Two collection agencies - which bought delinquent credit card and other bank debt - raised millions based upon misrepresented collection histories and the misapplication of GAAP.   

Bill Bartmann liked to be larger than life. After the death of his failed oil services company in the 1980s, and his traumatic experiences with creditors (whom he owed nearly $1 million), Bartmann became oddly destined to be what he liked to call a "collector with a heart."

Starting with $13,000 in working capital in 1986, in the midst of the thrift industry failure, Bartmann formed Commercial Financial Services (CFS) and began purchasing defaulted loans at steep discounts from the Federal Deposit Insurance Corporation (FDIC) and the Resolution Trust Corporation (RTC). By 1992, Bartmann's collection success produced a viable business with 45 employees.

Six years later, CFS had sold $2 billion in delinquent credit card receivable and other bank debt-backed securities and CFS had grown to 3,900 employees, making it Tulsa, Oklahoma's ninth-largest employer. CFS made Inc. Magazine's fastest growing company list three consecutive times and Forbes estimated that Bartmann and his wife, Kathryn, were worth $530 million - each.

A generous man, Bartmann became a civic leader, magnanimous not only to his community, but to his employees. He financed a fledgling singing career for one of his staff and sent another to Calcutta to realize a life-long dream of working with Mother Teresa.

But Bartmann was no saint. His empire would soon crumble - the victim of misrepresenting the collections he was realizing from the delinquent debt he purchased and the disregard generally accepted accounting principles (GAAP). Fooled once, Wall Street should have understood the scam that cost investors hundreds of millions of dollars. But in early March of this year, the FBI raided the former offices of Creditrust, a one-time CFS competitor, and while it may be too early to tell, Creditrust may well have committed the same fraud using the same techniques. Déjà vu all over again?

CFS generated purportedly massive profits by purchasing delinquent credit card debt from credit card issuers at a fraction of face value. Banks and other issuers typically sold the delinquencies at 2 cents to 12 cents on a dollar of face amount, depending on the age of the receivable and other factors. Issuers had incentive to sell the delinquencies because the recovery process was lengthy and labor intensive even if the receivables were placed with outside collection agencies.
Using a kid-glove approach - no pressure or threats - with debtors, Bartmann told analysts that for every 10 cents he spent, he could collect an average of 35 cents. With CFS collections apparently, but falsely, supporting his contention that CFS could more than triple its investment, Bartmann pioneered the securitization of delinquent credit card receivables by pledging thousands of delinquent accounts to secure the bond holders who funded his debt purchases.

Asset-backed Securities Nothing New

Industry experts were skeptical that CFS was collecting three times its cost basis. But the phenomenal growth of CFS was keyed to Bartmann's ability to borrow substantial sums of relatively low-cost money based upon estimates of future collections, which in turn, were based upon false CFS collection histories.

Not content with limitations imposed by traditional lenders, Bartmann literally invented the securitization of delinquent credit card receivables. Wall Street was receptive. Indeed, the great investment banking firms had already figured out the synergy of asset-backed debt instruments - initially bonds secured by pools of home mortgages, each paying off over time. The more ways bankers parsed debt instruments - their cash flows, their risks, their security - the more commissions and trading profits there were to be earned. And the more complex the vehicle was, the more dependent customers became on the investment banker and rating agencies who presumably understood the nuances of these new products. Spreads were large and life was good.

Slicing the Asset-backed Pie

Asset-backed securities are generally little more than loans secured solely by a pool of largely homogenous smaller loans, each with similar credit risks. In the event of default, bond holders can look only to the assets that collateralize the bonds, but not to the issuer. The most common asset-backed securities are mortgage-backed bonds, secured by large pools of individual home mortgages, although almost any kind of debt has or can be securitized.

In the late 1970s, investment bankers became increasingly creative in the ways these securities were carved up. Typically, asset-backed securities are broken down into two parts:

• principal and interest, and
• priority of payment.

Mortgage-backed bonds, for example, are often split into interest only and principal only components and then each sold at a discount from the price as-a-whole; but just as typically, the parts often have more value than the whole. Purchasers of the interest-only component buy at the steepest discounts, betting on stable to rising interest rates. In such an environment, the homeowner's disposition to refinance is diminished so the stream of interest payments likely continues to flow. Conversely, purchasers of the principal-only component speculate that interest rates will decline to a level, which makes refinancing the logical choice. Because principal-only components are also bought at a discount, the faster the principal is repaid at the face amount, the greater the return on the principal-only investment.

Priority of Payment

Asset-backed securities have also been structured with varying priorities of payment. Suppose, for example, one issues $100 million of bonds secured by $100 million of credit card receivables. In such a case, the risk of repayment is spread equally among purchasers of the $100 million of bonds.

However, if the first $20 million of bonds had the absolute right to the cash flows of the entire pool, then this piece - typically called a "tranche" - would be far less risky that the last tranche, that had the right to the cash flows of the entire pool only after the senior tranches were paid. In many such transactions, the issuer would hold the last tranche - called the first loss tranche - in effect, over-collateralizing the senior loans.

The CFS Wrinkle

Bartmann's genius was to securitize delinquent credit card receivables in a way asset-backs had never before been structured. Using CFS collection histories, he persuaded his bankers and bond purchasers that the basis for the loan should not be a percentage of the cost of the collateral - the pool of delinquent accounts - but upon its estimated value.
If for example, CFS paid 10 cents on the dollar and collected 35 cents, Bartmann argued that the loans should be based on 60 percent of anticipated collections. In other words, under some circumstances, CFS was borrowing 21 cents on a pool of assets that cost it 10 cents. And the bonds were safe, Bartmann argued, because he was borrowing only 60 percent of the value of the collateral that secured them.

The Work of the Devil

In theory, the practice made sense, but human behavior, being human behavior, isn't so easily predicted. If idle hands are the work of the devil, what does that make idle cash? The simple fact was that economics of these transactions were based entirely upon histories of CFS collections of similar assets that no one apparently verified in detail.
Despite some industry critics, CFS thrived. The more successful CFS became, the more its competitors' imitated CFS's behavior. Delinquent receivables became more expensive as the market became more active. Nevertheless, the ratings agencies literally gushed over CFS, assigning investment grade ratings to its securities and to its servicing abilities. Moreover, theoretically correct, but (in my judgment) misguided accounting rules, assured that Bartmann would get very rich very quickly.

Auditors Fail Again

Accounting for these transactions - both the purchase of and subsequent securitization of receivables - was, and remains, a reasonably straightforward process in theory but a quagmire of detail in practice.

Accounting for receivables purchased at a discount is guided by AICPA Practice Bulletin 6. The essence of the guidance is that when a loan is purchased at a steep discount, each payment received must be allocated between interest income and the repayment of principal.

In its simplest terms, the calculation is made on management's best estimate of the amounts ultimately collectible.
Similarly, when CFS pooled such receivables and securitized the package, it didn't give up the rights to all the cash flows from collections, it merely pledged all the cash flows until the asset-backed bonds were repaid in full, both principal and interest. The remaining collections belonged to CFS and, theoretically, had substantial value. After all, CFS typically paid 10 cents with the anticipation it would collect 35 cents within a few years. And because the bonds received investment grade ratings, the interest payments were reasonably cheap, generally within 200 to 250 basis points of the 10-year Treasury.
Thus, the residual had substantial, theoretical value. What cost 10 cents and sold for 18 cents might permit the recognition of 16 cents of gain before a single dollar was collected. In accordance with GAAP, in this case FASB Statement No. 115 regarding Accounting for Certain Investments in Debt and Equity Securities, the fair value of the residual interest reduced the cost of the pool of receivables sold. And how was fair value determined? By an estimate of management, of course.

The Cookie Crumbles

Bartmann's money-making machine chugged along for three years, despite the fact that CFS auditors were required to test both CFS internal controls as well as its estimates.

Statement on Auditing Standards No. 57 requires auditors to obtain sufficient competent evidential matter to provide reasonable assurance that those accounting estimates are reasonable in the circumstances. 

Furthermore, auditors, in evaluating the reasonableness of an estimate, are required to: Test the calculations used by management to translate the assumptions and key factors into the accounting estimate. 

Despite some clear guidance, the CFS auditors didn't discover the fraud. In fact, what was apparently occurring was that affiliates of CFS were purchasing receivables and the proceeds of such purchases were being accounted for as collections. In some instances, 60 percent of collections on some pools were, in fact, purchases. The impact of this falsified accounting was that CFS collection histories continued to look good when, in fact, CFS collection forecasts weren't being realized.
And because both interest income (in the case of owned receivables) and gains on securitizations (in the case of pools that were securitized) were determined largely upon estimates based upon these collection histories, the financial results of CFS were as big a fairy tale as Bartmann's billions. Reportedly, the auditors never discovered the scam. As is often the case, an internal whistle blower, in October 1998, wrote an anonymous letter to the ratings agencies exposing the fraud. Their reaction was to suspend the ratings and ultimately, to withdraw them altogether.

Predictably, CFS filed for bankruptcy and the complaints and other pleadings began to fly. The matter received some national publicity, but because the bonds had been sold in private placements, the full impact of the debacle has never been adequately described. CFS assets, the massive pools of delinquent credit card receivables, were sold off at steep discounts. As we shall see, some of these assets were sold to Creditrust, a one-time competitor, that may have been committing the same kind of fraud.

History Teaches Nothing

With the discovery and playing out of every financial scandal, the financial press invariably runs post-mortems about lessons learned which, invariably, rarely are.

In this case, results are mixed. In July 1998, for example, Creditrust - a CFS competitor - was also reporting substantial profit from securitizing delinquent credit card receivables. Based upon earnings inflated by securitization gains, Creditrust went public. The firm raised nearly $29 million for the company and $7.2 million, in a secondary offering, for its founder, chairman, and CEO. The combined sales represented 31 percent of the company, giving it a market value of approximately $124 million. Such values were commanded not because of Creditrust's earnings of $456,000 in 1997 or its first quarter earnings of $415,000 in 1998, but because of some $6.1 million of gains the company recognized on a securitization completed in June 1998, just prior to its initial offering.

In January 1999, shortly after the CFS debacle became apparent, Creditrust returned to the equity markets to raise another $45.6 million for the company and $11.4 million for its founder. At $19 per share, the market capitalization of the Creditrust grew to $197 million. This 59 percent increase in value was apparently the result of $2.8 million of reported earnings for the three-quarters ended Sept. 30, 1998. Without gains on its securitizations of $7.4 million, however, Creditrust would have realized significant losses for the period.

The 1999 Creditrust offering was to be the last hurrah of "gain on sale" accounting. While the price of Creditrust steadily climbed throughout much of the year, it peaked at $34 per share in September. Then the dam broke.

Wall Street Loses the Faith

In early September 1999, Conseco, the Indiana-based insurer, and itself something of a Wall Street darling, announced it would no longer recognize gain on sale accounting in connection with its securitizations of high-risk, high-yield mortgages. (Conseco had been a high-flyer until its 1998 acquisition of Green Tree Financial, a consumer lending concern that also used gain on sale accounting. At the time of acquisition, Conseco had taken a $350 million charge on Green Tree's loan portfolio, but continued to recognize massive gains as a result of the securitizations of Green Tree's consumer loans.)

At the time of its announcement, Conseco said the accounting change was to shore up investor confidence and concurrently announced its earnings would likely decline 48 percent for the remainder of the year.

In the face of the Conseco announcement, Creditrust also acknowledged it would no longer report gains currently from its securitizations. Creditrust had issued unusual interim financial statements, computing earnings both with and without such gains, for its second quarter but gave no indication it would abandon its aggressive accounting.

But in September 1999, Creditrust took the pledge and foreswore gain on sale accounting for good.
"In keeping with our business model, our conservative accounting standards, and our desire to provide our shareholders with a steady and predictable earnings stream, this securitization will not be recorded utilizing gain on sale accounting methods," a spokesman for the company said in a press release.

The market reaction was favorable. Creditrust shares rose to $34 per share. Having raised almost $100 million of equity on the gains from securitization, Creditrust could afford to be magnanimous.

Déjà vu All Over Again - a Whiff of Fraud

In early December 1999, Creditrust soberly announced that a senior executive of the company had attempted to divert some $500,000 of company funds into its collection streams. In other words, a senior executive was apparently attempting to falsify crucial Creditrust collection histories.

Creditrust said it learned of the scheme in October 1999, during the period its stock price was collapsing from its high of $34 per share to $20. The company made no announcement of the diversion at the time, however, nor did it report the attempt to its audit committee, its auditors or the public.

With the December disclosure, Creditrust's stock continued its slide, declining from about $19 per share to the $12 to $13 range. In mid-December, the company announced it had engaged advisors for strategic planning and the stock declined to the $8 to $9 range.

Since then, class action litigation against Creditrust officers has been commenced, Creditrust failed, and its assets have been sold. Similarly, the officers of CFS have been sued as have their auditors by angry institutional investors who know they have been had. In the meantime, Conseco announced it was planning to sell Green Tree and angry investors quickly filed suit alleging false and misleading representations concerning the value of Green Tree's residuals from its consumer loan securitizations .

Lessons Learned

So what have we learned? Accounting estimates always have and always will invite fraud. The more complex the estimate, the more easily the public is fooled.

Even so, the debacle of CFS didn't prevent Creditrust from raising almost $95 million in equity from the public, much of it subsequent to the CFS meltdown.

So while what we should have learned is that accounting is neither exact nor a science, we must conclude that even the most sophisticated analysts are frequently swept up in apparent innovations of the moment.
Moreover, the very complexity and innovation involved in the CFS and Creditrust debacles may have simply been a smokescreen, masking the ugly answer to the fundamental question: will these instruments pay? Perhaps because of the volumes of accounts involved, no one apparently verified Creditrust or CFS collection histories or assured themselves that new pools of receivables were likely to collect as well as the old pools, which had demonstrable, verifiable histories that could be tested.

Further, industry critics of both CFS and Creditrust maintained for years that collections on delinquent receivables have always ranged from 15 cents to 23 cents on the dollar. Both CFS and Creditrust publicly represented that their collections histories were substantially greater and the ratings agencies, Wall Street, and the auditors believed them, apparently unaware of industry norms. Abnormal performance, however presented, should be viewed skeptically and if possible, tested objectively.

Wall Street and investors took a massive leap of faith on CFS and Creditrust, that with hindsight, was clearly misplaced. In the final analysis, the CFS and Creditrust experiences reinforce the simple, but time-honored notion that even in the most complex of circumstances, if some things seem too good to be true, they often are

Gordon Yale, CFE, CPA, is principal of Yale & Company, a Denver, Colo.-based forensic accounting and financial consulting firm and the Yale Group Inc., an investment bank. Yale & Company specializes in securities fraud and audit negligence matters.

Sources
Alaimo, Diane; Punch, Linda. "An End to an Era?" Credit Card Management, p. 29, Jan. 1, 1999, UMI Company, Faulkner & Gray, Inc.
"Creditrust Comments on Market Activity," Creditrust press release, Oct. 29, 1999.
"Creditrust Buys $546M Commel Finel Sves Portfolio >CRDT," Dow Jones & Company, Aug. 9, 1999.
Creditrust 8-K, dated December 13, 1999
Haring, Bob. "Kinder, gentler debt collector finds huge success," USA TODAY, June 24, 1998.
"Insurer Conseco Says Accounting Change Will Hurt 1999 Earnings," Dow Jones Business News, Sept. 8, 1999
Rigby, Bill. "USA: Shareholders sue Conseco over former Green Tree unit," Reuters English News Service, April 7, 2000.
Shafer, Shaun. "Tulsa, Okla., Debt Collection Agency's Secret Dealing Led to Failure," Knight-Ridder Tribune Business News: Tulsa World via Dow Jones & Company, March 18, 2000.
Shafer, Shaun. "Chicago-Based Accounting Firm Faces Suit for Bad Audits," Knight-Ridder Tribune Business News: Tulsa World via Dow Jones & Company, Feb. 9, 2000.
Shafer, Shaun. "Evidence Emerges of Illegal Tulsa, Okla., Bond-Form Collaboration," Knight-Ridder Tribune Business News: Tulsa World via Dow Jones & Company, Oct. 7, 1999.
Shafer, Shaun. "Bondholders File Lawsuits Against Tulsa, Okla.-Based Financial Firm," Knight-Ridder Tribune Business News: Tulsa World via Dow Jones & Company, Oct. 14, 1999.
Shafer, Shaun; Fallis, David; Bryant, Julie. "The Rise and Fall of CFS," Knight-Ridder Tribune Business News: Tulsa World via Dow Jones & Company, June 1, 1999.
"SMART 1996-4 ABS Certificates Downgraded by Moody's," Dow Jones & Company, Dec. 15, 1998.