Chipotle Arb Opportunity

There have been a couple of Seeking Alpha posts recently about the mispricing of Chipotle A shares relative to B shares.

The A shares closed today at $82.30. The B shares closed at $76.34.

From the Chipotle 10-K:

The restated certification of incorporation authorizes the issuance of an aggregate 230,000 shares of common stock consisting of 30,000 shares of class B common stock with a $0.01 par value and 200,000 shares of class A common stock with a $0.01 par value. Prior to Disposition, each share of class B common stock was convertible at the option of the shareholder into one share of class A common stock, and each share of class B common stock generally also converted into one share of class A common stock if a transfer of ownership occurred. Shares of class B common stock are no longer convertible beginning October 12, 2006. Shares of class B common stock participate equally in dividends with shares of class A common stock. Shares of class B and class A common stock generally vote as a single class of common stock Shares of class B common stock shares have ten votes per share whereas class A common stock shares have one vote per share, except that for purposes of approving a merger or consolidation, a sale of substantially all property or dissolution, each share of both class A and class B will have only one vote.

It seems clear that the B shares should trade equal to or higher than the A shares. However, the spread has persisted for quite a while.

New Short Possibilities: MBI and NCT

This is from Deep Survival by Laurence Gonzales:

The stages of getting lost apply to more than just hiking in the woods. A company, such as Xerox, ignores cues from a changing world and from inside its own research facility in Palo Alto and nearly destroys itself. In 1959, Xerox introduced its 914 copier. Fortune said it was "the most successful product ever marketed in America." By 1969, Xerox passed $1 billion in sales. In 1971, flushed with success (an emotional state of high arousal), the company's officers were in a state of deep denial. The world was changing and they weren't taking in any new information... Xerox's leaders had decided to take on IBM, despite all the clear evidence that it would most likely kill them to do so. They were like snowmobilers, flushed with emotion, who went up that hill, despite the clear evidence that it would probably kill them. They were bending the map, too...

Xerox spent $1 billion to purchase a computer company... In the meantime, scientists and engineers at PARC were inventing the mouse, Ethernet, the [GUI], the flat-panel display, and the laser printer. Others got rich off of those inventions. Xerox, busy with its mental models that did not match the real world, saw none of that profit.

I was surprised by the magnitude of this decrease in land prices:

The slowdown in housing construction has led to a sharp fall in residential land prices in the [Minneapolis-St. Paul] metropolitan area, according to a midyear market study by Bloomington-based United Properties.

Values for developable residential land in the metro area are down about 35 percent from 2005, United said... some residential developers were carrying 18- to 36-month supplies of land on their books, compared with a typical four- to six-month supply.

SPF's land positions are in markets that were far more overheated than Minneapolis-St.Paul. They have a high degree of leverage, and the recent revelation that their joint ventures are subject to margin calls and recourse to the company is stunning. Only a fool would buy their stock when you are going to be able to buy lots from them and other builders at a pittance.

Another name I am paying attention to is MBI. They are an insurance company that guarantees the performance of "municipal bonds, asset-backed and mortgage-backed securities, investor-owned utility bonds, bonds backed by publicly or privately funded public-purpose projects," etc. Mish has given some great MBI coverage on his blog. See also MBI on Seeking Alpha.

The degree of leverage in bond insurers like MBI is astonishing. From their capital base of approximately $7B, they guarantee the performance of over $600 billion in debt.

The great thing about MBI in terms of blow-up potential is that because of the high leverage, you don't need to be very specific about the possible cause. Will it be a few municipal defaults? That would be enough, and every housing collapse has caused some. Or will it be from the mortgage securities?

Take a look at Newcastle (NCT), a REIT of high-yield, asset-backed bonds. It's run by Fortress Investment Group, which gives me pause. You would think they would know what they are doing, but maybe FIG is using this as a vehicle to dump sludge-quality paper on retail investors.

That seems to be what was happening with Everquest Financial, the fund that Bear Stearns was going to IPO. They were essentially forced to cancel it once people recognized the moral hazard problem.

The stock has a 10.4% dividend, which has attracted quite a few yield-hungry investors and funds. It trades at 1.5x book, and it’s levered almost 10:1. Market cap is $1.45B.

My issue is with the credit quality: the securities have a weighted average rating of BBB+.; 41% is rated BELOW BBB- or is unrated, etc.

Worrisome parts of the portfolio:

• $77M in manufactured housing backed loans.
• $629M in home equity loans rated BBB+.
• $978M in unsecured REIT debt rated BBB-.
• $300M in BB rated commercial loans.
• $1B in subprime loans.

My final thought today is regarding the perception of a "subprime" menace.

It's wishful thinking to believe that the only problems are with low-FICO, "subprime" borrowers. Eventually, the market will realize that these problems extend to alt-a borrowers and beyond.

Corus Update

There was nothing languorous about the atmosphere of tropical Miami during that memorable summer and autumn of 1925. The whole city had become one frenzied real-estate exchange.

Yes, the public bought. By 1925 they were buying anything, anywhere, so long as it was in Florida. One had only to announce a new development, be it honest or fraudulent, be it on the Atlantic Ocean or deep in the wasteland of the interior, to set people scrambling for house lots.

A lot in the business center of Miami Beach had sold for $800 in the early days of the development and had resold for $150,000 in 1924. For a strip of land in Palm Beach a New York lawyer had been offered $240,000 some eight or ten years before the boom; in 1923 he finally accepted $800,000 for it...

It began obviously to collapse in the spring and summer of 1926...

By 1928 Henry S. Villard, writing in The Nation, thus described the approach to Miami by road: "Dead subdivisions line the highway, their pompous names half-obliterated on crumbling stucco gates. Lonely white-way lights stand guard over miles of cement side- walks, where grass and palmetto take the place of homes that were to be .... Whole sections of outlying subdivisions are composed of unoccupied houses, past which one speeds on broad thoroughfares as if traversing a city in the grip of death." In 1928 there were thirty-one bank failures in Florida; in 1929 there were fifty-seven...

By the middle of 1930, after the general business depression had set in, no less than twenty-six Florida cities had gone into default of principal or interest on their bonds...

The cheerful custom of incorporating real-estate developments as "cities" and financing the construction of all manner of improvements with "tax-free municipal bonds," as well as the custom on the part of development corporations of issuing real-estate bonds secured by new structures located in the boom territory, were showing weaknesses unimagined by the inspired dreamers of 1925.

That was from Only Yesterday: An Informal History of the 1920's, written by Frederick Lewis Allen in 1931.

Corus is the lender on 85 loans in the Southeastern U.S. for a total of 25,601 units. Of those deals, 68 are in Florida totaling 21,451 units.

Make sure you see these pictures of the Miami condo construction boom.

We had a debate going on in the comments section about CORS exposure to condos.

I prepared the following graphs regarding CORS deals in the SE region.

This is computed by calculating the average loan amount per unit for every CORS deal in the southwestern U.S. It would seem that the CORS loan size skews toward small amounts per unit.

The next graph illustrates that much of the skew comes from condo conversion loans. Excluding conversions, the average loan per unit is higher.




More bad news about the Miami condo glut:

"Boca Developers is considering a bulk sale of unsold units at three South Florida projects... New York-based Cabot Investment Properties has signed a letter of intent to pay $168 million for 236 units at Marina Grande in Riviera Beach, Peninsula II in Aventura and Oaks I at Biscayne Landing in North Miami.

Under the proposal, many of the units would be rented following the bulk sale with the idea of later trying to sell them when the residential market improves.

Cabot would purchase the units at a rate of $342 a square foot, which is much less than it would cost to build the same buildings at current construction prices, according to the documents outlining the proposal."

See the rest of this HBB post about Miami, which also says "Florida’s Miami Dade County has a 31-month supply of existing condos on the market."

Standard Pacific Update

I'm still short Standard Pacific (SPF), which gave an investor presentation yesterday.

"Net new home orders for [Apr & May 2007] were down 16% from the year earlier period and nearly 20% below the Company's business plan for the two-month period. The Company's cancellation rate for the 2007 two-month period was 28% compared to 35% in the year earlier period. The overall decrease in orders was driven by continued weakness in Florida and Arizona, while order activity was up over 13% year over year in California. The improvement in the California order comparisons was primarily a result of an increase in the number of active selling communities."

From the presentation:

• They have been somewhat successful at squeezing their contractors.
• 57,000 lots controlled; down 25% from Dec 2005. 60% of these are owned. They expect to sell about 8000 units this year, so they have a 7 year supply of lots. (That's assuming sales don't slow further - which they probably will.)
  
• They have only impaired one-third of owned lots.
• They expect to pay off the $350M revolver by the end of the year. The revolver expires May 2011. They renegotiated, trading a lower leverage covenant for a more flexible interest coverage covenant.
• They take impairments only once a community has a negative operating margin.
• They have 730 completed and unsold units, which is 3.17 per community.
• The JVs have an average 58% leverage.
• SPF is subject to LTV maintenance on the JVs. They are required to post additional equity as the value of the land/lots in the JV falls. Evidently the lenders on these JVs can slap SPF with a low appraisal and demand more equity. Their only "remargin call" so far was in SoCal during Q1.

More Bad News for Downey as Foreclosures Surge

One of my hedge fund manager friends sent me this:

Downey Seen As Mortgage Lender That Will Hold Up (Dow Jones Hedge Fund Trades, May 29)

“[Hedge funds] just feel that Downey is so conservative that they will not have the same problems that some of these other companies have had,” said Paul Miller, an analyst at investment bank Friedman Billings Ramsey & Co.

"Downey doesn’t issue the controversial “piggyback” loans and has trimmed its loan portfolio to cover areas with better housing markets."

The article goes on to say that certain hedge funds have been piling into Downey in recent months.

It's news to me that Downey was "so conservative" in its underwriting. Downey is the victim of adverse selection, as I explained in this anecdote. They were known among mortgage brokers for their "Downey Express" loan program that made loans with no verification of income or employment.

What kind of borrower would pay hundreds of bps extra interest to avoid having to document their income or employment? A borrower whose income is made up, of course.

I think these hedge funds just don't understand the role of the speculators, who were responsible for 20% of the house buying in SW markets and who are probably heavily represented (adverse selection) in Downey's loan portfolio. These are people who, prior to 2004 or so, were in multilevel marking schemes, or daytrading, or real estate or mortgage brokers.

As prices went up, these speculators piled in to real estate. They "bought" ten houses at a time - highly levered, of course. The result was a positive feedback loop of rising prices, with no basis in rising rents or buyer incomes.

That 20% doesn't count the straw buyer organized crime rings, or the grow-ops.

And even actual, legitimate home buyers bought houses far bigger and more expensive than they needed because "real estate only goes up."

Meanwhile, "U.S. foreclosure filings surged 90 percent in May from a year earlier... led by California, Florida and Ohio."

From the WSJ: "Small-bank investors who hope to snare big profits from a wave of takeovers in the industry may want to put their money elsewhere..." Says Yardville National Bancorp CEO: "The pendulum has swung back from a couple of years ago when it was a seller's market [for banks] to being a buyer's market at this time..."

Initiating Coverage on Corus (CORS): Part Two

This is the latest from Miami:

"...the most foreclosures at any condo development in [Miami-Dade county] this year."
"The 326-unit complex at 1331 Brickell Bay Drive had more foreclosures [17] than any other building in Miami-Dade and was second in South Florida..." That's 5.2% of the units, so far.
"Attempts to sell now are being compounded by falling appraisals and hesitant lenders. At least one has sworn off Jade already."
"Today, Miami-Dade has 28 months of existing inventory — without taking into account the new units that haven’t closed yet, said Ron Shuffield, president of [Esslinger Wooten Maxwell]." "In April 2005, there were 5,125 condos on the market. One year later, the number of available units spiked to 15,581, and the number rose to 22,924 in April, according to Multiple Listing Service research by EWM."
"Another problem for Jade owners trying to sell is that even willing buyers may have trouble finding a mortgage. Greenpointe Mortgage Funding stopped lending at Jade because of inflated appraisals and prior resale activity."

One of the themes of the nationwide condo bust is condo towers being converted to apartments, or condo conversions going back to apartments. From the Corus Q:
"An additional 3.8% of the loan portfolio ($147.5M) consists of one conversion loan and two construction loans that were previously classified as condominium loans. The borrowers ultimately failed to sell enough condominiums to make a condominium exit viable. For the conversion loan, the borrower has opted not to convert the property and will retain it as an apartment building. In the other two cases, the borrowers are negotiating the sale of the properties as apartment buildings."

Here's the two big pieces of bad news from the CORS Q:

• NPLs / Total Loans are 5.09% - up from 0.01% in 2006.
• "At this point, most of our problem loans are concentrated in the condominium conversion loan portfolio. We have a total of 48 condominium conversion loans totaling $1.0 billion, of which nine loans totaling $458 million are of particular concern."

Corus has an unusual loan officer compensation program:
"A significant portion of commercial loan officer compensation is ... the CLO Program... [which] generally holds back much of their commissions for up to nine years, during which time it is at risk of loss in the event the Company suffers a loss on the loans. Management believes the program motivates officers to make safe loans and aligns the officers’ goals with the Company’s interests."

This is admirable, but probably not enough because it gives the loan officers the same free option on the loan portfolio and they still don't have any skin in the game.

There's been an ongoing discussion in the comments section of the last CORS post that is worthy of publication.

Matt suggests that CORS will be able to take any foreclosed projects, sell them to REITs at a 5% cap rate, and break even.

There is a glut of condos for rent and for sale in Miami, for example. We could be very generous and assume that some REIT would be able to take a building and rent all the units for $2500/mo. We will assume that they can do this when 20,000 units are on the market simultaneously.

A cap rate of 5% would value one of these hypothetical units at (2500*12*60%profitmargin)/(.05) = $360,000.

Of course, if spreads widen to reasonable levels and condo rentals are yielding 8%, the value would be slashed by 38% to $225k.

If caps were at 8% and rents were only $1500 because of the inventory glut, the unit would be valued at $135,000.

Downey REO Watch II

This chart of the DSL Real Estate Owned pipeline is based on a sample of Alameda, Contra Costa, Kern, San Diego, San Joaquin, Solano, and Orange counties conducted on June 6, 2007.

Those 7 counties are only a fraction of the counties where DSL makes loans - DSL has at least one REO in over 40 different counties. However, the 7 counties do account for 40% of the REOs listed on DSL's website.

Not all counties make this data easily available online (especially in California). Los Angeles does not provide online access, and many of the counties that do have extremely cumbersome interfaces.

A key rule is that it takes at least 90 days to go from a default notice to a notice of sale.

So, the April spike in sale notices and trustee's deeds probably came from borrowers that received notices of default prior to 2007. Judging by the graph, that would suggest that virtually all of the defaults from Q4 2006 went to trustee's sales.

I encourage everyone to help monitor this pipeline. Go to the county recorder or county clerk websites (1,2,3,etc.) and search for "Downey," as they have a number of different entity names.

Downey REO Watch

In this April post about troubled DSL markets, I wrote, "Modesto, Stockton, Vallejo, etc. - that's Downey Financial."

Recently, a commenter asked, "How did you determine DSL is in these three markets specifically? None are mentioned as a top 5 market in the Q. Thanks."

It's true that the Q only tells us that Downey's loans are 89% in SoCal. So, first check the list of Downey branches. Then check the list of Downey foreclosures for sale.

Downey has two REOs in Modesto: 2512 Riverdale Avenue and 1209 Tenaya (wishing prices $240,000 and $210,000). It has one REO in Stockton: 3804 Wind Cave Circle ($479,000). It has two REOs in Vallejo: 130 Cottonwood Drive ($439,000) and 205 Haven Court ($357,000).

In Nassaw County, NY, Downey has this house listed as a REO, asking $950,000. The most recent sale was in December 2004 for $995,000. Prior to that sale, it had sold in January 2003 for $583,000.

In San Diego county, Downey added 4 REOs during Q1 2007. I predict they will add 14 more in Q2 since they have taken in 10 since Mar 31.

In Contra Costa county, Downey had no REOs until May 2007, when they took in two.

In San Joaquin county, Downey added 2 REOs during Q1. They have taken 4 so far in Q2.

In these three counties, Downey will go from 6 to 20 REOs in one quarter.

Initiating Coverage on Corus (CORS): Part One

Corus Bankshares Inc. (CORS) is the holding company for Corus Bank, N.A., the Chicago bank that is a pure play on condo loans in bubble markets. Corus’ lending focuses almost exclusively on condominium projects including both construction of new buildings and conversion of existing apartments.

As of March 31, 2007, their $3.85B loan portfolio was 64% condo construction ($2.46B) and 25% condo conversion ($0.98B) loans.

To fully understand the CORS story, you need to be familiar with the adverse cyclical events that are affecting condo towers and conversions and will be persisting for several years.

Just as with single family residential construction, an astonishing proportion of people signing condo purchase contracts were speculators. Because there is such a long lead time before the closing of condo purchases, speculators can make highly leveraged bets on real estate prices far in the future.

After the last real estate crash in the U.S., and after the 1990's crashes in Asian markets, economics and finance researchers began modelling these purchase contracts as call options. This is strictly accurate only if the developer has no specific performance remedy for buyers who back out of the purchase (e.g. when the market crashes).

However, that is a common circumstance: the contracts usually call for the deposit to be the liquidated damages, and specific performance is difficult even when your counterparty is not an overleveraged amateur speculator. (See also: As Condos Rise in South Florida, Nervous Investors Try to Flee)

Further, many of these contracts are contingent upon buyers being able to finance the purchases. But if there is a credit tightening associated with a real estate bust, then the speculators lose their financing and thereby have an assured "out" from any performance obligations.

(A related wrinkle occurs when buyers who want to close on their units can't, because the units no longer appraise for the sale price and the buyers can't get financing. This is an easy way for a tower project to fall into a death spiral.)

Anyway, under the "presale" model that developers use, the speculators get a call option (often underpriced), and the developer gets a unit "sold," which brings them closer to the threshold of sold units needed to close construction financing.

The 25% y/o/y real estate price increases combined with the high leverage of these implicit call options attracted legions of speculators. This led to increased development activity, and a positive-feedback loop was born.

Unfortunately, there is hardly any actual demand to live in these units at any price that will allow the developers - and possibly Corus - to be made whole.

Therefore, we find ourselves back in 1983, when the New York Times wrote "AUCTIONEER'S GAVEL FINALLY MOVES LUXURY CONDOMINIUMS IN MIAMI." That year, Miami condos sold at auction for "30 to 45 cents on the dollar."

As a percentage of funded balances of Corus' commercial real estate loans: Florida is 38% (Miami/SE FL is 22%), California is 16% and Las Vegas is 10%.

Which means that Corus' lending is concentrated in the most overbuilt bubble markets. In Miami-Dade County alone, 8,000 new condo units will be completed this year and nearly 12,000 more in 2008.

In Las Vegas, there are "4,214 existing [condo] units with another 13,409 under construction. Las Vegas also has a 16.5 month supply of residential inventory.

Jack McCabe talking about Miami condos: “When you drive by in the daytime, they are gorgeous, but when you drive by at night, there’s no furniture on the patios and only one light on out of 10.”

Stay tuned for the next post about Corus' portfolio, already showing signs of trouble.