Summary

First Marblehead (FMD) is a very misunderstood company that currently trades at $1.55 or about 70% of cash liquidation value.  It comes with a large embedded call option as the returning founder Dan Meyers rebuilds the business.  I call it a smoldering cash pile with an unknown quantity of (free!) dynamite underneath.  The reason it is so cheap and misunderstood is because of tarnished recent history with investors and unusually strict recent GAAP accounting rules.  The private student loan market and their previous business model of securitization were put on ice in 2007 with the financial crisis and have only recently begun to thaw.  Their GAAP financials also completely obscure the fact that they trade below cash liquidation value as excessively strict trust-consolidation rules (think Citi's SPV abuse...) confuse what shareholders own and do not own.  The stock has also dropped quite rapidly in the last month on high volume with no news.  This is possibly due to a forced liquidation but who knows, I like to buy when Mr. Market is headed for the door in a hurry.

Background

First Marblehead was founded by Dan Meyers and has been a facilitator of private student loans for over 20 years.  As college tuition grew faster than federal loan caps and inflation, a double-digit growth market for private student loans to fill the gap emerged, peaking at about $20B in 2007.  FMD securitized about 20% of these loans at the peak in 2007, only keeping a small residual piece of the trust.  Their business model was to help underwrite, package and sell the loans and take gain-on-sale revenue which was roughly 50% cash now and 50% residuals that would pay out towards the back end of the loan.  The securitized loans were guaranteed by non-profit TERI (think PMI mortgage insurance but for private student loans) so they could sell the subordinated pieces of the waterfall.  Tom Brown published a very graphical presentation on it back in 2007 which you can find online. 

The fact is, the world and the securitization market changed after 2007 and it hasn't been the same since, so they needed to change their business model to adapt.  While the securitization market has returned for most conventional loans, private student loans are pretty niche and have come back slower - Sallie Mae just did an ~$800M securitization this spring which was the first real deal since the crisis.

Dan Meyers left in 2005 and the new management ventured into some riskier ventures that cost some money during their brief tenure until Dan returned to rebuild the company in 2008.  Under Meyers, FMD's marketing budget was near $0, being more of a "nerd shop"; the new team spent $100M on marketing in 2007.  This seems to have as much logic as marketing root canals, as student loans are not exactly something people WANT to get but do so out of necessity.  It should be noted that no matter who running FMD in 2008, they would have disappeared from the market for a while as the problems were the world shifting rapidly much more than any internal issues at FMD.

Current Balance Sheet and Downside Protection

It's one thing to find a cheap option in the market; it's rare to find a marriage between a net-net and a cheap option.  Most net-net's are just liquidations where the upside is very limited.  The downside protection for FMD comes from two points (note there are 100M shares so the math is easy):

1)       The actual balance sheet that the shareholders own has a cash liquidation value between $1.60/share (ultraconservative) and $2.48/share.  The range comes from :

a.        $49M: The valuation on TMS that they just recently acquired for $49M: $1.60/sh assumes it's worth $0 and $2.48 assumes it's worth $49M.

b.       $40M: There is a tax dispute in the state of MA: FMD is fully reserved that they will have to pay $40M in back taxes.  Based on their recent losses, I am guessing they probably won't have to pay this so reduce their "other liabilities" line by $40M if they win.  If they do pay then you have only 1 year of "free" option instead of 2 years.

2)       The cash burn rate has been $10-11M per quarter for the last 2 years which is HR and general overhead (utilities, rent, etc.).

If you use the more reasonable liquidation value of $2.48/share vs today's stock price of $1.55/share you are getting a free option that they will be able to turnaround the business in the next 2 years.  I view this as being highly probable given that the founder has done this before and is highly incentivized to do so, which I will get into later.

More details on their balance sheet:

$278M cash

-$57M deposits (at bank sub)

=$221M net cash.

-$62M other liabilities (incl tax dispute reserve with MA)

=$159M most conservative net cash liquidation value.

+$49M TMS (Just bought for $49M)

+$40M tax dispute (fully reserved in 62M other liabilities)

=$248M = $2.48/share. (100M shares)

They do a good job at separating these "non-GAAP" numbers in their press releases and recent presentation (see last page from KBW presentation at: http://edg1.vcall.com/irwebsites/firstmarblehead/KBW%20Presentation%20Deck_6-8-11vFinal.pdf) though any kind of CapIQ or Bloomberg screen would show the misleading GAAP numbers.

Accounting Restatements for Trust Consolidation Obscures Downside Protection

An important point is that if you look at their GAAP numbers it looks NOTHING like this, it looks like they have a -$1B net worth.  They show up in quant screens and online news alerts under headings like "the worst debt to assets ratio of any consumer finance company" because of this.  They were forced to consolidate the student loan trusts that they securitized previously ($9B roughly) which due to the recession/high unemployment are being marked at a $1B+ paper loss currently.  This has no recourse whatsoever to them and is due to excessively strict consolidation rules that were adopted after all the off-balance-sheet abuse that took place up to 2008.

More Downside talk: Where the cash is from

The cash they currently have comes from two places: some from Goldman Sachs PE which purchased shares at ~$11/sh right before the stock nosedived in early 2008 (there were not many people that thought the whole securitization market could shut down for 3+ years!).  The other more significant source was tax refunds that Dan Meyers got once he returned.  The "other 50%" of their revenue I mentioned in the old model was residuals on the student loans that were back loaded to a 10-20 year student loan repayment.  These are like zero-coupon bonds and the phantom interest was taxable and they were paying real cash taxes on these even through 2008.

The problem was that this was like buying long dated zero coupon bonds on a company that goes bankrupt right before it repays - even if you paid zero for it you could have a big negative NPV because of the taxes!  Fortunately this was shifted to being someone else's problem by selling the residuals for $0, thus realizing the loss and reclaiming the taxes paid which totaled over $200M. 

This is technically still an open issue by the IRS - they are slow at closing audits - but given the straightforward nature of the refund it seems very unlikely that the IRS would invalidate the refund.  Nevertheless this is the only and most severe risk to the downside protection from the balance sheet that I am aware of so it merits consideration.

The Private Student Loan Market

Sallie Mae has a good primer on the private student loan market on their website (https://www1.salliemae.com/NR/rdonlyres/50F355EE-8FA7-49FA-AABF-D4A4B507A89C/14544/PrivateEducationLoanABSPrimer.pdf).  Private student loans fill the gap between the federal grants; federal loans and family spending (either savings or financed via home equity loans) and the rising cost of tuition.  In 2010-11 this gap is estimated to be $9B ($109B Fed grants + $103B Fed loans + $189 family contributions - $410B tuition).  2007 was the peak year at about $21B and it had been growing at 20% per year for the decade leading up to 2007.  It is currently supply constrained (as opposed to the majority of business and consumer loans currently which are demand-constrained due to the balance-sheet recession we are in).  This supply constrained market is evident in two ways.  One is that since 2007 the numbers of players in the industry has consolidated as most banks have exited the business.  The other observation is the high spreads (8-9% interest) being earned by lenders to high credit quality borrowers (760+ FICOs).

The overall market risk is that tuition stops rising, federal grants/loans increase significantly, or college enrollment declines.  This would be a significant break from historical trends that is certainly possible but there doesn't seem to be much evidence for this happening.  If you're worried about college enrollment dropping off, I'd consider buying ZAGG as a hedge, as this stock will likely continue to perform well if the education level of the population significantly declines.  Note that this whole write up doesn't overlap with the controversy around Federal student loans and the for-profit education sector, that's an entirely separate set of issues.  FMD is not a subprime student lender or anything along those lines.

One interesting point is that FMD is the only real dedicated player left in the market as the crisis wiped out a huge number of smaller competitors like MyRichUncle and the departments at smaller banks.  The main remaining players are WFC, SLM, and Discover (Citi sold their student loan division to Discover).

Getting Fundamental: Why does FMD have a reason to exist?

Previously FMD had a similar business model to the mortgage securitizers such as Delta Financial, Novastar, Countrywide, etc. though they never took quite as much credit risk from warehousing the loans prior to securitization.  Those businesses had no reason to exist except perhaps some better marketing in a given location.

Private student loans are different because they are such an unusual type of loan.  Unlike auto and house loans there is no collateral to borrow against and unlike credit cards there is no current income to measure against.  The repayment capacity is instead based on future increased income gained from the education.  This can be measured but requires much different models than banks are used to and so most are not equipped to underwrite this type of risk intelligently.  FMD is effectively an outsourced underwriter for student loans.  They have been in the business for 20+ years and have a payment history database for a wide range of students and so know how to measure credit quality more accurately than conventional bank metrics like FICO score.

Rebuilding the Business: Monogram

FMD has spent the last 2 years retooling their business model to work in the absence of securitization markets.  Their new product, Monogram, is designed as a partner-lending "cradle-to-grave" model where they act as specialized designers, underwriters, and portfolio managers of student loans on behalf of partner banks.  The loans would then be held on the bank's balance sheet and FMD would earn a significant portion of the spread.  In exchange, they also have to put up a first loss piece for the banks which they earn back over time.  They also recently announced that they are keeping their small UFSB S&L and will originate off its balance sheet.  Currently UFSB is capitalized to be able to offer roughly $35M of student loans so I view this more as an educational and testing tool for FMD more than the core business.  They can definitely get a higher ROE with more of a service model and I expect they will focus on that though it can be useful to understand your customers better when you can "pretend" to be one with your own small bank.

The product concept is finished and this season they are proving the concept.  They have a small $200M partnership with SunTrust, $75M with Kinecta FCU, and over 200 partner schools and we will know how the "trial" went by this fall.  The real growth would start next summer at the earliest if they can sign up more partners after this lending season.  If things go well and they successfully facilitated a lot of loans, expect growth to be rapid next year, if not then we have 2 years of free cash burn at the current stock price to wait for them to learn a better model.

Partner Lending Unit Economics: Very Back of the Envelope

While the details of the numbers haven't been fully disclosed yet, my rough guess of the numbers would be something like this:

For $100M of loans, FMD puts up $10M in first loss capital.  The loans are made at 8% interest which is split 5% to the bank, 3% to FMD and have an average term of 15 years.  They also get some fee up front, say 2% of principal.

Total revenue is $2M up front and ~$25M in interest over the 15 year life of the loan (using straight line amortization), call that $14M at a NPV10 so $16M total revenue per $100M loans.  This is a really rough estimate and assumes minimal defaults but gives the idea of how they make money.

They had 20% of the private market in 2007 before the crisis.  If they could achieve 15% of the current $9B market that would be $1.35B/year of loans facilitated * 16M NPV revs per $100M loans facilitated = ~$200M in revenues.  At 30% net margins that works out to $60M or $0.60/share of net income.  This doesn't include the $135M they would have to put up as the first loss piece but depending on the 10% assumption I don't think the cash they need to front will be the main bottleneck.

Note: the net margins seem high for a financial company but that's because their revenue is basically the net interest margin - look at bank and see what their net margins look like if you treat net interest margin as revenue, it's high...  FMD had 41% net margins in '05-'07.

I think the assumptions here are generally conservative, especially the $9B private student loan market, which will probably not stay that low for very long.  I built in credit risk to the 3% number (I think it's probably more like 4% so took off 25% of revenue for defaults) though that could be more of an issue once the cycle improves and capital returns.  15% market share could be low too considering they had 20% pre-crisis and there are far fewer competitors now.

The bottom line though is that I don't know what the upside looks like exactly, it's very fuzzy.  If the securitization market comes back faster, they could make far more than for the partner lending model since volume can be much higher.  There are scenarios where you could see the stock at $6 or at $25, the main point is that it's currently below net cash, so there's no downside for at least 2 years as they burn that cash for liquidation value to decrease to the current stock price.

Dan Meyers: The Comeback CEO

I have learned, perhaps too slowly, that with small companies especially in dynamic situations, the leader is a huge factor.  I don't know what the private student loan market will look like in 2-3 years and frankly, Dan doesn't either.  But he has every interest and the capability to be creative and adapt to the new environment.  I would not be excited about this slow burning cash-pile of a company if not for someone whose entire career has been about student finance. 

Dan owns 7.6% of the common stock and also received 2M options each at $6, $12, and $16 as compensation for returning.  He had potential financing to pursue another option, to create a "Second Marblehead" (he actually trademarked it) but chose to return to FMD because of the huge proprietary database they have as well as the moral consideration of finishing what he started.  Compensation was set at $1/yr along with the out of the money options (stock was at ~$2 then) until FMD was profitable but they gave him a $800k/yr salary after he got the tax refunds, which was impressive so I'm ok with that.

Summary

FMD is a very unusual financial services company.  It trades significantly below net cash liquidation value but doesn't appear to because of weird recent GAAP consolidation rules.  Unlike most net-net's it also has huge potential upside if they can rebuild their business and the perfect CEO (who has done this before) has the reigns and incentives to do so.  The catalyst will be seeing the initial results of the partner lending model and their own experience with UFSB this fall after the summer student lending season.  If it is successful, expect them to sign new partners at a rapid clip.  If it's not successful, then it'll just take longer - at $1.55, you have 2 years of "free" cash burn before liquidation value gets as low as the current stock price.

I would ask that you please not bid up the stock above $2.50 because then I actually have to start figuring out how to value the embedded option instead of just getting it for free.  It's the summer and I'd rather not have to until the fall when I have some more information.  Between here at $1.55 and $2.50 it's below liquidation value so it's a no-brainer on the downside protection.

P.S. Please tell your quant friends about FMD but not about this write up so their computer shorts it more based on the input GAAP numbers.