Is Downey Savings & Loan a Zero?

DSL released some interesting news today and so far the stock is down about 7% while I'm writing this.  I wouldn't be surprised if a market is made for some of the larger institutional shareholders to "quietly" get out of DSL in the coming weeks but what I find striking about DSL's press release is that the credit loss provision still seems conservative.  Further, the $9MM valuation reduction for real estate held for development is a non-factor.  To me, this is just the first in many accounting adjustments DSL will probably make in the coming quarters.

Based on DSL's latest 10-Q, $65MM is held in investments in real estate and joint ventures.  The $9MM write down appears to be related to this line item based on the press release - "$9 million valuation reduction to real estate held for development to reflect declines in the value of single family home lots in which the company is a joint venture partner."  Assuming that this is where the write down will occur, this represents a loss of 14% in the value of its JV assets.  These are lots concentrated around the same geographic area where DSL's mortgages are underwritten which is why this $9MM valuation reduction is both meaningless in the here and now, but also could serve as a major precursor of what's to come.

What I mean by stating that the $9MM writedown is currently meaingless is that I expect the Street to say that on a material basis, the $9MM writedown really doesn't impact DSL from a long-term EPS standpoint (one-time item to the Street).  Further, the 1.22% loan loss allowance is already factored into DSL's share price based on Price to Book, and other static valuation standpoints.  As a result, the Street may continue to recommend shares in DSL and this 7% sell off may be a non-event in the coming weeks.

However, the "meat" on DSL's balance sheet is its Loans held for investment which total $11.7B based on its latest Thirteen Months Data filing.  The majority of these loans are residential mortgages in southern California and Arizona.  Now, why would DSL be inclined to reduce its JV assets by 14% but only set aside a loan loss allowance of just 1.22% of its loans held for investment?  On one hand, DSL is writing its own investments in a JV down by 14% but if one takes a commen sense approach to its Loans Held for Investments line item, it becomes apparent that a significant write down may be in the cards which could cause terminal damage to DSL.

Table I: DSL Pro Forma Assets (USD B)

Table I provides a quick and dirty look at DSL's asset structure, showing that its loans held for investment are essentially its asset base.  What's interesting is that DSL has written down its equity investments in real estate similar to the one its mortgages are back against by about 14%.  The JV is invested mainly in residential and some retail shopping centers but I suspect the $9MM writedown wholly relates to the residential part of the JV.  This could imply that the 14% writedown understates how low residential real estate is being valued by DSL because shopping center assets are in that line item.

This writedown could imply that the bank's credit condition is much worse than what is currently reported on a loan-to-value ("LTV") basis.  One of the attractive aspects of DSL (according to the Street) is its conservative LTV ratios, specifically its 110% cap rate on negative amortization loans.  This works fine in a period of static or rising home prices but is generally worthless in markets where prices are falling and DSL's writedown of its JV investments gives investors a way to benchmark surrounding real estate and apply that to the bank's loan book.

Consider a $500k home that secures a $390k mortgage, resulting in a 78% LTV.  Now, taking the 14% writedown to the $500k home results in a home value of $431k with the mortgage now representing 91% of the value.  Now factor in the 110% cap on that $390k loan and the loan expands to $429k leaving just $2k of equity.  DSL essentially owns the home and one can only speculate how challenging it is to recast a loan where the borrower has just a sliver of equity in place.  I believe this scenario could play out across a number of DSL's residential properties when one considers that DSL maintains about $7.6B in residential loans that have loan balances greater than the original loan, meaning borrowers have let the loans expand.

Further, DSL loans are relatively "fresh" and "unseasoned" with 11% of residential loans originated in 2007, 28% in 2006, and 33% in 2005.  The main focus in the news really pertains to 2003-2005 vintage subprime loans but I believe there's a much larger, more meaningful story in the prime option ARM area due to the newness of many of these loans and the fact that many are just starting to adjust.  What's more, given that DSL borrowers who are generally in the prime category have let their loans expand at lower rates, there's little to suggest that their payment habits will change as their payments increase (even against a lower index rate in some cases).  As a result, the LTV against DSL's mortgages and thus its overall credit condition will probably remained constrained for a long time.

Secondly, DSL really has no "out."  A lot of news is on moneycenter banks like Citigroup ("C") and Bank of America ("BAC") but in these cases, these banks have diversified, global operations.  The writedowns are painful but its likely that they will take some bruises but ultimately survive.  In DSL's case, 89% of its mortgages are in Los Angeles, San Diego, Riverside, Orange, and Riverside counties.  The bank is a thrift with no income stream other than its origination fees and interest income in a highly volatile market which makes the current valuation extremely expensive when one considers its likelihood for failure, however remote, is still much higher than larger banks with similar valuation metrics.

A 25% haircut in price from DSL's high of $77 or so is nowhere near enough to compensate for the problems brewing for the Company.  Investors should keep in mind that DSL's EPS through 9 months in 2007 is $1.87 based on the Company's press release.  I expect DSL will be lucky to generate $2.00 in FY 2007 EPS, let alone the possibility the bank recognizes a Q4 loss to further reduce FY 2007 EPS below $1.87.  This is in stark contrast to what the Street expects.  Based on Yahoo Finance, EPS estimates for 2007 are $4.81 and it was just in July when LEH analyst Bruce Harting suggested DSL could earn $6.00 in 2007.  Against those unrealistic estimates, DSL may look like its fairly priced at $55 which translates into a 9-11x P/E multiple.

However, against the likely, realistic EPS outcome of $1.50-$2.00 for 2007, DSL is valued at 27-30x 2007 FY EPS!  The problems DSL faces are just starting to surface and will take at least a year to resolve.  It took a long time for the problems to build within the bank and this won't be resolved until late 2008 at the earliest.  Investors at $55 are not getting a bank that will crank out $7.00+ in EPS in 2008, I expect that DSL's 2008 EPS will be similar to 2007, meaning even on an 18 month time horizon, the bank's valuation is extremely expensive.  Even going into 2009 and 2010, I believe it will take a long time to ever achieve the EPS levels DSL produced in recent years.

Investors need to keep in mind that loan volume and origination reached unprecedented levels in recent years and using even an average EPS figure across 2003-2006 may considerably overstate DSL's EPS power.  DSL also relied heavily on capitalized interest income from negative amortization ("CINA") to juice its EPS in recent years (link that covers how much of an impact CINA had on DSL EPS here http://seekingalpha.com/article/41906-downey-financial-deconstructing-book-value).  As the quality of those earnings have fallen and as investors and the Street catch on, the reality is that DSL will likely be producing much lower EPS in the coming years that what was seen from 2003-2006.

Book value is currently meaningless for DSL leaving just EPS as a viable valuation metric.  At $55, DSL appears to be radioactive.  If one assumes DSL's true EPS production will be in the $1.50-$2.00 range for 2007 and 2008, typical bank P/E multiples of 8-12x would suggest DSL's fair value should be $12-$30.  Keep in mind that's fair value, not a price that offers any real bargain.  What's truly amazing is that the stock is only down about 7% considering that 2007 EPS estimates are likely be 60+% below the Street's average.  I suspect once institutional investors actually try to understand what they own in DSL, the stock should receive the correction it's managed to escape.

DISCLOSURE: AUTHOR MANAGES A HEDGE FUND THAT OWNS PUTS ON DSL