First Marblehead's Deals Explained
Board: First Marblehead Corp.

Related Links
Discussion Boards

By Tiddman
June 15, 2007

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

I thought I'd post a simplified example of one of FMD's deals to help explain how they work. This is perhaps oversimplified and may omit important details, and is for illustrative purposes only, but might be interesting for people who want to understand the mechanics of the securitizations.

A group of lenders has 1,000 private student loans that they want to securitize and sell, so they hire FMD. The loans all carry a rate of LIBOR+500 and are all for $10,000. The number of years until the borrower's first payment is evenly divided between 2, 3, and 4 years, resulting in an average of 3 years. The remaining term on the loans is 20 years after the first payment, resulting in an average of 23 years.

Each loan carries a 7.5% application fee which is rolled into the loan. For each loan, $750 goes to TERI, of which $150 goes towards their costs and internal reserving, and $600 goes into the securitization trust as a cushion for defaults. This results in a $600K TERI pledge fund.

The ABS investor buys the security and gets a yield of LIBOR+40. Of the cash they put up to buy the security, 20% goes into a reserve account, 10% goes to FMD for their up-front fee, 5% goes to each lender as a marketing fee, and 5% goes into frictional costs. So the ABS investor pays $14M for a security that contains $10M of student loans, and the "extra" $4M goes to FMD, lenders, and the reserve account.

During the first 2 years, no borrowers are making payments (i.e. they're still in school), so payments to the security holder are made from the reserve account. With LIBOR at 5.35%, the LIBOR+40 rate for the security is 5.75%, so these two years of payments add up to $1.61M or 11.5% of the fund's face value, which is absorbed by the reserve account which started at $2.8M (i.e. 20% of face).

Starting in year 3, one third of our borrowers start to make payments. Borrowers make payments at LIBOR+500 (i.e. 10.35%) while the security pays the investor LIBOR+40 (i.e. 5.75%), so there is 4.60% of excess spread on these payments. The borrowers are on a 20 year amortization schedule so their payments are $98.83/mo. Multiply by one third of 1,000 borrowers and 12 months and, assuming no prepayments during this year, that's $395,320 received by the trust in the year, while it pays out 5.75% of face or $805,000, or a net cash outflow of $409,680. The reserve account started the year at $1.19M so it ends the year at $780,320.

In year 4, two thirds of our borrowers are paying. Using the same numbers as above, the trust has cash outflows of $14,360, so the reserve is down to $765,960.

In year 5, all of our borrowers are paying, so the trust has net cash inflows of $380,960 after making its payments to the bond holder. We have about 18 years left to run, and an annual surplus of $380,960, assuming no defaults or prepayments.

However, defaults and prepayments do occur. The default and prepayment assumptions are spelled out by FMD and TERI. Let's say that 8% of loans default and recovery on defaulted loans is 40%, resulting in a loss rate of about 5%. The TERI pledge fund was about 6% of principal, so we should completely cover those defaults with a little room to spare.

So what happens to all of this excess interest? Well, those are the residuals that everyone talks about. Those are split 75/25 between FMD and TERI if and when they do occur. Note that FMD says that they expect to start getting their residuals 5-7 years after the deal is done. This jives with my oversimplified analysis above.

The obvious risks to the residuals are defaults and prepayments. Defaults could come in way worse than expected. If this were to occur, TERI should be on the hook for some of the excess defaults, since they guaranteed the loans after all. Maybe if things really go pear-shaped they won't be able to cover them or may take a hit to their ratings. But they theoretically have the best database on private student loan performance and so should be in a decent position to avoid any reckless underwriting.

The prepayments are perhaps less predictable, and are driven by things like interest rates, availability of alternate forms of financing including home equity, changes in student loan products, etc. We've already seen prepayments come in higher than originally anticipated, which has put a pinch on the residuals.

However, things have to get pretty bad pretty quickly in order for this residual interest of $380,960 for 18 years to become a big goose egg.

Also, FMD gets their 10% fee up front in cash and can go on to the next deal. Similarly, lenders get a 5% cash fee up front and can use that capital to support a new round of loans. If they do this once per quarter, that is a pretty good return for them for nominal balance sheet exposure and essentially no credit risk. The bond holder gets a LIBOR+40 security that is AAA or AA rated. It is generally a good deal for everyone, except for TERI if they are way wrong on the defaults.

NOTE: in reality, almost all of these numbers are subject to adjustments and statistical spreads. The LIBOR-based rates fluctuate up and down, and different tranches of the security carry different rates. The up-front and residual breakdown changes, the default and prepayment rate assumptions change, etc. This is really just a simplified example for discussion purposes.


Become a Complete Fool
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.