Will the Housing Slump Raise Consumer Defaults?

"This is one of the greatest financial panics I've seen in fifty-five years in financial services... All of the money is going into T-bills at the moment."

(Angelo Mozillo, CEO, Countrywide Financial Corp.)

Must viewing for risk managers is Friday's CNBC interview of Countrywide Financial (NYSE:CFC) CEO Angelo Mozillo -- if you can get the CNBC web site to work.

Mozillo describes out how the investor run from commercial paper and other types of collateralized assets started and why the flight to quality, away from all types of short-term consumer and commercial assets, is threatening the US economy with illiquidity and recession.

In the interview, Mozillo notes that most of CFC's business has been done "outside of the bank" over the past few years, but that he and his management team are in the process of restructuring the business so all of the loan originations occur "in the bank." His comments about "out" vs "in" the bank are a telling comment on the state of the money markets and the ability of smaller, non-bank financial institutions to survive the current crisis.

The reason for CFC's corporate makeover is very simple: counterparties want the federally insured bank unit of CFC as the obligor in any transactions, not the publicly-listed parent holding company. As of Q2 2007, Countrywide Bank FSB had $99 billion in assets while the parent's consolidated assets were 2x that amount, including the loan production and conduit operation. Look for all of those business units to be "in the bank," to quote Mozillo, before year-end.

We continue to believe that the doomsday scenarios regarding CFC described by certain Sell Side analysts are a remote possibility, to put it mildly. But it is technically possible that were the parent company of a federally insured depository institution to sustain significant losses vs the 13% total capital of the consolidated group, then the FDIC could be compelled invoke its legal powers as receiver to conserve the assets of Countrywide Bank FSB, leaving the crippled parent to face bankruptcy and liquidation.

We tend to think that Bank of America (NYSE:BAC) would pull the trigger long before CFC's capital became impaired, but the scenario above is the technical possibility driving market hysteria regarding CFC.

Hopefully market participants will calm down this week and realize that credits like CFC remain sound, even if subprime assets are not, and that most asset-backed CP programs are likewise not at risk. As Mozillo told CNBC: "There is a serious problem out there in housing," but that does not necessarily mean that all commercial paper issued by any US company or bank is suspect. At least not yet...

Of interest, last week subscribers to the IRA Bank Monitor saw that the loan portfolio default rate of Countrywide Bank FSB jumped dramatically from the 15bp reported in Q1 and is now above the 21bp peer average for Q2. That said, 21bp of default is 0.21% of total loans -- hardly a crisis.

When defaults hit 100bp or 1% of total loans, then we'll have reached the 2001 peak. Still no biggie. Question is, where do we go in 2008-2009? Looking back to 1990, the peak for average defaults among the mortgage specialization peers of CFC was a whopping 1.3%.

According to our forward-looking estimates, default rates for the mortgage specialization peer group could hit ealy 1990s levels as soon as the end of 2007 -- that's how quickly the mortgage bubble is deflating. Imagine how such credit default experience will affect the happy world of "mark to model," not to mention the impending adoption of Basel II, both in the US and globally.

Click here to read our comment on Basel II in last Friday's edition of The American Banker.

Is Consumer Credit Next?

As the kettle steeps in the mortgage markets we turn our attention to risk management in the unsecured lending side of the world. Like most people, we make most of our monthly payments at or around the due date, using the handy facility of online bill pay. Our bank sends out the money on the designated day at midnight and goes into the books for the payee bank the next midnight. A basic enough cash management plan.

Watching our card issuers, we've discovered some interesting new loss management policies coming from the credit card industry. It seems that credit card issuers have drastically tightened rules about the timing of consumer payments. At midnight of the day the payment is due, if it has not been logged at the payee institution, then the bank shuts down the card until the payment arrives.

The decision to freeze the account -- a crude way of managing Exposure at Default or EAD -- turns out to be a 24-hour hole because of the way two bank's computer systems batch process systems. Naturally we've adjusted our cash management to allow for 24 hours of processing time at the payee institution. However this discovery leads us to ask: How is the unsecured consumer credit world doing these days? Why are credit cards (and even home equity lines) that get paid off every month and carry no more than 10% of their credit limit receiving Draconian loss controls applied in 30-day increments?

Using the data mining capabilities of IRA's bank analytics systems, we looked at the history of credit card loan principal outstanding over the last few years. The numbers are, relatively speaking, in about the same range since around the end of 2003, this after climbing steadily to that level after the end of 1989.

Even as consumers were inundated with offers to switch balances around from one card issuer to another, the size of the business, measured by the balance sheets of the credit card specialization peers, seemingly was a zero sum shell game, moving the same loan principal around -- like cell phone customers migrating to different service providers. Keep in mind that these static, end of quarter balance sheet totals do not show how the securitization volumes in the credit card business peaked in the 2005 period, but have since declined sharply.

Somewhat more interesting was our data mining of annualized projections of net defaults for credit card loans. Here we began to notice that the credit card industry seems to have adjusted policies in 2006, even as net defaults declined to five year lows. For example, BAC's credit card unit, FIA Card Services, the $143 billion total asset bank formerly known as MBNA, reported just 43bp of defaults in Q2 2006, but has since see run rate loan defaults jump to over 500bp in Q1 and Q2 2007.

That's right, charge offs by FIA totaled 10% of total loans in the first six months of 2007. Now you understand why credit card specialization peers have asset and equity returns that are 4x that of a mortgage lender like CFC and capital-to-assets levels in the high teens.

Indeed, as shown below, the 470bp (4.7%) rate for aggregate 1Q 2007 defaults for the credit card specialization peers indicate that the annualized default projection numbers are once again moving back to pre-2006 levels, possibly worse. Certainly just cause for unsecured credit issuers to possibly act to protect themselves by reigning in EAD, raising rates and tightening other payment terms.

EAD is one of our favorite Based II Pillar measurements and, for credit card issuers, a key business model indicator. This number is a measurement of how much additional credit an obligor can draw upon just before defaulting on a debt -- and how a bank's managers view forward probability of default or "P(D)" for their customer base. The number to watch in this case is the Unused Commitments for Credit Cards. That's the unused line of credit that could be tapped by a consumer before going belly up.

In 1Q 2007, that figure was $4,320 billion for all credit card issuers; an EAD ratio of 12.2 to 1 over the loan principal already outstanding in these credit lines. That may seem like a pretty large EAD, but it has become SOP in the credit card industry over the past decade. For the subset of credit card issuers whose portfolios are in excess of 50% unsecured credit card debt, the EAD ratio was an even higher 14.3 to 1 in Q1 2007, this based on a $186 billion share of the principal loans out standing and $2,668 billion share of the unused commitments.

By comparison, the EAD for Countrywide Bank FSB was just 0.19:1 in Q2 2007, half a standard deviation below the 0.35:1 peer average.

The unsecured consumer credit industry long has relied on the principle of large number theory -- that most consumers were good payers of revolving credit -- as the basis for safety and soundness against the minority of defaulting debtors. But the shift from generous lending practices to strict loss controls indicates that the population behavior statistics have changed dramatically.

So if you are a rational credit card issuer looking at your unsecured line holders as also being holders of marginal mortgages, you'd probably be installing short-fuse stop loss policies to keep people from accessing their full lines as part of a worst-case default scenario. That is, when dislocated consumers start tapping unused credit card lines to pay the mortgage for another month -- like the 3,000 people who've been laid off in the past two weeks from the mortgage units of Lehman Brothers (NYSE:LEH) and Bear, Stearns (NYSE:BSC).

What does that last sentence imply? Well if consumers are cut off from their primary source of liquidity, namely the home equity line, and also face a tightening web of loss-management measures on unsecured facilities, do these two clear data points not call into question the assumption that all is well in the US economy? Is the US economy really insulated from the secondary squeezing effects generated by the mortgage market downshift?

If Mr. Mozillo is right about the US real estate sector being in its worst ever crisis for the past half century, then the consumer credit industry may be next in line for a revision regarding forward P(D) estimates and policy regarding EAD. In deflating the US housing bubble, we may be toppling a lot more dominoes than is currently recognized by investors or policy makers.