Description

The subprime mortgage market has been very active in the investment community from participants on both the long and short side. But despite its popularity, it is not an industry that is very well understood. There is a lot of research out there on some of the major players, but this write-up aims to explain some of the nuances that seem to be missed, or at least misunderstood, by most of the analysts who cover the space. And it also aims to explain why a niche player called Delta Financial is one of the most misunderstood and undervalued stocks in the market today.

Despite being one of the best positioned and best managed companies in its industry, DFC currently trades at low-to-mid single digit multiples of its current earnings power even given recent tough industry conditions. This phenomenal valuation is due to a combination of factors including awkward accounting (highlighted by a general lack of investors’ understanding of the difference between real leverage and non-recourse leverage), recent industry conditions that have punished lower margin players, and conventional wisdom which suggests that the Subprime mortgage lending industry (like the housing market in general) is long overdue for disaster. We will address these three issues in detail below, but the bottom line is that we believe that this amazing valuation is real even in the event of an inevitable housing market slowdown or a moderate pickup in credit losses.

GAAP earnings are relatively meaningless, as detailed below, but our math says that they are earning over $1.50/share. With a stock price of $9.70, and excess capital of the loan portfolio of $3.60 - $9.50/share, you are paying
Price Earnings P/E
Stock Only $9.70 $1.57 6.2
Incl. excess capital (conservative) $6.10 $1.57 3.9
Incl. excess capital (fair value) $0.20 $1.57 0.1

In other words, when you net out the $9.50 of excess capital and loans (that are completely separate from the company’s ability to generate earnings and cash flow on future production) from the stock price, you are basically paying zero for a business that conservatively earns over $1.50/share at trough spreads and is growing volumes at 25% per year. Furthermore, these earnings are being achieved under trough pricing conditions that are driving weaker players out of business and causing larger players to take a lot of capacity out of the industry. And if you do not believe that current pricing of subprime mortgage loans accurately reflects the risks, then it is worth noting that even no matter how high losses eventually come in, it would not impair their current GAAP book value, which still means you have $3.60 of excess capital. Meanwhile, DFC is substantially insulated from many of the risks inherent to the subprime mortgage industry

Background

Delta Financial provides residential mortgage loans to non-prime (or “subprime”) customers. These customers typically have blemished credit and/or are unable or unwilling to provide the necessary documentation to obtain a traditional mortgage from a prime originator, and are thus charged higher rates and fees. Shortly after making the loans, DFC either sells them outright for cash (“whole-loan sales”) or raises secured financing for the loans while still retaining an economic interest. This latter method is done through the asset-backed securities (ABS) market and is the cause of much complexity and accounting confusion as we discuss below.

DFC is a small player in the subprime market, with less than 1% of the market. Competitors include both larger monoline Subprime mortgage originators such as New Century and Ameriquest (a private company, but you may have seen their Super Bowl ads), as well as more well-known mortgage finance companies like Countrywide, Washington Mutual, and GMAC. The industry, and in particular the monolines, has been growing rapidly for the past several years due in large part to advances in financing, improved risk underwriting, creative new mortgage products, and higher penetration of an underserved market. The subprime industry is driven largely by borrowers’ needs for cash, such as to pay off a credit card balance or to put towards a larger home purchase. As a result, particularly compared to prime loans, subprime loan demand is not as sensitive to interest rate changes.

The entire subprime mortgage industry stumbled in 2005 after several years of record profitability and growth for virtually all participants. A combination of several well-funded new competitors, an aggressive focus on gaining market share by the industry’s two largest players (among others), and—most importantly, in our minds—a rapid rise in funding costs led to a sharp contraction in the value of loans produced. This contraction occurred despite very strong underlying demand and credit fundamentals. The volume of subprime loans produced industrywide increased an estimated 26% in 2005, and the level of seriously delinquent subprime loans fell nearly 25%. The value realized by originators for each like-sized loan, however, fell precipitously.

Business Description

Part 1 – Underwriting Profitable Loans

Subprime mortgage originators make money when they originate residential mortgage loans that are worth more than the costs incurred in the loan origination process. The key levers of profitability, therefore, are (1) the total principal amount of loans originated, (2) the value of those loans (usually expressed as a percentage of principal, e.g. 103%), and (3) the costs incurred, directly and indirectly, while originating the loans and running the business (also typically expressed as a percentage of principal, e.g. 2.0%).

Why would the value of a loan be worth more than its principal (“par”) value? A $200k loan is worth more than just $200k if the expected net excess interest payments are greater than the expected credit loss (because borrowers can prepay the loans this is an estimate). Credit losses occur when a borrower defaults on a mortgage and the lender doesn’t receive the loan’s full value from selling the house. For example, if a borrower defaults on a $200k mortgage and the lender only nets $180k from selling the house (there are a lot of expenses associated with foreclosing and selling), the lender will lose 10% on the loan. If 10% of the borrowers default each year with a 10% loss per loan, the lender will lose 1% of the principal balance to credit losses. It is this propensity for higher losses that necessitates that subprime lenders charge higher interest rates than prime lenders; if they didn’t, then their $200k loans would be worth a lot less than $200k.

Getting back to a loan’s value, if a loan is expected to pay a 3% higher interest rate than the cost of financing that loan, and it is expected to last for two years before being refinanced, a total annual loss expectation of 1% would make the lender 4% in total profits over the two years. While we are ignoring the time value of money here, it should be clear that such a loan would be worth well in excess of the principal balance.

After origination, there are basically two ways that subprime lenders realize the value of the loans they make. First, they can sell the mortgage outright, either to a bank, Wall Street, or some other buyer in return for cash. This quick monetization is known as a “whole-loan sale.” Alternatively, the owner of a mortgage can retain an interest in the loan by financing the loan with asset-backed securities. Unlike with whole-loan sales, where the seller gets cash up front, it will take a few years to learn how much that retained interest in the loan will be worth (while the assumptions made earlier regarding losses and prepayments play out); that is one reason for messy accounting in the industry.

Part 2 – Differentiating Factors

The subprime mortgage origination business is actually much less subject to commoditization than one would think. Subprime lenders are service providers that among many other factors are differentiated by their marketing, reputation, speed, service quality, and their relationships. Those factors explain why the different lenders have vastly different levels of profitability. Indeed, DFC was making solid margins on their Q4 loans while most of the industry was losing money on their origination businesses.

Delta Financial is quite unusual among the subprime originators. For one, it is not based in Southern California, and in fact has very little exposure to the California market at all. This is despite the fact that the state makes up 30%+ of the value of the nation’s mortgages industry-wide. Second, DFC focuses almost exclusively on fixed rate mortgages, eschewing the hot and relatively new interest only, “pay option,” or hybrid ARM mortgages that have become very popular in recent years as a way for borrowers to lower monthly payments. A third differentiator is that Delta derives about 50% of its origination volumes from its retail channel, whereas most of its competitors are around 90% wholesale. Finally, DFC re-securitizes its securitized loans, selling “NIM” bonds collateralized by part of DFC’s equity in the securitization. This enables the company to rapidly grow its portfolio, but actually remain cash flow positive while it does so.

Part 3 – Securitizations and Asset-Backed Securities (ABS)

ABS are debt securities sold with either mortgages, auto loans, credit card loans, or other receivables as their only form of recourse. To isolate these mortgage or other loans for the ABS, a Special Purpose Vehicle (SPV) is created with its sole purpose being the acquisition and ownership of the loan collateral pool. The SPV pays for the loan pool by selling the ABS. Therefore, the SPV’s only real assets are the mortgage loans and the only liabilities are the ABS. To appeal to different investor appetites, the ABS are not created equal. Rather, they have different slices (called “tranches”) of priority and accordingly different credit ratings (the lower the bond’s priority, the higher the interest rate it pays). The excess value between the loan pool and the ABS goes to the equity owner of the SPV; this is why we said that ABS allows the mortgage’s owner to retain an interest in the loan (as opposed to whole loan sales). Putting it into our context, by selling ABS, DFC obtains financing for their loans while retaining ownership of the loan pool. In a simplified example, the originator/securitizer would finance $100 of loans by issuing $98 of ABS and contributing just $2 of equity financing.

Explanation of Accounting Issues

Unfortunately, when valuing or comparing the public subprime mortgage originators, traditional GAAP metrics are problematic and often very misleading. It will be useful, therefore to take a step back and look first at economic reality, and only then look at how that reality is represented (or misrepresented) by the individual company’s financial statements.

As we discussed earlier, mortgage originators make money by originating a loan of a given size (say $100) at a given cost (say $2) if the loan they originated is worth more than that cost. If the market value of a newly originated loan with those characteristics is $103, then the economic reality is that the originator created $1 of profit. The originator’s later decision to either sell the loan for cash or to finance it with ABS is a capital allocation decision with relatively little impact on the value creation process. On the one hand, because the originator can choose to sell that loan in the market for $103, management must believe that incremental value is created by securitizing and retaining that loan. On the other hand, the price that whole loan buyers are willing to pay is very much dependent on the expected value THEY could obtain through securitizing it themselves. It is therefore a safe generalization to think of the value of the loan being essentially the same whether it is sold for cash or securitized by the originator.

Let us now leave economic reality behind and talk about how to compare two mortgage originators’ GAAP financial statements. The main issue that makes valuation difficult is the timing difference between value creation and value recognition under GAAP. Value creation occurs when a loan is originated and then either sold or securitized (although as previously mentioned, in the case of securitizations the exact value created is unknown for years). Value recognition as reported earnings, on the other hand, depends on many factors including whether the loans are sold for cash (resulting in a cash gain on sale), securitized using sale accounting (resulting in a non-cash gain on sale and the creation of a residual security on the balance sheet), or securitized using portfolio accounting (where both the mortgage loans [assets] and the non-recourse securitization bonds [liabilities] are kept on the balance sheet, resulting in the deferral of gains and the recognition of spread income over the life of the loans).

In fact, it can get even more complicated than that, as it does with Delta Financial. DFC securitizes most of the loans that it originates using portfolio accounting, but then issues non-recourse “NIM” bonds secured by their residual interest in the underlying securitization. The NIM bonds have no recourse to DFC in the event of higher than expected defaults (remember, the ABS only has claim on the securitized pool of loans). This lack of recourse means that DFC essentially sells (for cash) all but a very small residual interest in a pool of mortgage loans while retaining no corporate liability in the event that those loans default. Despite this, they do not report any of the cash they receive as upfront earnings. For example, suppose that DFC originates $100 of loans at a cost of $2 (for a total cash outlay of $102). It would then securitize those loans and receive $98 from bondholders, keeping all value over that as an equity interest. Suppose now that DFC could then sell $5 of NIM bonds backed by its residual interest in the securitization. All in, DFC would have received $1 of cash and would have no liability no matter how poorly the mortgage loans perform. The company’s income statement, however, would show zero earnings (and in practice would show a loss because many origination costs are expensed and not capitalized). Amazingly, the balance sheet would be an even bigger mess. If we assume for simplicity that all of the origination costs were capitalized, DFC’s balance sheet would show $102 of mortgage loans at cost, $1 of cash, $103 of debt, and no retained earnings. Leverage would appear to be infinite ($103/$0), despite the fact that none of the debt on the balance sheet would have any claim on the company’s cash! Moreover, the value that DFC created by originating and securitizing the loans would not show up as book equity (again, despite the fact that the $1 of cash is the company’s free and clear).

What we are talking about is not just theory; DFC’s balance sheet completely distorts both the company’s true equity and leverage. Take the 3/31/06 balance sheet, for example. It shows $5.1 BN of mortgage loans held for investment and $5.1 BN of debt versus only $117mm of shareholders equity (recall that by issuing NIM bonds, DFC is able to issue non-recourse debt in excess of the principal balance of the mortgages it securitizes). If this were normal, recourse debt, that would imply that a mere 2.3% reduction in value of the mortgage assets would entirely wipe out DFC’s equity. To put that into perspective, subprime mortgage loans are an asset class where 1.0-1.5% annual loan loses is considered normal!

But of course that isn’t accurate. DFC has already received considerable cash upfront, which the ABS holders can’t get back no matter how badly the loans perform. They are limited in their recourse to just the loans in the trust. But among other things what this means is that, since loans on DFC’s balance sheet are held at a value below the amount of debt securitized against them, if basically everybody stopped paying then DFC’s book value would not be impaired.

Another nuance that understates DFC’s equity stems from the composition of the cost to originate. Earlier we simplified things by saying that a company’s cost to originate was 200 bps, 100 bps of which was capitalized and 100 bps of which was expensed as incurred. But DFC defers fewer expenses than most of its competitors. This means that their balance sheet is stronger than it appears, and its exposure to some sort of disaster in the residential mortgage sector is lower than for some of their competitors.

Valuations

Due to GAAP’s inability to measure the true economics of these businesses, we value the companies by going back to the basics of their business; how many loans they issue, how much it costs to originate their loans and how much those loans are worth. We then add the value of the balance sheet’s securitized loan portfolios, as it has value independent of the mortgage origination platform.

We take just this approach when looking at DFC. First, we isolate the mortgage banking operations from the portfolio operations, and then we measure the earnings potential of those units separately. In other words, what would the company’s mortgage banking operations earn if they were to sell all of their production for cash? The formula is fairly simple: (origination volume) x (gain on sale minus cost to originate) plus warehouse spread profit (interest earned on the positive carry of holding the loans from origination until sale). For DFC, we estimate that its 2006 banking earnings, if the company decided to sell all of its loans, would be $1.57 per share [$4.4BN x (3.1%-2.0%) + 14.0mm] taxed at 40% and divided by 23.9mm shares. This represents just 6.2x the recent trading price of $9.70 for DFC. Also note that we used a 1.1% gain on sale margin, which is much lower than the 1.4% they reported in the 4th quarter (and first quarter, adjusted for seasonality).

Of course, the company’s portfolio has substantial real value as well. To value the portfolio, we look at the fair value table at the back of their 10K. Congratulations if you actually made it all the way to the back of their 10K. But basically it says that if you marked their portfolio to market at year end it would have added $141M to the book value. If anything, this number would have grown in the first quarter, but they don’t update that table in the Q’s, so we will conservatively use $141M. So if you take the book equity of $117M plus the $19.3M they raised in April plus the $141M adjustment you get an equity value of $277M. From this we subtract the amount of capital that is required to run the mortgage banking business (i.e. the amount of capital that is used to generate the $1.57 of earnings). We think it is conservative to assume that the mortgage bank uses $50M. So ($277-$50)/23.9M shares is $9.50/share.

When you add up the pieces, you can see that YOU ARE GETTING THE MORTGAGE BANK FOR FREE. That’s right, when you net out the $9.50 of excess capital and loans (that are completely separate from the company’s ability to generate earnings and cash flow on future production) from the stock price, you are basically paying zero for a business that conservatively earns over $1.50/share at trough spreads and is growing volumes at 15% per year even in a declining market. Even if you don’t believe the MTM fair value of the portfolio (it does admittedly depend on a variety of management estimates including credit), if you strip out the $141M you still have excess capital of $86M or $3.60/share, which means you are paying less than 4X earnings. And as I explained above that $3.60 would still be solid no matter what happens to credit on their existing portfolio.

They are also rapidly improving their origination costs, both by spreading fixed costs over more originations and by improving process efficiencies through automation and other initiatives. Therefore this ridiculous valuation should get even better.

Catalyst

Catalysts:
- Rationalization of pricing due to presence of activist shareholder on the board of industry’s largest public competitor
- Rationalization of pricing due to weaker players folding
- Rationalization of pricing due to larger players raising prices and taking out capacity
- Continued growth in earnings power as origination platform grows
- Posting reasonably strong credit statistics even as rates rise and housing market slows
- GAAP earnings will begin to catch up with real earnings over time, beginning late this year, as DFC started portfolio accounting at the beginning of 2004 and these are generally 3 year assets