Description

• SLM is a long here.  The stock is near its lows (down from $17.00 to $12.25) in the past several months and has sold off with the broader financials complex.  Most financials are down due to yield curve compression and macro slowing, but I think SLM is less exposed to these two issues and is extraordinarily cheap at these levels.  The stock trades at only a 6.1x P/E on 2012 Street estimates and 5.3x my estimates.
• Street numbers I believe are too low because they don’t factor in a meaningful mix shift in the portfolio away from for-profit, low FICO, and non-cosigned loans, to not-for -profit, higher fico, and much higher cosigner loans.  I believe Street SLM numbers are substantially too low in 2012 and 2013.  The street is at $2 in both 2012 and 2013, I am at $2.35 in 2012 growing to $2.63 in 2013.
• To summarize, on a sum of the part basis, the FFELP run off portfolio is worth $7.  The remaining 2 businesses are private student loans and loan servicing.  Those two businesses should earn $1.07 in 2012 and $1.40 in 2013.  At an 8.5x PE for those, SLM is worth $16 today (+30%) and $19 (+55%) over the next year.  $16 and $19 only represents 7x consolidated EPS in 2012 and 2013.

SLM is composed of 3 pieces –

• FFELP Run off Book.
  • This is the business that the government insourced in July 2010.  SLM was left with its book of $142b of FFELP loans, but no longer originates these loans as the government is now the primary originator.  The cash flows from these loans are highly predictable because losses are 97-100% government guaranteed.  In addition, loan yields and funding costs are floating, so the recent flattening of the yield curve doesn’t impact the cash flows from these loans.  The $142b will run off over the next 20 years.  Currently, the servicing fees and net interest income on this portfolio generates a substantial percentage of SLM’s earnings.  In 2011 this portfolio will generate ~ $1.34 in EPS.  This runs off modestly over the next few years, $1.32 in 2012, $1.28 in 2013, $1.15 in 2014, and $1.04 in 2015.  As such this creates an earnings headwind as the portfolio winds down (although modest in the context of the overall earnings base).  Given the runoff nature, most analysts value this cash flow stream on an NPV basis.  I do the same and given the predictability and the government guaranteed nature of the loans, I use a 7% discount rate (materially above treasuries despite the government guarantee).  This generates a value for FFELP of ~$7.00/share.   Note my key assumptions for the FFELP EPS stream are a NIM of 93 bps, approximately $80m of service fees (predominantly late fees) that run off at a similar speed to the loans, 14bps of servicing expense (you can back into this number from the 3q10 conference call where the company gives the dollar amount of servicing expense), I assume the servicing expense also scales down with the loan book, I assume $20m of other annual overhead, 6 bps of loan loss provisions, and ~75m of annual corporate overhead (1/3 of corporate overhead).
  • There is approximately $250b of other FFELP loans not owned by SLM, but which SLM may look to acquire.  Because of SLM’s scale in servicing they are the best natural buyer.  In late 2010, they bought $25b of FFELP loans from Citi for $1b net of liabilities, in a deal which was 13c accretive to EPS.  As they buy more FFELP portfolios in the future there is potential for significant accretion.                          
• Private student loan business.
• SLM is the leader in the private student loan market, followed by WFC, DFS, and JPM.  BAC has discontinued the business and Citi sold their business to SLM and DFS.  JPM has deemphasized the business.  SLM has $36b in private student loans and has about 30% market share of the $9b origination market. 
• This is a high return business (1.35% ROAs) with limited competition.  The secular trends are favorable as the cost of education has been rising rapidly in recent years and the amount of public aid per borrower is limited by the government.  Loans are floating and match funded which eliminates rate risk.
• The story here is that earnings will rapidly expand in 2012 and 2013, because the portfolio is undergoing a significant positive mix shift.  There are a few pieces composing this mix shift -
• In 2008 SLM improved its underwriting standards, by stopping subprime loan issuance and increasing cosigner (parents) substantially.
• The book of legacy loans is becoming a much smaller piece of the pie as they roll through the books and are charged off. (see table below).  As you can see, dollars of legacy loans has been shrinking (far left column), and credit quality has been improving (far right column).

• During the first year after entering repayment (typically six months after graduation), is the period when most loans charge off.  In fact 60% of all charge offs happen in the first year.  After that the charge off rate drops dramatically.  (See tables below).  As you can see, in the second quarter of 2011, loans in the first year of repayment charged off at 4.5% for traditional and 18.9% for subprime.  Loans at 13+ months in repayment charged off at only 2% and 8.2% respectively. (second table).

• Losses (1-12 months in repayment)
  	Loans		Traditional	Loans		Subprime	Total	Total
  	Traditional	% loss	loss $	Subprime	% loss	loss $	loss $	loss %
  2010	8,399	5.5%	466	1,214	25.2%	306	771	8.0%
  1Q11	8,401	4.7%	395	1,110	19.0%	211	606	6.4%	<< (annualized)
  2Q11	8,707	4.5%	392	1,096	18.9%	207	599	6.1%	<< (annualized)
  Losses (13+ months in repayment)
  	Loans		Traditional	Loans		Subprime	Total	Total
  	Traditional	% loss	loss $	Subprime	% loss	loss $	loss $	loss %
  2010	14,824	2.5%	371	1,581	10.3%	163	535	3.3%
  1Q11	17,000	2.0%	342	1,608	9.5%	153	495	2.7%	<< (annualized)
  2Q11	17,427	2.0%	340	1,640	8.2%	135	475	2.5%	<< (annualized)

• Meanwhile the new originations since 2008 have been meaningfully better.  (See table below).  As you can see the cosign rate and the FICO are much higher than the old originations.  Loans without a cosigner charge off at 3x the rate of loans with a cosigner.  (second table).

  New originations
  		%	%	% for
  	($BB)	subprime	Cosigned	profit	FICO
  2009	3,177	0.0%	83.0%	NA	745
  2010	2,307	0.0%	89.0%	NA	739
  2011	2,529	0.0%	85.0%	NA	737
  Losses (1-12 months in repayment)
  			2008	2009	2010
  UG/Grad w/ co-signer			1.20%	1.90%	1.20%
  UG/Grad w/o co-signer			4.00%	5.90%	3.60%
  Loss multiplier			3.3x	3.1x	3.0x

• So, the volume of old loans is shrinking, and new loans are coming on with much lower charge offs.
• So in 2012, losses should be –
• 3.2b legacy entering repayment at 5.4% (down from 6.1% in q211 due to lower subprime mix) = $174m
• $25b of loans in 13+ months at 2.33% (down from 2.5% in q211 due to improving trends) = $574m
• ~$2b of new originations charging off at 150 bps (due to higher cosigner rate) = $30m
• For a grand total of $778m, or 2.70%.
• Consensus is at $957m, or 3.5% which doesn’t seem to make sense to me.  The difference is 22c in EPS.
• The same math gets better in 2013 b/c only 1.4b legacy loans enter repayment, in 2013 I get 2.2% NCOs or another 18c in EPS.
• Management commented at the Barclays conference recently that Charge offs in 2012 will be “sub 3%” and by 2014 should be “about 1.5%”, which validates my numbers (although the street is in wait and see mode).
• In aggregate this should take eps in the private student lending segment up from a 28c run rate in 2011, to 66c in 2012 and 92c in 2013 (post a 1/3 corporate allocation).  At an 8x P/E this segment is worth $5.30 today and $7.36 in a year
• Business services/collections business.
• This segment provides loan servicing to the Department of Education and other third parties, transaction processing services to college financial aid offices, default collections on behalf of colleges, third parties and the Department of Education, and program management for student college savings plans.  This segment has high pretax margins (in the 50% range) and requires no capital.  This segment should grow modestly for two reasons.  First, Defaults (and hence collection fees) are increasing on Dept of ED loans (which unlike SLM do not use underwriting criteria but are only need based).  Secondly, SLM is servicing only 1 years worth of Department of ED loans, and will begin servicing a new cohort of borrowers every year for the next three years.  I have this segment earning 41c in 2012 and 48c in 2013 (post 1/3 allocation of corporate overhead).  At a 9x P/E this segment is worth $3.70 today and $4.30 in a year.
• Total Valuation -

• FFELP valuation >>	$7.00	$7.00
  Implied PEx on FFELP >>	5.3x	5.5x
  	2012	2013
  Private and business services EPS >>	$1.06	$1.41
  P/Ex  on Private and services
  6.0x	$6.34	$8.47
  7.0x	$7.39	$9.88
  8.0x	$8.45	$11.29
  9.0x	$9.51	$12.70
  8.5x	$8.98	$12.00
  10.0x	$10.56	$14.12
  Total Value				% Upside
  6.0x	$13.34	$15.47		6.0x	8.9%	26.3%
  7.0x	$14.39	$16.88		7.0x	17.5%	37.8%
  8.0x	$15.45	$18.29		8.0x	26.1%	49.3%
  9.0x	$16.51	$19.70		9.0x	34.7%	60.9%
  8.5x	$15.98	$19.00		8.5x	30.4%	55.1%
  10.0x	$17.56	$21.12		10.0x	43.4%	72.4%
  Total SLM EPS >>	$2.33	$2.64
  Implied Total SLM P/ex at 8.5x…
  ...for private and services >>	6.8x	7.2x

• At a stock price of $12.25 with $7 for FFELP, you are creating the private loan and service segments for 5.0x 12 EPS  and 3.7x 2013 EPS. 

• Risks 

• In the FFELP segment, SLM pays Libor and earns a yield indexed to the 90 day commercial paper rate.  Typically there is a 10bp spread between the two.  In 2008 and 2009, Libor rose as the interbank funding markets collapsed, and CP stayed flat as the government artificially lowered CP rates via the Commercial Paper Funding Facility (CPFF).  This caused a significant (but temporary) decline in FFELP earnings.  More recently, with the downgrades of BAC and C in the CP market and lowered issuance by foreign banks, there has been a lack of CP issuance which has caused CP spreads to be volatile and tighten.  Currently the spread has been as wide as 20-25bps in the past month.  Approximately 35% of the FFELP book is tethered to 90 day CP, the rest is tied to 1 month CP (which has been much tighter to Libor than 90 day CP).  As such, Every 5 bps of widening is 3c to annual EPS.  Short of another US government intervention, I don’t foresee the spread gapping out like 08/09 but it bears watching.
• SLM has $25b of unsecured maturities.  When financial markets dislocate, wholesale borrwers get punished (which I think has contributed to SLM’s recent decline).  Versus the $25b of unsecured debt, they have $21.4b of unencumbered collateral.  In 2011, 2012 and 2013 they have 2.2b, 1.8b, and 2.3b of unsecured maturities.  Given that they have $5b of cash, generate $1b of cash per year, and have 11.5b of unencumbered private loans that could be securitized, I am not worried about them meeting these maturities.  Of note they repurchased $5b of unsecured debt in 2010 and $300m of stock in 1h11.  They also issued $2b of unsecured debt in q111.  Most recently they did an asset backed deal which refi’ed a more expensive securitization (which drove some outperformance the day of announcement).
• Political.  Where to begin?
• The White House recently introduced a plan to allow students with both Direct loans and FFELP loans to consolidate these loans.  The rationale is that by decreasing the paperwork burden from two servicers to one, it will be less confusing to borrowers and they will be less likely to default.  Insert whatever comment you want here about the sophistication/responsibility level that exists or that the government gives credit for.  Anyway, this affects approximately 18b or 13% of SLM’s FFELP book (the portion where students have FFELP and Direct).  The program is only available to current borrowers, which haircuts the eligible pool down to approximately 9% of their book.  The savings to the student is 50 bps, on an average $15,000 loan that’s a yearly savings of $75 and the paper work takes about 6 months to process from what I can tell.  What’s the take rate on a $75 savings with a bunch of paperwork from someone who needs help staying current because he can’t process two servicer loans at once? Generously, let’s say its 50% although I can’t even imagine that’s right.  That would take the pool down to 5% of the FFELP business, or 7c in EPS.  Since FFELP is valued at ~5x EPS at $7, that should be 35c to the stock (the stock ended down $1.82 or 13% that day).
• CFPB recently initiated a request for comment about student lending which sent SLM shares down 4%.  The request was not “new” news, it was mandated by the Dodd-Frank Act passed over a year ago.  The information request I believe is benign.  It relates more to disclosure than regulation of the business.  SLM has better disclosure and payment options than the rest of the industry, and the economics of a private student loan are not much different from a government one.    

Catalyst

• Earnings beats in 2012 and beyond due to portfolio mix shift.
• Reauthorization of buyback in q411.  SLM repurchased 7% of its current market cap in the first 3 quarters of 2011, and I expect they will continue to return capital aggressively in 2012.