Radian Group, Inc. (Report #215231001)

Radian Group, Inc. (Ticker: RDN, last price $10.62 as of 12/14/07 close)

Author: Anonymous

Position: Long

Submitted: December 14, 2007 (after market)

Radian is a specialty insurance company with two major business lines: traditional mortgage insurance and bond insurance. Radian provides both direct insurance and reinsurance in these areas.

Radian’s stock has been hit extremely hard during 2007, down 84% from a high of $67.35 on February 6. There are two major reasons for this. First, Radian had agreed to merge with another mortgage insurer, MGIC, but turmoil in the mortgage market caused the merger to fall apart. Second, and more importantly, Radian’s two lines of business are both directly involved in the current sub-prime mess.

As of 3Q 2007, Radian had a book value of $3.4 billion, or $42.86 per share. At $11.05, the stock is trading at a 74% discount to book. While Radian obviously faces some serious challenges in coming quarters, this giant discount to book value is unwarranted.

There are two major reasons why a company would trade at a large discount to book that apply to Radian. First is that the market believes that the book value isn’t correctly estimated. For a financial company, this would typically be because there are impaired assets on the books that haven’t been written down, or at least not written down enough. Second is that the market believes shareholders are likely to be diluted by a need to raise new capital. We have seen several other mortgage-related financials raise large amounts of new capital recently in response to either regulatory or ratings agency pressure.

Both the expectation of future losses as well as the shareholder dilution situation apply to Radian, or at least could potentially apply to Radian. However, I believe an examination of potential losses indicates that losses are unlikely to be large enough to justify the current share price. In addition, Radian is unlikely to need to raise new capital, particularly common equity.

According to the company, Radian’s insurance portfolio as of 9/30/07 breaks down as follows (in either billions of dollars or percentage of total par insured).

Direct/Indirect Public Finance: $59.8, 38.4%
Direct/Indirect Structured Finance: $53.0, 34.0%
Total Financial Guarantee: $112.8, 72.4%

Prime Mortgage Insurance (MI): $20.1, 13.4%
Sub-Prime MI: $8.5, 5.5%
Pool MI: $3.0, 1.9%
Second Liens: $1.0, 0.6%
NIMs: $0.7, 0.5%
International MI: $8.6, 5.5%
Other: $0.2, 0.1%
Total Mortgage Insurance: $42.9, 27.6%

Examining the likelihood of loses in this environment is extremely difficult. On one hand, the economy is experiencing the after effects of a massive overindulgence in mortgage lending. Many analysts, myself included, believe there will be significant price erosion in the housing market in coming quarters.

On the other hand, efforts by the Federal Reserve and other government officials will undoubtedly alter the foreclosure rate by some degree. At worst, programs like the Hope Now Alliance will at least delay foreclosure, which should spread Radian’s MI losses out over time, which will ultimately be less taxing on their capital.

Despite the difficult environment, there is no particular reason to believe that Public Finance losses will be larger than usual for Radian. While Radian tends to insure BBB/Baa-rated municipal issuers, municipal default studies conducted by both Moody’s and S&P show no cyclicality to municipal defaults. These results were consistent even when looking solely at higher-risk municipal issuers, like private education or health care facilities.

In Structured Finance, Radian has limited exposure to residential mortgage-backed securities (RMBS), with only about 1.5% of total insured amount outstanding. Radian stopped writing insurance on RMBS in 2005, and on sub-prime RMBS in 2004, although they continued to take reinsurance contracts on such pools.

Radian’s primary exposure in Structured Finance is in corporate collateralized debt obligations (CDOs). These CDOs make up 30% of total insured. Corporate CDOs are pools of investment-grade credit-default swaps (CDS). Radian’s exposure is at the top of the CDO structure (rated AAA), which means that Radian typically has a substantial amount of notes subordinate to what’s being insured. According to Radian, these CDOs can suffer defaults of between 10% and 15% before Radian would pay any amounts under the insurance contract. According to Moody’s, the worst single-year default rate on corporate bonds from 1983-2007 was 3.8% in 2001, and the worst 5-year period was 13% for the period ending in 2003. Although no information was readily available about Radian’s CDOs, most corporate CDOs have a 5-year life.

Insurance contracts on CDOs are structured as pay-as-you-go CDS, which means that the insurer only pays bond holders for any lost principal and interest as it becomes due. There is no ability to “put” defaulted bonds back to Radian. If Radian can indeed suffer 10% defaults without paying any amount on its insured pools, then what would be owed given a 13% cumulative default rate over 5 years would be small.

Given Radian’s minimal exposure to residential mortgage loans in their Structured Finance portfolio, it is likely that losses from this unit would be consistent with historical norms. Even if corporate bond default rates rise significantly, Radian has enough subordination that losses from this area are likely to be small.

The bigger concern is in Radian’s traditional mortgage insurance business. It is our opinion, and probably that of most observers, that 2008-2010 may see more home foreclosures than any period since the Depression. Getting a good handle on exactly what percentage of homes may be foreclosed upon is difficult, given that a larger percentage of total loans originated in 2005-2007 were sub-prime than in past periods. In addition, a much higher percentage of loans were made based on stated income during this period, which are likely to default at a higher rate than historical norms. Therefore foreclosure rates during past recessionary periods will not be a good gauge.

Fortunately, an exact estimate of Radian’s losses is not necessary to suggest that the current book value discount is unwarranted. Instead, we can make an estimate of what kind of losses would justify the current stock price, then opine on whether that level of losses are reasonable.

First, we note that Radian currently has $1.095 billion in loss reserves, and is currently priced at a discount to book value of $2.594 billion. Therefore in order for book value to decline to the current stock price, the company would have to suffer $3.7 billion in losses with no offsetting new business.

According to Radian, their default rate on sub-prime MI is approximately 18%, Alt-A is 8% and prime is about 3%. Note that sub-prime defaults have increased markedly, from 3Q 2005 in sub-prime (was 14%) and Alt-A (was 6%) but not in prime.

Putting these default rates into Radian’s portfolio, and assuming 100% severity, we get $1.9 billion in losses from MI. This is almost exactly half the amount needed to justify Radian’s current book value discount. Hence if Radian’s defaults immediately doubled, that would justify its book value.

Of course, realistically, Radian’s losses would not occur all at once and in the mean time, they would write new business. Logically, mortgage insurance premiums are likely to rise in 2008, as underwriters are likely to value MI more in a weaker housing environment. In addition, underwriting standards overall are likely to improve in 2008 and onward, indicating that loss rates should improve for the 2008-2009 vintages compared with the 2006-2007 vintage.

So in order to actually justify Radian’s current price, MI losses would have to more than double. If we assume that most of the current losses in Radian’s portfolio are coming from 2005-2007 vintages, loss rates are likely to “burn out” at some point. It is logically impossible for loss rates to remain elevated in perpetuity. If we assume that half of all sub-prime loans from 2005-2007 will eventually default, and 18% has already defaulted, then we’d only expect 32% marginal losses going forward. If we stuffed all those losses into 2008, that would still be less than the loss rate required to justify Radian’s book value.

I conclude that Radian’s stock price is undervalued. I am long Radian stock.