Residential loans fall into one of two categories: Agency — eligible for programs offered by Fannie Mae, Freddie Mac, or Ginnie Mae (FHA/VA) — and non-Agency. As of late 2015, Agency loans are approximately 90% of new originations. According to the Federal Reserve Board, mortgage debt outstanding in that year totaled approximately $13.5 trillion. By contrast, the private (non-Agency) market consist of just over $1 trillion of this balance. The importance of the non-agency market is disproportionate to its market share, because non-Agency programs, typically sold through whole-loan execution, are often where lenders can find a niche to differentiate themselves from the competition.
Agency vs. Non-Agency Loans
Fannie Mae, Freddie Mac, and FHA/VA issue guidelines defining the loans that that they will purchase. Banks, mortgage companies, and other originators, generate loans that meet those guidelines, allowing consumers access to multiple sources for home lending. Originators compete on rate, price, and service, while providing consistent access to standardized financing vehicles offered by the Agencies. This creates liquidity in the market by providing a known source of funding for mortgage lending.
The non-Agency market, comprising loans that are not purchased by the Agencies, serves a different tier of consumer, many of whom have been left out of the housing recovery due to the drastic reduction in capital sources for non-Agency mortgages since the market down-turn began in 2007. These customers may not qualify for an Agency loan for any number of reasons, including lack of credit history, self-employed income, inconsistent employment history, excluded property type, or loan balance. Non-Agency loans are typically originated by banks to be held on their balance sheet (“portfolio;” hence the term, portfolio lending). In order for this additional risk to be worth it to the portfolio lender/investor, they typically carry an interest rate of 50-150 basis points (bps) over Agency paper. As of late 2015, there is increased demand across various financial institutions for such higher-yielding paper, since prevailing yields have been unacceptably low for many investors for an extended period of time.
Funding for Agency Loans
The vast majority of the $13 trillion mortgage market is funded by mortgage bonds issued by the Agencies list above. The U.S. government issues mortgage debt via bonds and securitizations that are used to fund mortgages for consumers. These instruments are then traded on the open market, allowing institutions and individuals to invest in the housing market with a U.S. government guarantee, virtually eliminating credit risk.
Funding Non-Agency Loans
The source of funding for non-Agency loans primarily consists of private, non-government money from life insurance companies, hedge funds, REITS, large banks seeking yield for their portfolios, securitizations, and some other types of insurance companies. The low interest-rate environment in the mid-2010s compressed Net Interest Margins (NIMs; the profit a bank makes from its lending), and forced depositories and others to seek additional income sources.
This environment makes higher-yielding non-Agency mortgages especially interesting in the life insurance market. Historically, this capital has been reserved for the lowest risk investments with the highest liquidity. However, the reduced income from depressed interest rates, combined with the hit to the market value of underlying security accounts, has forced insurance companies to chase yield. Agency loans do not have as high a yield as non-Agency loans; therefore, insurance companies seeking yield often look favorably at purchasing non-Agency loans.
Hedge funds will purchase loans for the yield and cash flow, or, they will aggregate and securitize the loans for sale on the secondary market. Hedge funds have a higher cost of funds than traditional banks because they do not have access to the overnight window or other government money. As such, their primary capital sources are private funds, institutions, and retail investors. These investors require a higher rate of return, which is obtained through leverage compounding the higher interest rate on non-Agency mortgages. By levering their capital, a mortgage that yields 4.5% can be booked with a double digit return.
Hedge funds are also originating and purchasing loans (often via correspondent relationships) to securitize and sell. In the securitization process the debt is rated and tranched, creating a more liquid investment in comparison to raw whole loans. These instruments are sold for a higher premium due to the tranching of both prepayment and credit risk. This allows investors to purchase a specific piece (tranche) of the overall securitization with just the credit risk and cash-flow characteristics it desires to match its liability structure and risk/return profile.
Because hedge funds are not regulated by Dodd-Frank, they can lend to a tier of customers that the Agencies will not serve. Whether it is due to self-employment, credit impairment, down payment limitations, property type, or any number of factors, these investors fill an important need for consumers and the housing market. It is clear from the improvement in the housing market, that coincided with the expansion of the lending universe, that much of the recovery is due to the easing of credit restrictions in lending in the non-Agency market.
There are a limited number of residential mortgage REITs in the market. Due to their tax structure, REITs focus on real estate as opposed to debt. The existing residential REITs primarily participate in the market by purchasing loans on the secondary market as opposed to originating new loans. The REIT model is similar to a fund that only buys.
Types of Non-Agency Loans
In the mid-2010s, Mortgage Lenders were venturing into new products that are further down the credit curve and/or allow for impairments that previously may have resulted in prohibitive risk-based pricing adjustments. Specifically, larger community banks were underwriting portfolio mortgages to higher loan-to-value ratios (LTVs), lower FICOs, along with alternative doc types. Generally, only one of the three criteria deviates from what make a loan Agency-eligible; another way to say this is that portfolio lenders seek to avoid excessive risk layering. Banks still need to be able to justify to regulators why they make a loan, and “the relationship” is not an acceptable answer. As this style of community-bank lending becomes more widespread, the secondary market for these loans is beginning to expand. Other banks are willing to pay a premium for these loans, because they can avoid the overhead associated with retail loan origination, and the higher yields on the loans are adequate compensation for the (perceived or actual) incremental credit risk.
Underwriting for Non-Agency Loans
Non-Agency Loans allow for greater flexibility and customization of the underwriting process. While the underwriting standard must remain transparent, the secondary market is allowing banks to serve a diverse range of borrowers.
To serve the needs of landlords and larger property owners, funds sometimes lend against an entire residential portfolio. These are blanket loans, secured by as many as 100 properties, typically with a maximum LTV of 65%. The advantage is the cost-efficiency of collecting payments on one loan. The borrower is generally a professional manager, theoretically less likely to default as there is more at stake than one property. There is also additional security afforded by the diversification of the cash flow. If a few properties are vacant, they are funded by the remaining occupied rental properties. The loans typically feature a five- to ten-year term and are either interest-only, or 20-year fixed rate. The cash flow is underwritten in a manner similar to a commercial loan, with a Debt Service Coverage Ratio calculated to determine ability to support the debt repayment.
Borrowers with Low Credit Scores
Borrowers with low FICO scores, recent adverse credit events including bankruptcy, foreclosure, and mortgage lates, may still obtain a mortgage from a community bank, with the idea being that the loan serves a need in the local community. As an example, our firm has a bank client which will originate a residential, owner-occupied mortgage for a borrower that is one year out of a Chapter 13 bankruptcy with a perfect post-bankruptcy pay history. On the other hand, Agency guidelines require four years pay history from discharge. The logic is that if the borrower paid their Trustee payments on time and has re-established credit, and all the debts were paid as opposed to a Chapter 7 in which the debts are discharged then the borrower has proven the ability to pay obligations in a timely manner. Our client has determined that this is a good risk, and to date has not had a default.
Loans for self-employed borrowers who do not meet Agency guidelines virtually did not exist after the market meltdown until the mid-2010s, when the secondary market began to demonstrate an appetite for them. The loan criteria are consistent with Agency guidelines in terms of LTV, FICO, etc. There is opportunity for lenders to be flexible and differentiate themselves in how they arrive at an understanding of the borrower’s income, and the calculations used to determine Ability to Repay. These calculations include verifying income with both business and personal cash flow, understanding the ownership of the business, and understanding the borrower’s industry as a whole. Underwriting mortgages for self-employed borrowers requires a deeper understanding of each loan, and of the borrower. Again, it is acceptable to have one deviation from Agency underwriting. As long as the credit and collateral are acceptable, the lender has the ability to manage risk in the income or ability to repay.
Non-bank lenders are originating higher-LTV loans for investment properties. These mortgages require much higher FICO scores, and full income documentation, and significant reserves. The borrower is more likely to default on an investment property than their primary residence, and the occupants of a rental home are less likely to maintain the home; therefore, the bank is less likely to recover a foreclosed property in a marketable condition.
Pricing of Agency and Non-Agency Loans
Funds issuing non-Agency mortgages require a lower LTV, since the loans are not insured by the U.S. government and any losses incurred during the default period are born directly by the investor which owns the loan. Non-Performing Loan (NPL) expenses include deferred interest, default interest, attorney fees, court fees, maintenance fees, and more. Therefore, the equity in the property must be sufficient to cover all fees for foreclosure plus the transaction costs associated with re-selling the property on the retail market. Alternatively, the seller may dispose of the assets via a bulk sale to other funds that specialize in distressed assets. In the mid-2010s, the market for these NPLs priced out at 65 percent, or less of the value of the property. This level represented a significant increase in market values of NPLs from the early stages of the 2007-2008 market contraction, when NPLs typically sold for 30 to 45 percent of the property value. Banks do not typically hold non-accruing assets, including non-performing loans, on their books. The secondary market will purchase the NPLs, allowing the bank to recapitalize.
Pricing of any asset is a relatively transparent indicator of the risk involved in the investment. The lack of mortgage insurance, any government guarantee, or subsidized capital drives the yield required by the secondary market higher in an attempt to normalize risk and reward. To put it another way, what would an investor have to earn from his investment to make it worth his while to lend to someone that the Agencies would not lend to? This premium is a function of just how far out the loan is on the credit curve. The higher the perceived risk, the higher the yield the investor will require to fund the loan. As of late 2015, this premium was 50 to 150 bps. In the early/mid 2000s the premium could be as high as 650 bps, as lenders originated many loans that ultimately did not perform.
Note that yield is correlated with the market’s perception of risk. This creates an opportunity for portfolio investors which may have the expertise to safely originate loans for which the market imposes a high risk premium. Not only does this provide the opportunity to earn higher returns in portfolio however once the loans are sufficiently seasoned and the lender can show actual performance of the loans, there may be an opportunity to profitably recapitalize, at a premium, by selling those loans into the secondary market.
A loan’s value is a sum of all parts: LTV, Doc type, FICO, and so on. Any one factor can be a compensating factor, and any one can be a deal-killer. In the portfolio space, there is typically the ability to have flexibility in accepting one impairment, assuming the other two are strong enough to compensate. This is the underlying theme of portfolio lending: the ability to use compensating factors in underwriting a mortgage. In the 2000s, lenders such as BNC, Countrywide, New Century, First Franklin and Fremont successfully navigated three criteria, providing 100% financing on investment properties, or 100% financing on owner-occupied properties down to a 580 or even 560 FICO, a credit range that is not generally lendable at all in today’s environment.
A large Midwest bank has a footprint in a major California city with a large concentration of borrowers that are typically self-employed, and whose tax returns do not reflect their cash receipts. This bank took the time to understand the borrowers and the market, including the very strong real-estate values in the city, and the fact that actual bank deposits illustrate a much stronger cash flow than the tax returns would leave a lender to expect. The combination of deeper understanding of the borrowers’ cash flows, the exceptional real estate values, a maximum LTV of 65, weighted average LTV of 61, and strong credit scores, allowed the bank a high degree of comfort with the credit risk.
The bank built out a set of guidelines that allowed them to originate hundreds of millions of dollars of loans to this clientele, with not one single 30 day late payment. MIAC was able to identify buyers of these loans on the secondary market. Buyers were able to get comfortable with the underwriting and loan characteristics once the entire story was communicated, and as a result we were able to develop counterparty relationships with investors which would purchase the whole loans from the bank. The creation of this market improved liquidity for the bank, showed bank regulators that the loans were marketable, and provided the buyer a loan portfolio with an above-market yield.
Trending in Lending
Fannie Mae in the 2000s was mandated by politicians to expand lending to borrowers in “underserved” areas in order to broaden homeownership opportunities. Loans were created, such as the My Community, that allowed 100% LTV, and in some case greater than 100% financing.
Underserved is typically defined as a geographic area where there tends to be lower income and lower homeownership rates, without regard to why these situations exist.
Combining more liberal debt-to-income ratios with credit requirements directly led to greater default rates in the neighborhoods that could least afford the instability. The default rate began to accelerate, eroding property values and further accelerating the deterioration of many neighborhoods.
As political pressures increase, lenders will be compelled to provide home mortgages for more and more borrowers, particularly borrowers who currently are not served by the Agency market. We saw this happen in the early 2000s as FHA was compelled by the government to compete with the Sub-Prime market. The government agencies began to originate loans with smaller down payments, or no down payments, in an attempt to recover business from subprime lenders. In 2015, the Agencies re-introduced 100 percent financing, plus seller-paid costs, allowing borrowers to buy a home with little to no financial contribution from their own pocket.
This is relevant for the whole-loan market, because Agency guidelines are generally the starting point for portfolio loan programs. Lenders will take an Agency product and push one or more criteria a bit further out the credit spectrum in the hope of attracting more borrowers and higher yields, the idea being that they are only slightly riskier than what the federal government insures. In the last credit cycle, the Agencies loosened their guidelines, and private and public sectors began to compete on credit—a recipe for disaster.
What this means for the secondary market is that as yields increase, the funds that are sophisticated enough to understand the risk and get out in front of the market stand to profit nicely, assuming they get out at the right time. We have the benefit of hindsight, having the recent mortgage industry contraction of 2007-2008. The current near-zero interest rates have made it difficult for funds to produce a profit for their investors with traditional investments. Portfolio managers have determined that the mortgages available on the market in the mid-2010s provide an acceptable risk/reward relationship and are willing to continue purchasing these loans. Banks will continue to originate such loans for a variety of reasons, not least of which is an active secondary market willing to purchase their loans for a profit.
Non-agency Niche Products
Foreign National, Individual Tax Number (ITN) loans, and loans to borrowers without Social Security numbers, serve a very specific subset of borrower. There is very specific audience for banks willing to underwrite loans for non-U.S. citizens, Resident Aliens, etc. These loans are originated in an even more secure position than the traditional portfolio loans. Typically, the LTV maxes out at 65%, and the documentation type is usually full doc or some sort of asset depletion loan that will provide a documented measure of security. Ultimately, understanding the nature of these borrowers, the security lies in the real estate and the potential to be made whole via foreclosure.
The borrowers are commonly in very specific industries, such as the oil industry in Houston, where a portfolio lender can gain additional relationship business from originating these loans and serving this population. This strategy continues in vacation areas frequented by a concentration of people from a common home country. Miami, Florida, for example, has a large concentration of second homes and vacation condos owned by foreign nationals. The banks that originate these mortgages typically earn a 75bps yield premium.
There is also an even smaller subset of the lending market that originates non-consumer loans. These loans are exempt from Consumer Finance Protection Bureau guidelines. The comparatively unregulated market serves a very specific audience that is commonly more sophisticated and uses such loans as an investment tool. These include investors that participate in renovation or rehab projects with the intent of reselling the property in less than 12 months. The return on the investment is what makes these loan a practical tool for these borrower. These loans will not be issued against a primary residence. The yields on these loans are typically 500-800 bps through the traditional mortgage market.
What Trading on the Secondary Market Looks Like
Best execution typically comes from a counterparty that is comfortable with the collateral and debt, and is willing to purchase at an above-market price. This may be due to pressure to diversify their holdings, or a lack of yield on the current portfolio, or the belief that, at the margin, the purchase will improve the overall composition of the institution’s loan portfolio.
That being said, considerable cost efficiencies can be created by establishing a flow arrangement between two parties. Once the buyer and seller have gone through the resource-intense process of negotiating legal documents, it is easier for them to do future business with each other. There may be an agreement, either formally drafted or, more commonly, an ad hoc relationship, with a mutual understanding that the parties will transact as frequently as appropriate with agreeable terms. The idea is that both parties benefit from the transaction and thus are motivated to continue the relationship with future transactions. If the market moves away from either party there is the flexibility to forgo a trade until the market is more advantageous.
On a bulk sale, the seller seeks to obtain the best combination of price and surety of close, with the lowest risk of “fade” and the most efficient or fastest close. The more experienced buyer from a pool of bidders will often provide the best execution. That is not to say that the highest buyer is the best buyer. There have been known to be bidders who will submit an above-market bid with the intent of locking up the trade. Then the bidder either attempts to re-trade the seller, broker the loans out, raise funds to complete the transaction, or disappears altogether. Of the above, re-trading is the most common adverse outcome. The bidder will dig deeper into the files and realize they are not what they thought. This is generally no fault of the seller, as all bidders had the same information. The bidder in this example simply did not have the experience to understand the nuances of the individual pool, whether it is specific geography, loan type, seasoning, or collateral that is not what they thought it was. This became especially common in the late 2000s, as many firms flooded the market, chasing deals without having much relevant experience. The frequency of this occurrence destabilized the market, as sellers were fooled and buyers would fall out of the space as they realized it was not as easy as they had thought to close these deals.
The Secondary Market for Non-Agency Loans
Mortgages are either sold as mortgage-backed securities (MBS) or as whole loans. Buyers of MBS are purchasing a security that is backed by mortgage loans. Buyers of whole loans are purchasing the actual loan, the note, and mortgage. Whole loans have the potential to be more profitable, however the purchaser must have the infrastructure to underwrite, service (unless loans are sold servicing retained), and perform due diligence.
Some less sophisticated buyers of whole loans will throw a number in a pool, hoping it works out. At the other end of the spectrum, buyers perform nearly complete due diligence on the loans prior to submitting a bid. This cuts both ways, as they may have the best, most accurate bid, however, the people they are bidding against have a wider margin of error that allows them to offer a higher bid. It is not unusual to see bids on a pool of loans ranging from 74 to 92 percent of the unpaid principal balance.
The vast majority of whole loan trades are from aggregators (banks like Wells Fargo or Chase, that can sell directly to the Agencies) or large originators to the Agencies (like Quicken), on a daily or bulk basis. There are additional origination channels, in the form of mortgage banks and retail or commercial banks, that originate non–Agency loans intended to be sold to other institutions, funds, banks, etc. seeking the higher yields offered by this product. These non-Agency trades will vary in size depending upon the arrangement. It is not uncommon for a large community bank to originate $10-15mm per month of this product to be sold explicitly to a fund or bank that needs the mortgages, but which doesn’t have the origination platform to fill this need.
Putting Trades Together
The firms best suited to identify the best counterparties (buyers or sellers) will be firms that have the greatest exposure to market activity. This includes Agency trades, pricing execution, portfolio valuation, and portfolio hedging. It is important to work with a firm that sees as much activity as MIAC, which, by virtue of the breadth of its business activities, has frequent contact with most of the largest originators, funds, banks, and portfolio companies.
The firm you work with must gather the market intelligence needed to identify and engage the best execution for a given trade. This includes an intimate understanding of firms’ business models on both the buy and sell side, as well as their pricing requirements and investors’ “appetites.” Some dealmakers have, at best, a cursory level of understanding of investors’ objectives. The lack of depth of understanding often yields a sale that has been put out so widely (shown to too many unqualified buyers) that it becomes a trade that many serious funds will avoid.
Best execution comes from a complete understanding of the seller’s needs, both in terms of price and timing, and also in the nuances of the portfolio. The dealmaker you work with must understand these and match the trade with a small number of serious buyers who will focus on the trade and put in a no-fade bid that will close with a very high degree of surety.
Equally important is understanding the current market. There are timing considerations, external events that affect a buyer’s ability to focus on a trade, as well as other market considerations. Turning points in Federal Reserve policymaking, such as in late 2015, show that bid levels can vary widely depending upon any given investor’s opinion of “what the Fed will do.” It is rarely a good idea to “blast” (mass market) a pool, which is what some internet-based firms will do in an attempt to increase their pool of contacts for future deals at the expense of a trade at hand. The best trades are executed through relationships and experience, not the internet. There is much to be accomplished with a phone call or meeting.
Secondary Market Pool Size
Larger pools typically attract a different audience than the typical $10mm trade. There are certain fixed costs associated with a loan trade. For example, legal departments will draw up the same documents whether the trade is one loan or 100 loans. There are certain economies of scale that exist when a buyer schedules man hours for due diligence and closing. This makes the larger trades more attractive to buyers who have the capital to invest. The market participants in larger trades are also generally more sophisticated and experienced when it comes to executing transactions, and tend to have a better understanding of the market and more accurate pricing.
Size can also cause a strong buyer to not participate in an auction that could yield a particularly attractive trade for the seller. When this happens, sometimes a carve-out of the seller’s pool may be sold to this high bidder, with the balance being put out for re-bidding by other parties. With this strategy, there is the risk of the previous high bidder(s) bowing out of the trade because the most desirable assets were carved out. Alternatively, an “all or none” sale will be executed if there is not enough spread to warrant multiple trades with multiple counter parties. In this case, the seller will frequently solicit bids via a whole-loan trader or broker from multiple counterparties. A final bid process provides the potential buyers one last chance to win the pool with a final best execution offer.
There is some value in counterparty diversification by trading a pool with multiple partners; however, there is an obvious increase in transaction costs, as discussed above. There have been times in my career when a deal did not trade for completely unforeseeable reasons that were neither parties fault. Had there been multiple buyers, the trade would likely have closed with one party or the other.
Brendan Teeley, Vice President, Whole Loan Sales & Trading, MIAC Capital Markets
MIAC Perspectives – Summer 2016
Whole Loan Execution