Description

MTG is the largest mortgage insurance company in the United States[1].  By way of background, private mortgage insurance (“MI”) covers the first ~25% of losses on a given loan in the case of default by the borrower.  The mortgage insurance industry exists largely to facilitate the purchase of low-downpayment loans by Fannie and Freddie, who cannot buy loans with LTVs higher than 80% by charter absent some third party insurance.  Thus, the GSEs (i.e. Fannie and Freddie) and the mortgage insurance industry are co-dependent; the GSEs are the MI’s customers, and mortgage insurance currently covers ~$140 billion of GSE exposure with MTG being their largest MI counterparty. 

Outlined below is a simple arbitrage opportunity whereby one creates a 5 year out-of-the-money equity option by going long MTG’s 5% senior unsecured convertible bond and long MTG 5-year senior unsecured CDS (i.e. short credit).  Importantly, the convert is “deliverable” into the CDS in the event MTG defaults and is putable at par in the event of a change of control.

Simplified “fixed” cash flows for both instruments (assuming 1x hedge ratio though we actually are a bit “shorter” for now given overly bullish Street estimates) through maturity:

As can be seen, the total return of the “fixed” cashflows of this investment is +4%.  This is also our downside case.

What interests us about this investment, however, are the potential “variable” cashflows as a result of the equity option embedded in the converts, which we believe the market misvalues.  The 5% convert can be exchanged into 74.4 shares, which is equivalent to a conversion price of ~$13.44.  This strike price over 500% higher than MTG’s current stock price of $2.19, which is why the market assigns virtually no value to the equity call option. 

1. Market Underestimates Earnings Potential If MTG Survives

Given MTG’s high leverage and unquantifiable regulatory risk, it is difficult to precisely handicap their chances of making it through this cycle as a going concern.  The market currently thinks it is fairly likely and while we are more optimistic, it's difficult to have a lot of conviction. What we have issue with is what the market thinks will happen if MTG survives. The convertible bond’s trading levels imply that MTG will just limp around.  We do not think this is likely.  MTG’s earnings have the potential to spike coming out of this downturn.  As a quick-and-dirty illustration of their earnings potential, prior to the crisis in 2006, MTG generated $565 million of earnings ($2.81 EPS using today’s share count) on similar revenue as 2011 even though its fixed costs were higher then.  While we do not foresee $2+ of EPS in the near future, several factors make us cautiously optimistic that this scenario may happen before our convert matures in five years.

First, MTG’s new delinquency notices have been declining since 1Q 2009 and MTG’s inventory of delinquent loans peaked 4Q 2009.  New notices have been declining at year-over-year pace in the high teens for the past two years. 

Second, defaults are cleansing.  MTG has tightened underwriting standards and is now writing what is likely very profitable new business.  As old vintages of mortgages roll off, new delinquencies will continue to drop.  In addition, loss severities will be lower as housing prices stabilize. 

Third, MTG’s revenue opportunity is enormous.  During the downturn, private MI lost significant market share to FHA (the government’s alternative to private insurance) as a result of the industry’s capital constraints and more conservative underwriting.  However, this trend began to reverse in 2010 as FHA increased pricing.  Depending on how the GSEs are reformed, the pace that the industry claws back market share has the potential to meaningfully accelerate if the economy picks up steam and additional capital is freed up for new business. Within the MI industry, MTG now has significantly less competition after PMI and Republic were put into run off by regulators. 

While MTG is writing far less business then before the crisis, if the mortgage market recovers over the next five years, there is a big opportunity for MTG.  We expect NIW to continue to expand.

While short-term earnings are very difficult to forecast (though are particularly bearish heading into Q2), we believe MTG will return to profitability in the next few years and generate substantial earnings if regulators allow them to operate.

2. Operating leverage / Earnings volatility

MTG has significant operating leverage and slight deviations in fundamentals translate into big earnings swings.  For example, after 11 consecutive quarters of substantial net losses, MTG almost miraculously generated a GAAP profit of 13 cents per share in the second quarter of 2010 before retreating in Q4 to a loss of almost $1 a share. 

The reason for the earnings volatility is that small changes in fundamentals (e,g, delinquencies, cures) have a magnified impact on profitability.  The “cure” ratio (i.e. the amount of delinquent loans which cure divided by new delinquency notices) has been especially volatile ranging from 71-149% over the last couple of years.  A 10% swing in cure rate translates into approximately $400 million swing in annual pre-tax losses or ~$2 per share.  Average analyst estimates for 2012 EPS is $(1.64).

3. Potential value even in run-off

While estimating MTG’s book’s ultimate losses is a futile exercise, one can estimate losses (in runoff) using a framework similar to the one Affton1 used in his/her VIC write-up in January.  I won't rehash it here, but my asusmptions are a bit more pessimistic.  Importantly, the last time MTG provided a lifetime loss forecast in 1Q 2010, they anticipated $11.9 billion in total claims paid.  MTG has already paid almost $5.5 billion of claims through 1Q 2012.  We expect losses on business written more recently to be minimal.

In addition to $6.4 billion of cash and investments, MTG’s book of business generates annual premiums 63bps per annum.  In 1Q 2010, MTG estimated its book would generate cumulative net premiums of $5.8 billion in runoff.  Given the book is about 18% smaller now, ~$4.5-5 billion is a reasonable estimate of future premiums in runoff.  As investment income should approximately cover other operating expenses, we believe MTG’s cash resources of $11 billion will be more than sufficient to cover insurance claims and holdco debt maturities even in runoff and leave substantial residual value for equity holders.

If a few things go right for MTG, we believe EPS of $2+ a few years out is not unrealistic.  In that scenario, the convert value will be well in excess of the "bond floor".

Why does this opportunity exist?

1. Stock Price Volatility vs. Earnings Volatility

"Volatility is a symptom that people have no idea of the underlying value" –Jeremy Grantham

One of the key inputs in the traditional convertible valuation model is volatility (“vol”).  Vol measures the variation of the price of an asset (in this case MTG stock).  Vol can be either historical (calculated) or forward-looking (implied).  Historical vol is clearly useless when valuing a five year option as the world will look very different in five years.  Forward-looking vol, which equals the implied vol based on equity options’ trading prices, is limited by the trading markets for equity options.  There are no five year equity options for MTG and, thus, this metric is also fairly useless.

For these reasons, we much prefer to look at earnings volatility, which we believe is extremely high for MTG as discussed earlier.  The market does not appreciate MTG’s earnings volatility.

2. Accounting Makes Modelling Very Difficult

MTG’s liabilities (insurance reserves + debt) represent over 11x its market capitalization.  However, this number vastly understates MTG’s expected losses.  This is because GAAP requires MTG to book reserves only after a loan defaults (i.e. 45+ days delinquent); even if MTG expects a loan to default, they will not book a reserve until after the default actually occurs.  Because MTG has not disclosed its internal loss forecast since early 2010, the market has trouble valuing the company and shies away from an investment in MTG. 

MTG’s 9% subordinate convert due 2063 is a good substitute for the senior bond as its conversion price of $13.50 is nearly identical to the 5% convert.  In exchange for less seniority, significantly longer duration and a coupon that can be deferred at the Company’s option, one gets a much bigger (and longer-dated) coupon and a significantly lower entry price. 

We do not particularly mind the longer duration given we expect things will play out one way or another in the next few years.  With regard to seniority, given the insurance “leverage” at the opco, we don’t think the converts will recover materially less than the senior notes; it is such a small sliver of the cap structure after all.  The coupon deferral bothers us, however, 20 pts plus the larger coupon seems more than adequate compensation for the additional risk.

Hedge Ratio

At inception, we have used a hedge ratio of ~1.2x (net short), trading a little equity upside in order to profit some more in the default scenario.  Given MTG’s leverage and unquantifiable regulatory risk, we believe this is prudent for the time being.

Investment Risks (related to the value of equity option not necessarily the value of our position)

We lose if MTG plods along (actually we make 4% over five years).  Thus, things that mute MTG’s earnings power coming out of this downturn are our risks.  (It is worth noting that the 9% long-dated coupon of the subordinated convert will provide additional protection in this scenario.)

 1.     Dilution

While the conversion price does adjust for certain dilutive events, it will not adjust in the event of additional equity or convert offerings.  If MTG continues to suck wind, MTG may decide to raise additional capital.  In addition, the GSEs and regulators may pressure MTG to raise capital.  MTG may also raise additional capital in order to capitalize on new business opportunities.

2.     Seizure by Regulators: Risk-to-Capital Ratio Breach

MTG is required by certain state insurance regulators to maintain certain levels of statutory capital.  New insurance written in jurisdictions that have capital requirements represented approximately 50% of MTG’s new insurance written in each of 2010 and 2011.  One of the most common metrics used by regulators is the risk-to-capital ratio.  MTG will breach its risk-to-capital ratio within the few quarters. 

MTG’s primary insurance regulator has granted MTG a waiver through December 31, 2013 allowing it to continue to write business even in the event of a breach of capital requirements.  While not all state regulators have granted waivers yet, MTG has implemented a contingency plan, which has been approved by both the GSEs and state regulators. 

Under this plan, MTG will write new business out of a newly formed Tier II subsidiary (wholly owned by Mortgage Guaranty Insurance Corp.) in states where the primary insurance company is not approved to write insurance.  The subsidiary, MIC, has been capitalized with $400 million of equity already, which will enable MIC to continue to write new business even in the event of a capital breach.

Both the GSEs and state regulators may change their mind regarding approvals. 

3.     Political

There is currently a high degree of uncertainty regarding the future role of mortgage insurance in the mortgage market.  In addition to the question of how congress will restructure the GSEs, mortgage insurers face additional political risk. 

As one example, Dodd- Frank requires a securitizer to retain at least 5% of the risk associated with mortgage loans that are securitized. This risk retention requirement does not apply to mortgage loans that are “Qualified Residential Mortgages” (“QRMs”) or that are insured by the FHA or another federal agency.  As currently drafted, the exemption from the risk retention rule does not apply to loans with private mortgage insurance. The regulators requested public comments regarding an alternative QRM definition by August 1, 2011.  The final rule has not yet been issued, but, depending on its drafting, may have a material effect on MTG’s revenue prospects.

We do not have a high degree of conviction regarding how things will play out.

4.     Countrywide/BofA Litigation

In December 2009, Countrywide filed a complaint that alleges that MGIC has denied, and continues to deny, valid mortgage insurance claims submitted by Countrywide.  From January 1, 2008 through December 31, 2011, rescissions of Countrywide-related loans mitigated our paid losses on the order of $435 million.  A hearing is scheduled for September 2012 regarding certain of the loans at issue.  The total damages Countrywide seeks is ~$1.1 billion.

We do not have any view on how things will play out.

 5.     Crappy Business

Mortgage insurance is not a good business.  The product is highly commoditized and insurers have very little pricing power.  It is also highly cyclical.  During boom times, mortgage insurers chase volume and print money and then they tend to lose it all (and oftentimes more) during busts.  MTG management is unsophisticated and has not effectively hedged credit or interest rate exposure.

 6.     Lack of Catalyst

In some ways, this investment boils down to time arbitrage.  Thus, there is no catalyst.

Catalyst