Wednesday, August 22, 2007

This is our most desperate hour...

Many mortgage originators have declared bankruptcy in the last 6 months. The market is now placing a fair probability that Countrywide may be forced into liquidation in the coming months. A couple days ago, Countrywide bonds maturing in December 2007 could be bought with a 20% yield. Now they are a little better, around 13%, but obviously the market is concerned.

Corporate bond guys use the term "jump to default" to describe what might happen to Countrywide. On August 1, Countrywide's credit rating was a strong A-/A3, how could the company be so close to default so quickly? There is, not surprisingly, a lot of confusion in the main-stream media about Countrywide and their troubles. Below is a simple model for how specialty finance companies (like Countrywide) operate. I don't have intricate details about how Countrywide specifically finances their business, but this example should give the reader a fairly accurate idea of how things work.

Countrywide takes most of the mortgages it underwrites and sells them to investors in the secondary market. But they can't sell anything to the public until they have actual loans actually closed. Of course, at closing they have to actually write a check to the borrower, so somehow they have to come up with some cash.

One way firms like Countrywide can secure attractive financing, at least until recently, was by pledging some of their investment portfolio. This could be to back a credit line at a bank or to back a commercial paper program. Historically banks have always liked having specific assets pledged to secure a loan, because that prevents the borrower from taking on more loans and diluting the bank's recovery in liquidation or similarly selling off assets which could prove valuable in liquidation.

Let's assume that Countrywide wanted to make $30 billion in loans (which was their July production.) In order to raise that cash, let's say they pledge $30 billion of their investment portfolio (mostly mortgages) to an Asset-Backed CP program. Let's say that the term of the CP is 30 days. Then when the loans close, they sell those loans into the secondary market for $30 billion, and repay the CP. Their assets are then free and clear to do the whole thing over again next month.

But what happens if they can't sell the loans into the secondary market? The $30 billion they lent to home owners goes out the door, but they still have to pay off the CP. They might like to just roll over the CP, but now the value of their assets is in question. If the lenders can't value the assets being pledged, they are unlikely to make the loan.

Because of the long lead time in the mortgage business (you commit to a loan as much as 90 days ahead of actually paying out the cash), if Countrywide has $30 billion in CP outstanding, that's probably already being used to pay for loans originated in May and June. The $30 billion they committed to in July is yet to be funded. By that calculation, they would have something like $60 billion in cash needs. This is the figure estimated by Kenneth Bruce of Merrill Lynch in his August 15 report on CFC. Suddenly Countrywide is in desperate need of $60 billion but no one wants to lend them money. Maybe the only route they have left is to sell their investment portfolio. But even if the portfolio is relatively high-quality, the odds are good that any non-prime MBS will have at least a 5-10% discount from their par value. At least. And their liquid investment portfolio is only about $50 billion, again according to Merrill Lynch. So there is probably a $10 billion shortfall even before you assume any discount on their liquid assets.

So notice that the problem has nothing to do with suffering losses on bad loans, or insane leverage, or any of the other ills that have plagued the mortgage industry. Its simply that they cannot sell mortgages.

Now, Countrywide may be able to sell some of their mortgages, and they have been able to tap existing bank credit lines to avoid an imminent cash crunch. They could choose to drastically reduce production and slowly sell off assets, but if they get rid of all their assets, what's left for lenders? If they go down the asset sale road, they may have trouble getting their credit lines re-approved even if the market for mortgage loans recovers.

So while Countrywide's core business is weak, it's the sort of weakness most companies could weather. We know that mortgage underwriting is a long-term viable business. But because Countrywide faces this constant need to refinance itself, the liquidity crunch could cause a jump to default.

Enter Warren Buffett. Consider the scenario I just described: long-term Countrywide is a powerful player in a long-term viable business who is getting squeezed due to short-term problems. Someone who is looking to buy a nice asset on the cheap and who has adequate access to capital to ride out the liquidity crunch. That means that Countrywide, at a certain price, is a
very attractive takeover candidate. Hence the Buffett rumors. His history of getting involved (or trying to) in Salomon Brothers and Long-Term Capital Management has a lot of parallels. If he is willing to suffer a little short-term pain, he could own a dominant player in the mortgage business when things finally improve.

Help us, Oracle of Omaha... You're our only hope.

UPDATE: No... there is another... Bank of America Invests $2 billion in Countrywide


Anonymous said...

Most impressive! does it take long to search for a good phrase from SW to contextualize your postings? :-)

Scurvon said...

Great read - thanks.

So what do you make of $2 bln of preferred sold on what appear to be very generous terms? Do they need $2 bln that badly?

Anonymous said...


Thanks for a great post!

The $2b presumably goes into the bank, which can then buy $10b in mortgages with the capital. Do you agree?

That certainly helps. However, if the CP cannot be rolled, its not enough (given your numbers). So the company's future STILL hinges on mortgage sales. If it can't unload them, then back to BAC they go for more liquidity. Only what price would the second convert be done at?

Credit cycles cause lenders to do multiple converts or rights issues at successively more dilutive prices. Examples: Citi in '91, Mexican banks in '95, J. Banks in late nineties.

Why should this one be different?

Accrued Interest said...


Sadly, it doesn't take me long. I'm that big a geek. I started doing it a while back because I thought it was funny. The nice thing about blogging is you really have the freedom to write for yourself. If I think its funny, I'm damn well doing it.

Anyway, I'm just as big a geek on the bond market, which is much more lucrative that SW.

Scurvon & David:

Remember that BAC was widely rumored to be acquring CFC about 6 months ago. So now...

1) BAC does not benefit by CFC going under. If the whole mortgage market is damaged, that wouldn't bode well for BAC's business either. Everyone in banking relies on securitization (for MBS, CLO, ABS etc), so no one wants to see that market permanently damaged.

2) So BAC injects a relatively small amount of cash into CFC, but at the same time tells the market that BAC has a vested interest in CFC's survival. Other financial institutions will be more willing to lend to CFC given BAC's involvement.

3) BAC can now buy a very large percentage of CFC at a steep discount. I'm thinking of it as a quasi-option to acquire CFC at some future point.

Unknown said...

surprised to NOT see any mention of Buffets recent INcreased stake in bofa as a quasi buffet involvement in CFC ... whatever and whoever the case may be, for NOW the winner appears to be kenny lewis of bofa, but really, it all strikes ME like getting a MasterCard bill in the mail and sending along a 'NO-interest' check from Visa to pay the former bill ... the bad 'stuff' still exists on someone's balance sheet and oh yea, when bank earnings come 'round again (september?) one would imagine the mark-to-market dance will once again get interesting ...

fbeckenbauer said...

Interesting post.

But let's compare Buffett's stated acquisition criteria:
1. Large Purchases -above 75M pretax profits- OK maybe
2. Demonstrated consistent earning power (no turnarounds) - NO
3. High ROE with little or no debt - NO
4. Management in place - NO (Mozilla doesn't strike me as Buffett's kind of manager)
5. Simple businesses - NO
6. An offering price - OK maybe

So I think the Buffett rumours are ridiculous. He has stated these criteria every year for aeons and CFC does not qualify.

Accrued Interest said...


You seem to follow Buffett more than I do, but I wonder how his offer to buy LTCM fits your model?

Unknown said...

fbeckenbauer: not for nothing, i did say quasi involvement, didn't i? can you say that buffet's decision to increase stake in BofA fits? suppose what i'm really thinking kinda scares me most: do NOT agree even with buffet's INcreased stake of BofA as i can't help but think banks too have some 'risk' and i guess we'll find out IF ... when banks issue earnings and are forced to mark some stuff to market ... i do really hope i'm wrong, cause that simply bodes poorly for stocks and 'things' generally speaking ... on the other hand, being involved in the bond mkt, hopefully biz will pick up! (thats right -- when life gives ME lemons, i'm gonna try and mix up some lemonade!!)

fbeckenbauer said...

Well its not really my model. The criteria are lifted from those he puts in the BRK AR every year.

Those are his criteria for acquisitions - ie companies which he buys outright, and I was assuming from your post that you were considering this possibility. For the reasons stated above I think this is not going to happen.

But his criteria for trading securities is very different - as you mention he did offer a price for the LTCM positions (NOT for LTCM -he explicitly told them he did not want the management etc)and he has bought junk bonds, options etc.
So it is not completly out of the question that he might buy a stake (up to 15%) in CFC or some other lender.

Accrued Interest said...


Yeah. Good point on not wanting the management. I think BAC's involvement with CFC takes WB out of the equation honestly.

fbeckenbauer said...

By the way I think this is an excellent blog.
I'm hoping that someday you might post a reading list with good books about bonds/bond market

Accrued Interest said...

If anyone is young and thinking of getting into the bond business, either on the sell or buy side, first read Liar's Poker and watch Wall Street, Trading Places, and Office Space. I'd say at least 1-2 times a week one of these movies/books is referenced on every trading floor in America. If you don't know what a Big Swinging Dick is, you'll just look like an idiot.

I will sometime make up a list of the actual books I reference a lot in my day-to-day work.

fbeckenbauer said...


Anonymous said...

How could you leave Boiler Room off that list