Thursday, December 14, 2006

More on First Marblehead (FMD)

I can’t help cleaving to the old models, and will continue to hammer away at the themes touched upon in earlier dispatches, patiently awaiting further enlightenment as to the obvious flaws of my analysis. As far as offering specious arguments, we’ll let the wider audience decide on that one, since no real alternative to valuing the stock using a method with any theoretical underpinnings has been provided. The feedback on First Marblehead (FMD) suggests the temperatures haven’t cooled at all. The “debate” is reminiscent of the feverish arguments over Freddie/Fannie in the late Nineties, when voices of skepticism were shouted down or ignored. Of course we all know how that ended. In the spirit of adding to the debate rather than engaging with those with an axe to grind, I’ll put my two cents as to why this stock is more expensive than the whole universe of 80+ financials that form part of the S&P financial sector. I will not (at least for now) delve into the detailed overview of GOS and securitization margins, residuals, (marginal issues at this juncture). Nor will I address the worrisome issues advanced as to potential growth in student loan demand.

One of the most bullish analysts on First Marblehead (FMD) does appear to be the team. I won’t bother with the “minutiae” (they’re words not mine), but will attempt to wrap my arms around the stock’s valuation and what is imputed from the current stock price. Unless convinced that a different valuation framework can be valid here, using price/book for assessing fair value for FMD (and all financials, broad as that term may be to even include servicers) seems like the only sensible thing to do. The bulls don’t really address the valuation matter directly, and the 2007 (split adjusted) estimate of some $3.50 (consensus) suggest approximately 50% growth in EPS (over 2006) is expected. But I would hazard that 50% per share earnings growth may be closer to what the “market” is discounting not just in fiscal 07 (ends June), but in 08, and 09 as well. But here is a big disconnect (between Street estimates and the “owners”). Consensus expectation in 08 and 09 is for low-teens growth, though it they seem like low conviction estimates with FMD being given very little credit, as of today’s date (chances are these are low balled numbers). To their credit, report went to great lengths to detail the mechanics of FMD’s business model, but to a much lesser extent in my opinion, the economics of the model. So far, so good. There was plenty of detail, and even quite a bit of honest assessment that “it is a lending business after all”.

So what am I missing? The best lenders (bar none, except for AMEX, another short) generate low twenty ROEs. The top-tier underwriters/originators/servicers have sustained these numbers for ten years running in one of the most lucrative periods for credit businesses. But the burden of proof that FMD can sustain 40% to 50% ROEs is on them.

Investors buying into the story should look at history and the sector for perspective. The highest ROEs among all financial institution businesses (besides FMD of course), can be dug up in the footnotes of companies like CFC, WB, BSC, or GS. But a brief glance at parent company consolidated ROE's and valuations suggests 40%_plus ROE business lines (relative to these whole companies) are few and far between, and generally well hidden from competitors’ view. And for the most part even medium risk businesses with 20% ROEs would be very welcome additions to the portfolios of JPM, WFC, BAC. So does FMD’s “competitive advantage” shelter it properly? In my opinion, little risk is being priced in, while a lot of intangible value is being assigned to a business model with a relatively short track record, despite 20 years of student lending data (Come on guys; is that really worth much? COF’s army of analysts could slap something together in weeks looking for an angle here.) Perhaps excess returns (mid 40’s ROE average from 2004 thru 2006) will endure for a few more quarters or even a couple of years. By my estimation even with a decline in the ROE to 35%, it will take the company four years to earn its way into the current valuation SLM is also on the skids, with unusual relative underperformance over the last two years versus financials. Any suspicions as to why?


Christopher Caputo said...

Just curious, does your analysis of FMD's book value take into consideration the carrying value of the Teri database? Have you researched the name enough to know how much they paid for it before the private student loan biz became what it is today? I doubt anyone could come close to paying the bargain price FMD's founders managed to pay. I think the economic value being realized off of that low carrying value is skewing your assessment. If someone were to pay for the data today so that they could use a similar model, those eye popping ROE's might not be the 'unsustainable problem' you think they are.

As to your question, is it (the data) really worth that much, I think the very obvious evidence is yes...otherwise why would JPMorgan sign on through 2010? Keybank, in the student lending business for years signs up with you think they would do that if the model had no value? There has to be a reason these clients are signed up with FMD...what do you think that reason is?

As for barriers to entry, seriously, if as you say "COF’s army of analysts could slap something together in weeks looking for an angle here" why couldn't JPMorgan do that too? They have the muscle, but they have made their choice.

Surely, over time new entrants will appear, it is normal in a profitable competitive market, however I am doubtful that anyone can just swoop in and eat FMD's lunch by slapping together data in the matter of "weeks."

In addition, this article seems to contradict your assessment regarding the importance of the data. Have you talked with anyone in the industry or even the rating agencies that can offer a different view? If you have further questions about it, you might just want to shoot the author an e-mail. They are more than willing to have open dialogue and much more studied on the name than I could ever hope to be.

One other parting comment, I think a big risk to your sell thesis(?) is the very real potential of FMD to monetize some of the balance sheet residuals which they could then use to buyback shares. It appears they are seriously entertaining this idea, and if such a move was enacted it would translate into significant price appreciation.

Glen W Peterson said...

The argument being offered by the gentle poster reminds me of the shifting sands of the Sahara Desert.

Why not start a discussion of valuation by looking at the success analysts have had at predicting FMD earnings and stock price valuation. Ask yourself how successful their models have been at predicting earnings growth, margin growth, growth in the size of the securitizations, monetization of residuals in terms of the recent BBB tranches, etc.

The answer of course is they have been completely unsuccessful. Why wouldn't that make a thoughtful poster think twice about cleaving to their earnings predictions for 2007. Which are, split adjusted, at least (from a consensus standpoint)10% below the high prediction, and still a bit under my own predictions. WS has consistently argued that margins will compress, FMD's residuals are valued to highly and the competition will commoditize the sector. They have supposed that prepayment assumptions, credit assumptions, defaults, etc. are all to aggressive, even though FMD's tranche performances have clearly shown that FMD's assumptions, are in fact, conservative when compared to actual performance.

They get this by evaluating the student loan biz like the credit card biz and have theorized that the same cycle of commoditization will quite obviously occur to FMD.

This is the trap you have fallen into IMHO. Indeed, years from now, we might, in fact, see such a thing. But using that speculation to underscore a theory is specious and not based on any material fact. Show me one, I say again, one instance where someone has successfully entered this business (private student loan securitization) in this niche market in the recent past. If securitizing is just securitizing, then FMD would already have been put out of business. Something is quite obviously not quite right with the thought process.

From even the most bullish WSJ analyst (Bear Stearns)we are finally seeing estimates that will come close to determining FMD's earnings power. Without understanding the sector or it's growth prospects, how can you properly set up any kind of valuation method in terms of FMD's potential growth. The old saw, "it's a financial stock" ergo, someone has to come up with a better argument before I'll switch my valuation method," is as tiresome, as your lack of research on the business model.

I would offer the following questions for your consideration that you have quite obviously ignored.

How large is the entire student loan market? How fast is it predicted to grow? How much of the market is made up of non-federal (private) loans? What percentage of the market does Direct to Consumer type loans currently make up. How fast is the DTC loan market growing? Are college tuition costs rising and at what rate? Are matriculation rates rising. What is driving the trend in college martriculation? What is driving the trend in increased tuition costs? What percentage of college expenses are being currently financed by credit card debt? What percentage of private loans are made through Direct to Consumer outlets? How does inflation effect the demand for DTC student loans? Has the pace of average hourly earnings and or real personal savings kept up with the rate of growth of inflation in terms of real dollars. How does a declining U.S. dollar effect the demand for U.S. educational services at the graduate and post graduate level?

In short, I think you are looking at the trees and not the forest. The market is growing at a very rapid pace and has several macro drivers in place that are not going away any time soon.

I think your ROE and book value models will wallow in the mire of analysis which leads to paralysis as long as you don't have a grasp of the global, macro factors that are driving the need for more access to private student lending funds.

Simply put, you are buying growth potential here. Yes there is risk that someone else will step in and you think it's not being discouted correctly. How long do we have to wait for your theories to play out. One year, two years, a decade?

And buy the way, I agree with Chris' comments regarding the TERI data base. It was literally stolen by this investment group and is worth 5 times what they paid for it. Your lack of understanding of this is in line with the entire group of analysts.

I would also point out that SLM is not in the private student loan business. It securitizes Federal loans which is, without further legislation, a capped market for growth. Valuing one like the other is like comparing a mining stock to a hockey puck.

However, I'll give you credit for cleaving to analysts beliefs that FMD should trade at a discount to SLM. I would argue that if FMD simply monetized it's residuals, that this argument would evaporate into thin air. The fact is, that the current management simply disagrees that monetizing these assets will lead to greater shareholder value. Their tranche performance data is telling them a different story.

One thing we do agree on is the difficulty of evaluating the valuation of growth related stocks. And I do agree that if another big competitor were to get into the business that it would be very negative to FMD's near term business, especially if it were a customer like BOA. However, you are discounting the securitization business as if any Tom, Dick or Harry Investment bank on Wall Street could do this. So far, this analysis has proven to be about as correct as your estimate of the direction of FMD's stock price.


FIG Trader Intelligence said...

In response to GWP, your logic becomes rather circular when you get into discussions on the "size" of the student loan market, and stimulants to demand (matriculation rates, tuition hikes, swapping credit cards/HEL for funding tuition). Talk about speciousness... unless of course, you really believe all the stuff you are writing is relevant to the company's value today. Your demand-based "credit" model takes no position on affordability. Should we assume because everyone wants a house that everyone will get one without regards to "credit scores". Talking about falling into a trap.