cfo.com

Print this article | Return to Article | Return to CFO.com

Longer Paper Routes

Banks have gone to greater lengths to keep assets off their balance sheets. That means higher prices for commercial paper.
Hilary Rosenberg, CFO Magazine
October 1, 2003

Once upon a time, everyone in the asset-backed commercial paper (ABCP) market was happy. Companies got access to cheap financing and bankers reaped lucrative fee income.

The recipe? A company sells trade receivables or other high-quality assets to a bank-sponsored entity known as a conduit. The conduit pools those assets with those from other companies. The bank provides liquidity and credit enhancement to cover potential losses in return for fees. And the conduit issues commercial paper, the proceeds of which are used to buy more assets. Critically, the financing of those assets is not reflected on banks' balance sheets.

But Enron's implosion in 2001 convinced regulators that the good times for ABCP and other forms of off-balance-sheet financing had gone too far. So last winter, the Financial Accounting Standards Board issued a new rule, called Financial Interpretation No. 46 (or FIN 46), that requires companies that control off-balance-sheet entities that generate variable interests to shift the assets and liabilities of those operations onto their balance sheets. The ruling was meant to prevent companies from doing what Enron did—using off-balance-sheet entities to hide debt.

Because it applies to ABCP conduits, FASB's proposal threatened to curtail access to this inexpensive source of capital. But after intense analysis of the new rule, which took effect last July, banks have found it possible to cost-effectively restructure the conduits instead of consolidating the assets and liabilities. Most of the banks have just finished the job, which will be reflected in their third-quarter results. The bottom line for corporate participants: ABCP financing will remain available, but at a higher cost. It remains to be seen whether that extra cost will make the financing less attractive than the alternatives.

Perhaps it should be. Some critics contend that continued off-balance-sheet activity of this sort poses a risk to the health of the financial system (see "Conduits to Hell?" at the end of this article). Much, of course, depends on the course of the economy, simply because a recovery could mitigate much of the risk.

Financial Reengineering
It's clear that FIN 46 makes life more difficult for banks sponsoring conduits. Under the rule, the primary beneficiary of a traditional ABCP conduit is now required to consolidate the assets and liabilities on its balance sheet. The rule says that the beneficiary is the party that stands to absorb the majority of expected losses, to reap the majority of the operation's expected returns (portfolio gains or fees), or both.

The banks involved in this business have played both roles. Under FIN 46, it first appeared that the banks would have no choice but to bring conduit assets and liabilities onto their balance sheets. This was no small matter. Conduit assets often run into the tens of billions of dollars—about $20 billion for JP Morgan Chase, $39.5 billion for Bank One Corp., and $55 billion for Citigroup. Typically, conduit assets represent about 5 to 10 percent of total bank assets, says David Reavy, a partner with KPMG LLP's financial-services practice. The capital required as a reserve on these assets would equal 8 percent of the total, which would have led to a commensurate reduction in a bank's return on equity. Raising ABCP prices to bring ROE to previous levels would not have been realistic. "It would certainly depress the conduit market," says Mellon Financial Corp.'s Jonathan Widich, manager of its multiseller ABCP program. "It doesn't give an advantage to this type of financing over [traditional] corporate lending."

Then there was the potential impact on banks' net interest margin—the ratio of what is earned on assets to what is paid on liabilities. "Analysts and management look at this as a key measure of profitability," says Reavy. The high-quality assets in ABCP conduits are low-margin, lower than the lending business. Their impact on the balance sheet would be to depress the net interest margin. As a result of consolidating $10 billion in conduit assets and liabilities on July 1—leaving open the possibility of restructuring some to return to off-the-books status by September 30—Wachovia estimates that the full-year net interest margin would drop by about 5 basis points. Return on assets (net income to total assets) would also be squeezed.

What's more, banks have to consider possible future lines of business, and that would mean other demands on the balance sheet. "We might be fine today, but constrained in the future by having this on our balance sheet," says Brad Schwartz, managing director in charge of ABCP at JP Morgan Chase.

Estimating Losses
As a result, most banks have chosen to restructure their conduits instead. In the process, they've examined at least three possible means of doing so (see "One out of Three Ain't Bad," at the end of this article). But the banks have found only one, involving a so-called expected-loss tranche, that is truly viable. In this arrangement, the assets and liabilities of a conduit are reflected on the balance sheet of a third party that invests in a security whose returns exceed the expected losses on the underlying financing. By press time, HSBC, Citigroup, Mellon, Fleet, and Royal Bank of Canada had all completed restructuring at least some of their conduits using this method, while JP Morgan Chase and Bank of America were considering the possibility.


According to Mellon's Widich, the expected losses on its $1.2 billion in conduit assets were extremely difficult to calculate. "What's the expected loss when we've never come close to having a loss in 12 years?" he asks. "It's zero. Clearly zero wasn't going to cut it with the accountants."

And why should it, when past performance is no guarantee of future results? So Widich and his colleagues began studying FIN 46 and working the numbers. Looking at rating-agency data on the likelihood of downgrades and defaults, "we took a conservative approach and assumed anything that was downgraded defaulted," he says. Next, they figured in historic net losses in corporate defaults. Finally, they ran a Monte Carlo simulation of thousands of combinations of possible default frequency and net losses. At that point, says Widich, his team arrived at a number that satisfied Mellon's auditors that they had covered the expected loss. He won't disclose what that number is, except to say it's somewhere in single-digit basis points.

Next, Widich's team had to sell the subordinated note, which naturally meant paying a rate higher than that of the expected losses. He says demand was strong enough that Mellon was able to choose an investor that "knew the market and understood multiseller conduits and with whom we could partner."

Tallying the Cost
Despite Mellon's unwillingness to reveal the exact terms of its deal, it's easy enough to estimate the cost of a hypothetical transaction. If the rate of expected losses is 20 basis points and the return to investors in the note is 25 percent, the cost of participating in such a conduit would be an extra 5 basis points (25 percent of .0020). On a $100 million transaction, the additional cost of using such a conduit would be $50,000. Of course, that doesn't include the one-time costs of doing the actual restructuring, including lawyers and accounting fees.

Market participants differ on whether banks will decide to absorb these costs or pass them along to customers. (At press time, Mellon had not yet decided what to do.) Currently, the cost of participating in a traditional conduit ranges from 25 to 200 basis points, depending on a deal's liquidity and credit risk. A cost increase could hurt banks required to comply with U.S. accounting edicts, giving foreign banks a competitive advantage. Agreements for purchasing receivables for ABCP usually contain a provision that makes transference of cost increases automatic when the increase is due to an uncontrollable event. Some clauses specify that the pass-along is automatic when regulation increases regulatory capital, according to Sylvie Durham, chief investment officer at Abney & Holloway Asset Management in New York, which plans to invest in expected loss notes. But FASB is not a banking regulator, so simple consolidation may not trigger the clause, she says.

Still, most experts predict that the costs will be passed along. "It is a legitimate cost that should be borne by the users of the conduits," says the head of ABCP origination at one major investment bank.

Will customers accept such a cost increase? Bankers think so, if only because many of their ABCP clients are loan clients as well. "Marginally increased costs won't cause them to pull out," Durham says. At least one corporate finance executive agrees. Although "we never like costs to go up," says General Motors's assistant treasurer, Sanjiv Khattri, the company recently used a restructured conduit that is "slightly more expensive" to refinance the office space it leases in Detroit's Renaissance Center.

But some observers point out that returns on the subordinated notes, and thus costs of conduits restructured through their use, will have to exceed 25 percent if expected losses exceed 20 percent, which is likely for assets whose credit quality is less than stellar.

"I'd be willing to bet that those now paying 200 basis points to participate in a conduit will pay a lot more," says Jeff Wallace, a principal in Greenwich Treasury Advisors, in Greenwich, Connecticut.

The question of whether conduits will remain worthwhile may come down to whether they will continue to be less expensive than such alternatives as revolving credit. That depends on whether the banks also boost their pricing on revolvers and the like to maintain the current cost advantage of conduits. The answer to that question at this point is unclear.

Hilary Rosenberg is a kindergarten teacher and writes on finance from New York.

One out of Three Ain't Bad

Banks have examined three different means of restructuring their conduits:

The Silo Method. Bank of America reportedly used this technique to restructure about $6 billion of its roughly $22 billion in conduit assets. In this arrangement, companies that have sold assets into the pool consolidate the assets and liabilities on their own books, with each set of assets being a "silo." Such an arrangement isn't really asset-backed lending, but more of a traditionally collateralized loan. And the participants in the new entity already consolidated the assets and liabilities. This technique will not become more common unless other companies are willing to do the same, and so far there's little sign of that.

Joint Venture. A few banks reportedly are attempting to form a joint venture for purposes of restructuring their conduits, although they hadn't completed the transaction at press time. The concept: If three or more conduit sponsors combine their conduits, no one of them could be the primary beneficiary—and thus all would escape the confines of FIN 46. Each sponsor would provide liquidity and credit enhancement, and the commercial paper would be backed by all of the conduit assets. The trouble with this approach, besides its complexity, is that "you are exposed to the credit risk of a competitor's clients and the underwriting of a competitive sponsor," says Deborah Seife, managing director at Fitch Ratings.


Expected-Loss Tranche. By far the most popular method for restructuring ABCP conduits, this carries the greatest ongoing cost to corporate customers. The concept here is that a third-party investor in the conduit covers the majority of the expected losses (before the bank's credit enhancement), which by FIN 46's definition makes it, not the bank, the primary beneficiary. The bank has to estimate the expected losses from the pool, construct a subordinated note worth more than that amount (leaving room for the conduit to grow), and pay an interest rate high enough to sell that note to an investor willing to consolidate the conduit's assets and liabilities on its own balance sheet. Critically, the bank's credit enhancement does not make it the primary beneficiary under FIN 46. —H.R.

Conduits to Hell?

Critics contend that banks's increasing use of asset-backed commercial paper (ABCP) conduits threatens the health of the global financial system, because they make it harder to keep tabs on bad debt. Proponents suggest the opposite is true, because the off-balance-sheet financing technique serves to spread the risk of credit losses to outside investors. But the fact is, those investors understand credit risk less thoroughly than banks do. And a proposed change in bank regulations may inadvertently encourage riskier credits to be transferred.

Under rules that went into effect in 1988, the capital required to support a high-grade, AAA loan is the same as that required for a single-B loan. Because more-creditworthy loans carry lower interest rates, superior assets are more expensive for a bank to keep on its balance sheet than lower-quality assets, so better assets have more frequently been taken off bank balance sheets through ABCP conduits.

Indeed, these rules drove the banks into the business in the first place. Lenders quickly found it profitable to use high-quality corporate assets such as trade receivables to support the issuance of commercial paper by creating an entity outside the bank to contain the operation while collecting fees from corporate participants. The banks' only risk would be in providing credit enhancement and liquidity.

"It was regulatory arbitrage," says one investment banker who asked to remain anonymous. "On the balance sheet, capital would be required [to be set aside] on 100 percent of the assets. When assets were kept outside the bank, the bank was charged capital only as it interfaced with the conduit in providing credit enhancement. The point was they were saving regulatory capital."

Today, the ABCP market is a $734 billion business involving thousands of corporations and most of the major banks, according to David Zion, a banking analyst for Credit Suisse First Boston. Starting in 2007, however, banks will have to adjust their capital reserves to reflect the risk of losses on their loans, so lower-quality assets will require more support than higher-quality ones. That will create an incentive for banks to move poorer-quality assets as well as higher-grade ones off their books through ABCP conduits. And risk transfer of this kind has some regulators worried. "The transferors may have proper insight into the nature of the risks," Andrew Large, deputy governor of the Bank of England, was recently quoted in the Economist magazine (CFO's sister publication) as saying, but "are the transferees actually aware of the risks they have taken on?"

To be sure, new rules from the Financial Accounting Standards Board may inhibit banks' ability to put lower-rated assets into conduits, because the rules encourage banks to turn to third-party investors willing to put the assets on their own balance sheets. Experts say investors therefore could become warier of risky assets. Of course, that assumes that the investors have an adequate understanding of credit quality. But greater vigilance on their part would suggest that FASB has done more to protect the financial system than have banking regulators. —Ronald Fink




CFO Publishing Corporation 2009. All rights reserved.