Rosetree Mortgage Opportunity Fund
Company Name :
Company Address :
Country : USA
In early December 2008, Isabel Villegas and her team at Rosetree Capital Management were evaluating the purchase of a pool of U.S. residential mortgages. Rosetree managed over $20 billion of high yield and distressed debt portfolios for institutional clients, and had recently formed an investment vehicle to take advantage of dislocations in the U.S. residential mortgage market. The Rosetree Mortgage Opportunity Fund had $1.2 billion in capital and a mandate to acquire troubled mortgages from banks and other motivated sellers. The goal was to purchase whole mortgage loans at a discount and to work with individual borrowers to restructure their debts. Performing mortgages could then potentially be resold in the secondary market. Villegas had already deployed a quarter of the fund’s capital, and expected to be fully invested within 12 to 24 months.
The Global Credit Crisis
In late 2008, the Unites States was in the midst of its worst financial crisis since the Great Depression. The crisis was triggered, in early 2007, by rising default rates on subprime residential mortgages, issuance of which had surged along with home prices between 2001 and 2006. In 2006, approximately 20% of all mortgages generated were subprime, up from 5% in 1994 and 8% between 2001 and 2003. Holders of subprime mortgages and related instruments included banks, investment banks and hedge funds, many of which were highly leveraged. Subprime-related losses forced these institutions to reduce their leverage and resulted in the broad liquidation of assets of all kinds, including equities. Contraction in credit ensued which fueled a global vicious cycle of asset price declines causing further deleveraging and distressed selling. By the end of 2008, global public equities alone had declined by more than 50% from their highs, a loss of over $30 trillion in market capitalization. Financial institutions generally were in a state of severe financial distress, and were being aided by government interventions including injections of capital, asset purchases, and the provision of guarantees
The U.S. Housing and Mortgage Markets in 2008
There were approximately 80 million single-family homes in the United States. U.S. housing prices more than doubled between 1999 and 2006 (Figure 1), peaking at around $20 trillion in total value. As shown in Figure 1, housing futures markets were predicting a 40% decline from peak by year-end 2009 and for prices to stabilize thereafter. The regions most affected had seen massive recent new construction, including Miami, Phoenix, and greater Los Angeles.
Residential mortgages totaled 55 million in number, with outstanding principal of around $11 trillion. Of these, 10 million mortgages with initial principal of over $2.2 trillion were considered “subprime” and “Alt-A”. Subprime borrowers had significantly impaired credit histories and/or high leverage; Alt-A borrowers had more minor credit issues, nontraditional or unverifiable income, high leverage and/or non-owner occupied properties.
In 2008, U.S. residential mortgages were held in essentially three ways: as “whole loans” on the original lenders’ balance sheets (approximately $4 trillion); as pools of mortgages insured and/or held by Government Sponsored Entities (“GSEs”), primarily Fannie Mae and Freddie Mac (approximately $5 trillion); and in special purpose trusts as private label securitizations (approximately $2 trillion).
For a fee of typically 12-18 basis points per annum, the GSEs would insure interest and principal payments on mortgages that met certain criteria. These typically included loan-to-value (”LTV”) ratios no higher than 80% at the time of origination, a maximum loan amount (for example, less than $417 thousand for single family homes), and an adequate borrower credit score. Loan originators would purchase GSE insurance on pools of qualifying mortgages and sell them in the public market, primarily to institutional fixed income portfolios. Provided that the GSE’s remained sound, the holders of these mortgage pools would not bear any credit risk, only the uncertainty of mortgage prepayment.
Private-label securitizations involved the creation of debt securities collateralized by pools of mortgages but not insured by the GSEs. For example, a $100 million pool of mortgages might serve as collateral for $70 million of senior debt and $30 million of junior debt and equity. Principal payments received on the mortgage pool would first go to service principal owed on the senior debt before any payments were made on the junior liabilities. This way, senior debt holders would have a cushion of 30% in mortgage losses before their principal became impaired. With an adequate cushion, high grade mortgage backed debt instruments could be created without having to insure each mortgage.
The cushion required for the senior debt in a private-label securitization to obtain a “AAA” rating depended on the rating agencies’ assessment of the credit quality of the underlying mortgages. Over time, the standards for what constituted adequate cushion appeared to have declined, and by 2005 and 2006, debt rated AAA was being issued against subprime mortgage pools with cushions below 20%. Subprime and Alt-A mortgages had been the fastest growing segment of the mortgage market, increasing from less than 3% of all mortgages in 2002 to a peak of 13% by 2005.
Subprime and Alt-A mortgage delinquency rates were relatively stable until 2006, before rising dramatically (Figure 2). Delinquency rates were low in 2004-2006 in part because borrowers were easily able to refinance their mortgages or take out second mortgages when they ran into financial difficulty. The phenomenon was coined “a rolling loan gathers no moss”. Deliquencies were highest on mortgages issued in 2006 and 2007. Many were issued with low initial “teaser” rates, and borrowers could not afford to pay the higher market-based rates upon reset. Incomes also were lower in the weakening economy, which together with falling home prices meant that other sources of credit such as home equity loans were no longer available. By May 2008, there were 305 thousand loans in the process of foreclosure, and over 1.1 million subprime and Alt-A loans with at least 30 days delinquency. By some estimates, 3 million subprime borrowers would eventually lose their homes, along with the $60-100 billion of personal wealth they had invested as down payments and mortgage payments.
The Residential Mortgage Investment Opportunity
Rosetree Capital Management had a long history of investing in distressed credit instruments, but the principals had never seen anything like the current breadth and scale of dislocation in the credit space. They were particularly intrigued by the opportunity in troubled residential mortgages. Large and small banks alike were under enormous pressure to liquefy their balance sheets, and residential mortgage loans and securities represented substantial portions of their balance sheet assets. Whole loans (mortgages that had not been securitized) seemed particularly attractive, since it would be possible to identify and work with individual borrowers to restructure their debt. Securitized mortgages by contrast were held in broadly-owned trusts controlled by third-party servicers, and gaining the right to negotiate with end home borrowers would require purchase of all of the ownership interests in any given trust.
Rosetree anticipated they would be able to purchase mortgage loans at prices that reflected internal rates of return in the range of 15-20% per annum based on expected loan cash flows less servicing expenses. Successful mortgage modification could add significantly to this expected return range.
Rosetree had hired Isabel Villegas to build the capability of investing in residential mortgage loans. Villegas had previously headed the residential mortgage lending operation at a large U.S. institution, and she was fairly quickly able to assemble a team of experienced professionals. The necessary capabilities included:
1) Sourcing mortgages in the secondary market, which required relationships with brokers and the secondary market desks of the selling financial institutions;
2) Performing due diligence, which included checking loan files, valuing the underlying properties, and verifying borrower representations;
3) Servicing mortgage loans, including collecting payments and working with end borrowers to restructure their debts; and
4) Pricing, packaging, and selling restructured mortgages in the secondary market.
In addition to building her organization, Villegas also spent significant time raising capital, primarily from large U.S. college endowments, foundations, and pension funds. Given the strength of her team and the unique characteristics and attractiveness of the opportunity, Villegas thought that Rosetree would have little difficulty raising $3-4 billion. Investors, however, were extremely skittish about the economy and the dire conditions in the capital markets. The losses that investors had already sustained in their portfolios, moreover, made them averse to committing to new investment strategies, particularly those containing the term “mortgage” in their description. Upon reflection, Villegas was pleased with the $1.2 billion they had been able to raise. The fund had a private limited partnership structure and charged a management fee of 2% per annum plus 20% of realized profits. The fund could remain invested for up to five years, and was permitted to use moderate leverage.
The Loan Portfolio
Villegas and her team currently were evaluating a loan portfolio that was being sold by a large West Coast bank. The portfolio consisted of 175 loans with a total outstanding balance of $65 million (Exhibits 1 and 2). The loans were all first lien mortgages, and over 70 percent of them were the only mortgages outstanding on the property. The majority of the loans were against owner-occupied single-family residences. The average FICO score on the loans was 679, and approximately one quarter of the loans had FICO scores below 650.
Although a “good” FICO score was considered to be above 700, the loans compared favorably to the nationwide subprime pool which had a median FICO score of 620.
The portfolio included loans from 20 different states, with 41 percent of the properties located in California and nearly 12 percent in Florida (Exhibit 3). All except 35 of the mortgages in the pool represented loans to finance original property acquisition as opposed to mortgage refinancing or to obtain cash for other uses.
The selected loans were relatively recent, with over 90 percent of the loans originated in early 2007. However, due to deteriorating lending standards, transactions closed between the second half of 2005 and the first half of 2007 generally were exhibiting worse performance with each successive quarter of origination. Overall, 13 out of 175 loans in the portfolio were in delinquency. However, the delinquency on these loans was below 30 days, and only a small fraction of recently delinquent loans would usually default. Moody’s was projecting that only 3.75 percent of loans delinquent for less than 30 days would default in the next 18 months, while 30 percent of loans delinquent between 30 and 60 days would default in the same time frame.
The average loan-to-value (LTV) ratio of the mortgages in the portfolio was 76 percent based on home values at origination (Exhibit 4). Given the fall in home prices, Rosetree estimated that the LTV now averaged 96 percent.
An important part of the investment strategy was to actively renegotiate acquired loans, as Rosetree believed that avoiding foreclosure was critical to maximizing loan value.
Loan renegotiation involved a Rosetree servicing agent contacting the borrower and offering to renegotiate the terms of the loan. Making contact was not always so straightforward, as many borrowers would take pains to avoid speaking with their mortgage lenders. Rosetree had, however, developed creative techniques for making borrower contact, including through the use of technology and incentives (such as rewards for returning phone calls). Unlike many loan servicers, Rosetree would also target borrowers who were still current in their payments but were projected by their models to become delinquent in the next 6-12 months absent intervention and restructuring.
The ideal case was that borrowers would refinance their mortgages from other sources, and Rosetree would indicate their willingness to waive refinancing fees and prepayment penalties. Mostly, however, borrower obligations would have to be restructured. In a typical situation, Rosetree might agree to bifurcate the existing first mortgage into new first-lien and second-lien mortgages. For example, an original $300,000 floating-rate first mortgage might be restructured into a new $240,000 first mortgage and a new $60,000 second mortgage. The interest rate on the new first mortgage would usually be a fixed rate deemed to be affordable given the borrower’s income and non-housing expenses, while the interest rate on the new second mortgage would be very low, often 0% for a period of time. The second portion would sometimes be forgiven altogether in cases of extremely high LTVs. Rosetree would insist, among other things, that the borrower provide full documentation of his or her income and assets, e.g. copies of wage and bank statements. This way, the renegotiated first mortgages would comply with GSE standards and could be resold at close to their par value. Only a third of the portfolio loans had full documentation and nearly 20 percent of the loans had no documentation at the time Rosetree was reviewing the transaction.
Cash Flow Projections
In projecting the cash flows on the loan pool, Villegas’ team modeled several economic scenarios including “slow economic growth”, “moderate recession” and “severe recession” (Exhibits 5 and 6). The projections included forecasts of realized losses due to foreclosure sales and the associated expenses such as brokerage commissions, property taxes, and maintenance and repairs. The total cash flows on the portfolio would be composed of receipt of principal and interest payments, proceeds from liquidations of foreclosed properties, and servicing costs. The team also modeled the improvement in cash flows that could be expected from loan modification, shown in Exhibit 6 for the “severe recession” scenario. Villegas felt that due to the complexity and uncertainty of this transaction, it was important that purchase of the mortgage portfolio would be an attractive investment even if there would be no modification to the current loan terms.
Villegas had three days to provide an indicative price to the selling bank. If her offer was accepted, she would have two weeks to conduct due diligence on the pool, with an opportunity to reject loans which did not conform to the sellers’ representations. Assuming the satisfactory completion of diligence, the transaction would close quickly, and the loans would transfer to Rosetree’s servicing platform in time for the next monthly payment cycle. Villegas believed that Rosetree was the only bidder currently reviewing the opportunity, but could not be sure.