
75 Re Conseco, Inc., Second Amended Disclosure Statement for Finance Company Debtors' Second Joint Plan, July
15, 2003.
76 See Re Conseco, Inc., Stipulation and Order on Continuation of Interim Servicing Arrangement, June 19, 2003.
These changes were in exchange for the grant of an adequate protection lien on CFC's assets in favor of the
securitization trust and the Trustee.

proceedings. In its final plan, the B2 guarantee claims were resolved in the aggregate
amount of $34 million, i.e., a low 6% recovery.77 This dramatic reduction in credit
enhancement, coupled with the weak performance of the underlying asset pools,
produced multiple downgrades and defaults. It was reported that out of a MH default
sample over 1991-2005, securities issued by CFC accounted for 36%. 78 This problem also
affected ABS deals backed by other types of assets. A record of 50 RMBS defaults were
recorded in 2003, of which 33 resulted from guarantor defaults by CFC.79
8 A Closer Look at the Meaning of “Ultimate Recoveries”

A final issue arising from this case study is how we should conceptualize “ultimate recoveries”.
Currently, post-default trading prices are more commonly used to measure LGD, given the
relative ease of collecting market data. However, there have been an increasing number of banks
which considered ultimate recoveries to be a more accurate LGD measure and decided to use
ultimate recoveries as an input in the Basel II regulatory capital formula.
However, in a case where creditors receive a debt-equity swap in the restructuring process, should
the ultimate recoveries be recorded as those at the time of emergence, i.e., the levels listed in
Figure 5? As discussed above, negotiation dynamics had, to much extent, driven the valuation of
the firm which was fiercely contested by the TOPRS. The final reorganization plan valued the
common stock at $16.40. Events following CNC’s emergence from Chapter 11 for the next 3
years suggest that the $16.40 estimate might be too low.
Ever since the stock of the newly-restructured CNC floated on the NYSE till the end of June
2007, 80 the average price was $21 and the price had only dipped below $16.40 for 4 days during
this period. In fact, CNC completed a recapitalization in May 2004 with a secondary offering of
44 million shares of common stocks at a price of $18.25. Based on the average price of $21,
there was a 28% increase in the ultimate recoveries for the creditors ranging from the Exchange
Notes holders to the TOPRS.
77 Note that this recovery is limited to guarantee claims in relation to the B2 certificates which were on the record as
of the day of the order confirming the final liquidation plan, i.e., CFC had no further contingent liabilities under
these guarantees – see Order Confirming Finance Company Debtors' Sixth Amended Joint Liquidation Plan of
Reorganization Pursuant to Chapter 11 of the United States Bankruptcy Code (September 9. 2003).
78 Fitch Global Structured Finance 1991-2005, Fitch Ratings (November 28, 2006).
79 U.S. RMBS Upgrades Are Down And Downgrades Are Up In 2006, Standard & Poor's (January 27, 2006).
80 In this analysis, we have excluded prices subsequent to that period, given the volatility and contagion generated in
the 2007-8 market turmoil.

Figure 12: Stock Price of the Restructured Conseco, Inc.
Source: Bloomberg.
Figure 13: Range of Recoveries based on Different Stock Prices
Discounted Discounted
No. of Shares Nominal Recovery Nominal Recovery
Recovery @ Recovery @
( millions) @ $16.40/share @ $21/share
$16.40/share $21/share
Exchange Notes 60.6 72.0% 92.2% 65.5% 83.8%
Original Notes 32.3 42.0% 53.8% 38.2% 48.9%
CNC Unsecured Claims 1 22.0% 28.2% 20.0% 25.6%
CIHC Unsecured Claims 1.9 100.0% 128.0% 90.9% 116.4%
TOPRS 1.5 1.3% 1.6% 1.2% 1.5%
At a price of $21, it appeared that the market's valuation was more than $4 billion – a level close
to the use of a discount rate of 10%, rather than 12% used in the plan).81 Part of the market’s
perceptions of CNC’s enterprise value also stemmed from it realizing a substantial net operating
losses (NOLs) carryforward. During the valuation trial, one of the TOPRS’ key arguments was
the company’s failure to sufficiently include NOLs in establishing the valuation. 82 This was
successfully resisted by management and senior creditors. In 2004, Conseco and the IRS entered
into “a closing agreement which determined that the tax loss recognized on the worthlessness of
81 Seeappraisal report done by Milliman USA Consultants and Actuaries.

82
For more details, see Re Conseco, Inc., Expert report of Wilfred J. Romero of Fox-Pitt, Kelton Inc. on behalf of
the Official Committee of Holders of Trust Originated Preferred Debt of Conseco, Inc., May 7, 2003, attached to
The Official Committee of Conseco Trust Originated Preferred Debt Holders’ Emergency Motion in Limine to
preclude debtors from offering expert opinion testimony from William T. Devanney filed June 16, 2003.

CFC was $6.7 billion, instead of our original estimate of $5.4 billion”. 83
Put simply, how appropriate is it to use ‘ultimate recoveries’ at the point of emergence as a
measure of LGD if the enterprise value was the result of ‘low-balling’ by parties with stronger
bargaining leverage in negotiations? Should the subsequent average stock price be used to arrive
at the recovery rate instead? Besides, how should the value of other instruments such as the
warrants and preferred stock be properly determined? In this case, the preferred stock issued to
the senior lenders was redeemed at par plus accrued dividends in 2004, 84 but the warrants were
worthless since the stock price never rose up to the level of the exercise price.
These questions are unlikely confined to CNC’s case. A 2005 report by Jefferies which maintains
an index tracking post-Chapter 11 equities reported that the average re-org equity, based on 62
companies which emerged since 2000, delivered excess return of 67% compared to the S&P
500. 85 In light of this, we propose further investigations in the other case studies and cross-
sectional data in addressing this issue.
83Re Conseco, Inc., Annual Report (10-K filing) for fiscal year ended Dec 31, 2004. The loss was considered an
ordinary loss for tax purposes, which meant that it could apply generally, instead of being limited to reduce future
capital gains.
84See Conseco, Inc’s press release in its Current Report (Form 8-K), May 12, 2004.
85Special Situations: Risk and Reward in Post-Reorg Equities, Jefferies Research (October 2005). The post-reorg returns
reported by Jefferies, which only started tracking the performance of companies emerging from bankruptcy in 2000,
could have been driven by the 2003-7 boom.


Conseco: Restructuring An Insurance Giant With A Subprime Lending Arm

Uploaded by

Copyright:

Available Formats

Conseco: Restructuring An Insurance Giant With A Subprime Lending Arm

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Conseco: Restructuring An Insurance Giant With A Subprime Lending Arm

Uploaded by

Copyright:

Available Formats

Case Study: Conseco, Inc.

Sarah Woo, Stanford University

Working Paper 2/4/08 Page 1

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

1 Objectives of Case Study

2 Overview and Chronology of Events

2.1 The Company

2.2 Organizational Structure

Working Paper 2/4/08 Page 2

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

refer to the complete organizational structure provided in Appendix A.

Figure 1: Organizational Structure of Conseco Group

2.3 Initial Signs of Distress

Working Paper 2/4/08 Page 3

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Working Paper 2/4/08 Page 4

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

2.4 Events Leading to Default

2.5 Chapter 11 Proceedings

10 Supra, n2, pp26.

Working Paper 2/4/08 Page 5

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Figure 2: Post-Petition Disclosure of Liabilities

Working Paper 2/4/08 Page 6

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Figure 3: Capital Structure of the Restructured Conseco, Inc.

Capital Structure of Conseco, Inc. at Emergence

Preferred Stock 860

3 Comparing Recoveries to Historical Estimates

Working Paper 2/4/08 Page 7

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Figure 4: Historical Mean Recovery Rates (1988-2003)

Nominal Ultimate Discounted Ultimate Trading Price at

Source: S&P’s Research (2004).16

Figure 5: Creditors' Ultimate Recoveries in CNC

Claim Description Plan Treatment Nominal Discounted

Source: Conseco, Inc.’s 6th Final Reorganization Plan.

Working Paper 2/4/08 Page 8

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Figure 6: Post-Default Prices of CNC's Traded Instruments

Post-Default Trading Prices

17 Altman, E. et al, Corporate Financial Distress and Bankruptcy (2006), pp328.

Working Paper 2/4/08 Page 9

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

4 Effective Leverage and Relative Seniority

CNC provides an instructive example as to certain types of adjustments which we should

Working Paper 2/4/08 Page 10

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Figure 8: Relative Seniority of Claims on CNC

Working Paper 2/4/08 Page 11

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

Claim Description Estimated Seniority/Collateral Relative

Source: Conseco’s 2nd Amended Disclosure Statement, March 18, 2003.

5 Deviations from the Absolute Priority Rule (“APR”)

Working Paper 2/4/08 Page 12

This working paper is licensed under a Creative Commons Attribution-Noncommercial-Share

5.1 Irregularities in Prior Restructuring Mechanisms

25 Supra, Weiss, n24.

Working Paper 2/4/08 Page 13