Commercial Financial Services: An Early Warning Nobody Heard

Book: Structured Finance and Collateralized Debt Obligations | Author: Janet M. Tavakoli | Publisher: John Wiley & Sons (2008) | ISBN: 978-0-470-44344-6

Chapter 7 is a case study. It’s the story of Commercial Financial Services, a Tulsa, Oklahoma company that issued roughly $2 billion in asset-backed securities during the 1990s, watched the structure collapse into fraud, and filed for bankruptcy in December 1998.

Tavakoli tells it as a detailed warning. Not a historical curiosity. A preview.

Because the exact same pattern would play out again in the subprime mortgage market, just at vastly larger scale.

What CFS Did

CFS was co-founded in 1986 by William Bartmann, who owned 80 percent, and Jay Jones, who owned 20 percent.

The business model was unusual. CFS bought charged-off credit card receivables - debt that banks had already written off as uncollectable after about six months of no payments. It bought these at steep discounts, funded the purchases through nonrecourse warehousing lines, and then securitized the receivables as asset-backed securities. Securitization proceeds paid down the warehouse debt and funded CFS’s operations. Servicing the loans was a second income stream.

During its brief heyday, Bartmann was rumored to have a net worth of $4 billion to $7 billion. He flew on the company’s Gulfstream G-IV. He threw parties for thousands of people, chartered Boeing 747s to fly them to Las Vegas, gave them each $500 to gamble, and dressed as Julius Caesar at one bash.

Sound familiar? Tavakoli notes that this pattern of lavish excess was copied by Enron, RBG metals, Tyco, and others.

Why This Business Model Was Fragile

CFS needed an investment-grade rating to access a large investor base at low cost. To get that rating, its securitizations had to be structured as static pool cash flow deals. That meant: a fixed pool of assets, no asset sales, returns generated entirely by cash collections from the receivable pool.

For CFS to meet its cash flow obligations, it had to:

• Convert enough nonperforming loans into performing loans through negotiated payment plans
• Collect on those performing loans
• Negotiate enough lump-sum settlements

The rating agencies required a 3 percent reserve account and overcollateralization of 30 to 50 percent. That high cushion may have created a false sense of security rather than signaling how uncertain the underlying cash flows actually were.

Charged-off credit card receivables were a new asset class. No one had much data. Three separate collection agencies had already tried and failed to collect on these same receivables before selling them to CFS. CFS’s business was still young. Arthur Andersen handled CFS’s books out of its Tulsa office, with a junior partner in charge.

Early Red Flags: Ignored

In the late summer of 1996, the lead underwriter’s ABS group sought credit approval for a warehouse loan. A senior credit officer raised concerns that CFS presented “a classic situation for fraud.”

The reasons were all there:

• Bartmann and Jones were not well known
• The loans were growing at a very fast rate
• It wasn’t clear how money was being collected from people who had previously failed to pay
• The securitizations were priced at huge multiples of purchase price based on optimistic collection forecasts
• There was no way to independently verify the assumptions
• Cash flow schedules were confusing

CFS was also underestimating its own servicing costs. Its financial statements did not show a servicing liability for the excess costs. This could have been verified immediately.

The underwriter could have requested a fraud audit. A fraud audit doesn’t mean you suspect someone - it means you want rigorous verification of assets and cash flows. Instead, the underwriter seemed to accept Bartmann’s explanation: CFS was more charming and diligent in collecting from debtors than previous collectors had been.

The rating agencies did no better. S&P used data from unsecured consumer loans as a proxy for charged-off credit card receivables, even though it was unclear these assets performed the same way. No relevant data was available. As more data came in over time, the agencies did not pursue the red flags.

In early 1996, a potential investor noticed that CFS data only covered performing assets, not total collections. CFS’s model was untested. Arthur Andersen had not tested it for accuracy. That investor walked.

CFS Gets Creative

The core problem: CFS wasn’t generating enough cash flow from collections to meet its securitization obligations.

When you have a static pool securitization, you cannot sell assets from the pool to raise cash. The only money available to pay investors is the cash flow the assets actually generate. If collections are insufficient, the structure fails.

CFS found creative workarounds.

First, it began putting assets back to the original sellers - the credit card issuers who had initially sold CFS the receivables. The sellers would repurchase those receivables at the original purchase price. The problem: CFS had paid a premium when it sold assets into the securitization trust. When assets came back at the original lower price, CFS was taking a loss. It was recording these as “resolved collections,” which implied moneymaking activity. It was losing money.

Then it started selling performing current-paying assets to an outside buyer called Cadle at the beginning of 1997. Investors were not informed. The CEO at the time, Mitchell Vernick, who’d been hired in mid-January 1997, figured out what was happening within five months and resigned. He signed a nondisclosure agreement and received around $10 million in severance.

Vernick’s assessment: CFS’s expenses were so high that new securitization proceeds were being used to meet operating expenses. He doubted the investors could be repaid in full. His description: a classic Ponzi scheme. Raising cash from new investors to subsidize payments to existing investors.

The Audit Report They Didn’t Look At

In August 1997, the lead underwriter received an independent audit report it had commissioned on one of CFS’s securitizations. The report was damning.

The auditor sampled 20 payments. Three of them - 15 percent - came from assets that had subsequently been sold from the trust. In a static pool, no current-paying loans should be sold. Finding 15 percent of payments coming from sold assets was a major red flag.

Another sample showed 75 percent of “resolved loans” were actually puts back to the seller - money-losing, not money-making.

Another sample of 20 so-called performing loans found that 20 percent were actually behind on payments. Some by more than two months.

The audit painted the picture of a cash flow disaster.

The lead underwriter’s response: the investment bankers later said they hadn’t looked at the audit they had commissioned after Vernick’s resignation.

DIMAT and the Shell Game

After Cadle stopped buying CFS’s assets in the summer of 1997, CFS needed a new solution.

Jay Jones incorporated DIMAT Inc. DIMAT bought essentially worthless assets from CFS’s securitization trusts at inflated prices - sometimes two or three times market rate. The money that flowed through DIMAT back into the trusts made it appear as though the trusts were generating cash. In reality, new investor money was being used to make principal and interest payments to previous investors.

The mechanism: Bartmann and Jones sold CFS shares to raise cash. Jones transferred that cash to DIMAT. DIMAT used it to buy overpriced dead assets from the trusts just in time to meet the trusts’ monthly collection requirements.

The officers of DIMAT were unknown. No one asked who owned it. No one performed due diligence on the purchases. DIMAT consistently paid above market rate for worthless receivables, was happy to continue paying CFS servicing fees on those dead assets, and always appeared right at the end of the month when CFS needed to meet collection targets.

In November 1997, CFS wrote down the value of its residual interest in the securitizations to zero - claiming this was “conservative.” The residual, previously valued at hundreds of millions of dollars by CFS, the lead underwriter, and the rating agencies, was probably already worthless. No one demanded a rigorous audit to determine the actual value.

The Anonymous Letter

In the early fall of 1998, the rating agencies received an anonymous letter. It alleged that roughly 20 percent of cash collections were coming from asset sales, not actual obligor payments. It named DIMAT and noted that DIMAT’s officers were unknown. It described delinquent assets being reclassified as current by a single CFS officer over weekends.

Duff & Phelps Credit Rating (later part of Fitch) downgraded CFS’s securitizations six notches from single-A to BB - solidly investment grade to well below investment grade - in October 1998. Fitch IBCA then lowered ratings to CC, ten notches below investment grade, saying it had relied on CFS’s representations and hadn’t independently investigated asset sales it was already aware of. Moody’s and S&P followed.

All four agencies said they reacted to the anonymous tip-off. Not to their own analysis.

The facts had hidden in plain sight.

Fallout

CFS filed for bankruptcy in December 1998. Jones pleaded guilty to conspiracy and was sentenced to five years in prison. Bartmann was indicted on conspiracy, bank fraud, mail fraud, wire fraud, and money laundering - and was acquitted of all counts at trial.

Bartmann later reinvented himself as a motivational speaker.

The Lessons

Tavakoli lists them explicitly. They’re worth reading in full because each one corresponds to a failure that happened again in the subprime market a decade later:

Heed early warning signs. A senior credit officer raised the fraud concern in 1996. The concern was noted and bypassed.

Do a background check. Bartmann had a past cease-and-desist order related to an oil venture. The lead underwriter discovered this in a background check they ran a year into the relationship. Neither they nor CFS’s lawyers insisted the information be included in future offering documents.

If costs exceed fees, investigate. CFS’s servicing costs exceeded its projections. The discrepancy was known and went unexamined.

If the model is untested, say so clearly. CFS’s collection model had no historical validation. The rating agencies should have flagged this prominently rather than substituting data from unrelated asset classes.

Hire a special auditor for new asset classes. Particularly for fast-growing ones where data is thin. If the issuer won’t cooperate, walk away.

Do not rely on the rating agencies. They issue opinions. Those opinions may be worse than your own analysis, especially for new asset classes.

Do not rely on the lawyers. They can opine on legal issues. Business decisions are yours to make.

If a key officer leaves after a short stay, question them. Vernick’s resignation after five months was a bright red flag. No one talked to him before continuing the relationship.

If the issuer writes down a residual to zero, believe it is accurate, not conservative. And demand an independent audit before treating any equity valuation as real.

The Rerun

Tavakoli closes the chapter with a pointed observation.

All of this information was publicly available. The Wall Street Journal, Fortune, the New York Times, and Tulsa news sources all covered pieces of the CFS story. Bartmann’s criminal trial was public record. The case was well-known to securitization professionals.

Despite that, strikingly similar failures appeared in subsequent years with Enron, Parmalat, manufactured housing loans, metals receivables, furniture receivables, and eventually subprime and Alt-A mortgages.

The chapter is a demonstration that the lessons of structured finance failures are available. They’re just not learned.

Previous: Early CDO Technology: How It All Started

Next: Subprime and Alt-A Mortgages: How the House of Cards Was Built