Description

Investors have largely abandoned financial stocks over fears of an extended period of low interest rates, loan losses during the pandemic, and an accounting change to loan loss reserving methodology. We see these risks as fully discounted in current valuations and remind ourselves that the next five years’ free cash flow equals only 20% of a company’s value, on average.  Free cash flow beyond year five drives 80% of the value, making near-term prospects of little consequence to true long-term value investors.   

SLM (commonly known as Sallie Mae; originally the Student Loan Marketing Association) is the largest originator of private student loans in the U.S. It maintains a number of attractive characteristics including dominant market share (52%), a healthy ROE (~18%), mid-teens growth in EPS and TBV per share, a high quality loan portfolio (avg. FICO score of 746), a strong capital position (13.7% RBC ratio), an aggressive stock buyback at attractive prices (~17% of shares in 2020), and most importantly a compelling valuation (4.7x normalized earnings and 1.0x TBV).  

Concerns include fears of a Democratic sweep this November (“free” college, loan forgiveness, higher student lending caps), aggressive competition from refi lenders such as Sofi, a depressed interest rate environment pressuring net interest margins, and loan losses during the pandemic. 

That’s the elevator pitch.

Company Description

Historically, SLM originated mostly government guaranteed loans through the Federal Family Education Loan Program (FFELP).  These loans were originated and funded by the private sector but fully guaranteed by the federal government as a way of lowering student borrowing costs.  In 2010 the Obama administration ended the FFELP and took government-backed student lending in-house through the Federal Direct Lending Program (FDLP).  As a result, the U.S. government now directly owns and services $1.5T of student loans.  (psss: it’s not going well).  

The demand for private student loans arises due to the gap that exists between the cost of a college education and the lending limits on government student loans.  The cost of a college education varies widely, however, the average approximates $30,000 annually, or $120,000 over four years, according to the National Center for Education Statistics (yes, that one).  The current loan limit under the FDLP program is approximately $9,000 a year (in a range of $5,500-$12,500 depending on what year the student is in and dependency status).  There is also a lifetime cap of $31,000.  This leaves the average student with a funding need of $89,000 over four years.  Most of this gap will be filled by mom & dad, however, 58% of students still graduate with a student loan (averaging $29,000).  Happy graduation!  

SLM’s total addressable market (TAM) in private student loans is roughly $11B in annual originations and $125B total outstandings.  The TAM is the difference between the cost of a college education and the collective funds available to the student via the FDLP, grants, and family contributions.  SLM offers a loan to the student to help bridge their funding gap, not backed by the government, but almost always backed by parents.  SLM has relationships with financial aid offices at 2,400 U.S. colleges & universities and are on the preferred lender list for 98% of them.  This network cannot be easily duplicated.  Private student lenders become part of the financial infrastructure of the school and integrated into its disbursement policy.  They have the largest salesforce in the industry supporting financial aid offices.  Also, schools generally only want the most recognizable and reputable lenders “representing” them on their platforms to avoid unnecessary scandal.  e.g. They don’t want another Wells Fargo situation subjecting their students to getting duped.  Schools receive no benefit from using a particular lender, making brand and name recognition important attributes.  

As mentioned earlier, SLM stopped originating government-back student loans in 2010 after the FFELP ended.  In 2014, SLM spun-off its government guaranteed loan book into a newly public company called Navient.  While this loan portfolio carried no credit risk, it was essentially in run-off and prevented SLM from allowing the growth of its private student lending business to shine.  Long story-short, the company no longer originates or owns almost any government guaranteed loans.  It is 100% unsecured consumer credit and takes on 100% of the credit risk.  

While SLM carries the inherent risks associated with unsecured consumer lending, it is differentiated from many other types of unsecured consumer lenders such as credit card companies due to its higher average credit quality, longer maturities, and lower loan rates. Importantly, student loans are not dischargeable in bankruptcy unlike other forms of consumer credit.

Market share in private student lending is very consolidated.  SLM dominates with 52%.  Discover is a distant second with 19%.  Wells Fargo is third but has been ceding share in recent years given their issues and asset cap.  Their market share erosion is set to accelerate as they recently stopped taking applications.  With Wells out of the market, SLM is making a concerted effort to gain share; however, we expect their share gains to be offset by losses to consolidation lenders such as Sofi and Earnest.  More on this later.  Approximate market shares:

SLM                52%

Discover          19%

Wells Fargo     15%

Citizens Fin.      7%

Other               6%

The typical undergraduate loan carries reasonable borrowing costs, no origination fees, and a term of seven years (ranging from 5-15 years).  Repayment is typically deferred until graduation.  About half their loans are fixed rate, half variable.  Customer satisfaction recently reached an all-time high of 80%.

Secular Growth

SLM’s organic growth is a key part of the investment thesis that we think many investors miss.  Only modest growth in originations leads to much higher growth in assets and book value due to the small starting size of the loan book (after the 2014 spin-off of Navient) relative to the dominant market share it has in originations.  At year-end SLM owned only ~16% of private student loans but has 53% market share of originations!  This is the linchpin to its organic growth and has fueled 18% growth in tangible book value per share since the spin.  All organic.

                                  2015     2016    2017    2018    2019

Originations (B’s)           $4.3     $4.7     $4.8    $5.3     $5.6

Growth in originations      8%      8%       3%     11%      6%

Loans outstanding        $10.5   $14.1   $17.2   $20.2   $22.9

Loan growth                    n/a     34%    22%     17%     13%

We expect healthy growth in assets and book value to continue for the next several years.  Pre-pandemic they guided to long-term growth in annual originations of 6%, which should drive low-double digit growth in assets (after prepayments, consolidations, and amortizations).  Loan growth assuming no asset sales and continued healthy losses to consolidation lenders should look something like this:

                                            2020e     2021e     2022e     2023e     2024e

Beginning Loans Outs.           $19.8B    $22.1B    $24.8B   $27.3B    $29.7B

+Originations (B’s)                  $5.0B      $6.0B     $6.2B     $6.5B      $6.7B

Growth in originations                N/A         20%        4%         4%         4%

-Consolidations                       $1.3B      $1.7B      $1.9B    $2.1B      $2.4B

-Paydowns, other                    $1.4B      $1.6B      $1.8B    $2.0B      $2.2B

=Ending Loans Out.               $22.1B    $24.8B    $27.3B   $29.7B   $31.8B

Loan growth                               n/a        12%        10%        9%        7%

As you can see, even assuming continued aggressive loan consolidation activity and origination growth below company guidance, revenue grows ~10% CAGR over the next five years.  Aggressive share repurchases will drive EPS and TBV per share growth even higher.  Note though - the table above should only be considered illustrative because SLM is planning on selling ~$3B in assets annually to fund an aggressive stock buyback program.  Guidance is for the loan book to stay flat via $3B in annual net loan growth and $3B in annual asset sales.  Nonetheless, you can see the dynamics at play between originations growth and asset growth.   

Asset Quality and Capital Strength

SLMs relatively high asset quality distinguishes it from many other lenders.  Their focus on middle income borrowers with stable credit histories has been consistent in the decade since the financial crisis.  Co-signers with high FICO scores are a critical component, as the borrower is an unemployed student with little/no income and credit history. 

                                                2015     2016    2017    2018    2019

% of loans cosigned                    90%      90%     89%    88%     87%

Weighted Avg. FICO at approval    749       748      747      746      746

The percentage of loans co-signed has dipped the last couple of years due to the launch of the “parent loan” which naturally does not need a co-signer.  The few loans that do not have co-signers are typically graduate students in their 30’s with an established credit history.

As mentioned earlier, student loans are not dischargeable in bankruptcy unless the borrower can prove repayment imposes undue hardship, further enhancing the quality of their loans.  In addition, SLM originates almost no for-profit college loans.  Loan approval rates are in the mid-40%’s.  

The quality of SLM’s private student loans has improved dramatically over the last decade.  They strengthened their lending profile during the financial crisis and have maintained this discipline.

                                                 2009    2019

% of portfolio with co-signer         57%    87%

Average FICO score                       713     746

SLM’s private student loans had a peak net charge-off rate of 2.7% during the financial crisis.  Management does not expect annual net charge-offs to exceed 2.5% this cycle.                                                                                                                                                                                                                                    

                                                              2009   2015     2016       2017    2018      2019     2020-21e

Allowance for loan losses as a % of loans   2.7%   .77%   1.05%    1.00%  1.10%    1.54%         2.50%

SLM ended 2Q20 with ~$20B in student loans outstanding, loan loss reserves of $1.9B, and an additional $1.6B in capital above “well-capitalized”.  Capital ratios are strong with RBC of 13.7% and CET1 of 12.4%.  GAAP equity plus loan loss reserves are a robust 15.7%.  In our opinion SLM has reserved conservative during the pandemic.  SLM adopted Moody’s economic model to estimate necessary reserve levels this year.  As of 2Q this model was estimating unemployment at 10.6% by 4Q21.  Given current unemployment of 8.4% we expect SLM’s aggressive reserving to ease.  (FYI, reserving increased in 2019 due to the impact declining interest rates had on TDR reserves.  That should normalize as rates stabilize.)

Relative to most financials, the balance sheet is pretty simple.  94% of assets are student loans.  SLM funds its loans primarily with deposits, both directly thru an online bank as well as brokered.  With a loan/deposit ratio of 89% it remains heavily reliant on high cost deposits.  With the competition for retail deposits as intense as ever, funding costs will likely remain high.  SLM also uses longer-term funding via an asset-backed facility (ABF) and a $400MM perpetual preferred.

CECL

Current Expected Credit Losses (commonly known as CECL) is a new accounting standard that was issued by the Financial Accounting Standards Board (FASB) to replace the current Allowance for Loan and Lease Losses (ALLL) accounting standard.  It went into effect earlier this year.  CECL requires SLM to recognize credit losses upfront for the life of the loan. It accelerates the recognition of future losses.  SLM is more heavily impacted by this accounting change than other unsecured consumer lenders because the life of a student loan is much longer (seven years on average).  While it does not change the economics of the business much SLM has to take a much bigger upfront capital charge under CECL to recognize life of loan losses.  In short, CECL will penalize growth.

(For those familiar with the software industry, you can think of CECL like what a software company goes through when it changes its revenue recognition from maintenance to subscription.  License revenue goes from being recognized all upfront to being recognized over the term of the contract.  Except in SLM’s case it is not revenue, it is costs.  And instead of shifting from maintenance to subscription they are shifting from subscription to maintenance (recognizing all expected credit costs upfront rather than spreading them out over the life of the loan). 

SLM transitioned to CECL in 1Q20, increasing the allowance for loan losses by $1.1 billion.  Due to the hefty capital set aside under CECL, management is now using Tier 1 capital plus loan loss reserves as a key measure of capital strength.  

                                Without CECL    With CECL

Total Reserves                  1.7%             7.0%

Tier One Capital              12.7%           11.6%

Total                              14.4%            18.6%

Loan loss reserves currently equals a life-of-loan loss of >11%.  The average life of a loan is seven years.  Thus 11% in life-of-loan reserves divided by a seven-year average life equals an annual reserve of 1.57% vs. the trailing five-year annual loan loss allowance of .97%.  SLM expects an average life-of-loan default rate of 9%, equal to 1.28% annually.  They look well reserved for the high losses expected in 2020-21 under the new and old methodology.  

Despite having no impact on cash flow, GAAP financials will be significantly distorted by CECL.  Reported tangible book value fell by 34% upon adoption.  Reported ROE will exceed 40%.  Reported EPS will be weakest when loan production is the strongest due to the heavy upfront reserves that must be set aside for new origination.  Those are the distorted optics.  Economically, CECL will increase the capital required to make a loan, lowering SLM’s ROE.  Management has commented that the additional capital required under CECL will reduce their ROE by “a couple points” and “coming out of CECL we will still have high teens ROEs”.  Using 2019’s ROE of 20.7% as the base implies an economic ROE post-CECL of ~18.5-19.0%. 

(On a side note, while SLM is required to recognize estimated credit losses for the life of the loan, they are not allowed to mark assets up to fair value.  Due to the high quality if their loans, SLM estimates the fair value of their loan portfolio is $2.6B higher than stated value, net of the fair value of its deposits.  (It’s in the notes section of their 10k).  That’s over $7 per share and would double book value.  The recent $3B asset sale generated over $300MM in gain-on-sale profits, offering some validation their assets are understated on the balance sheet. 

Pandemic Impact 

COVID-19 is expected to impact originations by ~$1B this year, or ~17%, from a planned $6B to $5B.  Lower prepayments and consolidation activity are helping to offset lower originations.  Competition for originations and refi’s has eased.  Also, the average loan size is increasing due to lower state subsidies and family contributions.  We expect originations to recover to normal levels in the 2021 school year.

In April SLM issued disaster forbearance to a large number of borrowers.  Loans in forbearance peaked in the mid-teens % and are already down to <6% of the loan portfolio.  They continue to trend down.  The average FICO score of those in forbearance is 727.  30-days delinquencies were 2.7% at the end of 2Q20.  As forbearance ends delinquencies are expected to rise to 4-5%.  Charge-offs are being delayed by forbearance but should rise sharply in the coming quarters.  SLM currently expects charge-offs of 1.7% in 2020, rising to 2.5% in 2021 using the Fed’s CCAR severely adverse stress test.  This is consistent with what SLM experienced during the financial crisis when charge-offs peaked at 2.7%. 

Valuation

We use three approaches in valuing SLM: A P/E on normalized earnings, a multiple of TBV, and a DCF.  We start with a base case summary EPS model, using normalized NIM and loss provisions.

Normalized base case         2019     2020N*       Notes

Interest earning assets       $28.2B     $30.2B        They have $6B in cash & $1.5B in ST bonds

*Net interest margin            5.76%     5.50%        5-yr avg: 5.7%, but 2020 ~4.9%

=Net interest income          $1,623     $1,661

-Provision for loan losses        $354       $295        $295 was original pre-covid 2020 guidance 

Provision as % of loans         1.54%      1.4%        5-yr avg: .964%, peak ~2.5% in 2021

=NII after provisions           $1,269     $1,366

+Non-interest income              $49           $0        excluding gain on sale

-Non-interest expenses          $574        $580

=Earnings before taxes          $744        $786

-Income tax                          $165        $197

=Net Income                         $579       $589

Preferred dividends                  $17         $17

Net income to common           $562       $572

EPS                                      $1.30      $1.60

# shares                                  431        358

*normalized for credit cycle, rates, and CECL

Normalized 2020 EPS of $1.60 compares to 2021 and 2022 consensus EPS of $1.44 and $1.72.  Normalized provisions are modeled fairly conservatively at 1.4%.  From 2013-2018 provisions ranged from .77-1.10%, then moved higher in 2019 to 1.54% due to the temporary impact declining rates have on TDR values.  Provisions were expected to move down in 2020 prior to COVID.  The most debatable model assumption is likely the NIM.  NIM has been under a lot of pressure due to the decline in rates this year.  It has averaged 5.75% over the last five years, which may be a good proxy for modeling the out years but not the next year or two.  Management expects 2020 NIM to average 4.9%.  If current historically low rates persist indefinitely then normalized NIM moves permanently lower.  A move to a 4.75% NIM would bring normalized EPS down to $1.30.  Under this scenario we would argue SLM is still cheap at only 5.8x.  In our base case normalized model SLM is currently trading at only 4.7x normalized EPS.  We believe 12-13x is more appropriate given their growth and ROE dynamics.  We considered their historical multiples as well as the multiple implied by a DCF (both are below).  At 12x $1.60 in normalized EPS, SLM’s intrinsic value equals $19.20, upside of 156%.

Next, we ran a DCF which assumes an 18% ROE, 4% secular growth, and a 10% discount rate, which generates a value of $16.12, 115% upside. 

Increasing the discount rate by 0.5% lowers the upside by ~15%.  We also considered ROE scenarios ranging from 16% to 19%.  Using the assumptions above but changing the ROE, we get the following valuations:

ROE     Fair value   Upside

16%        $13.82      84%

17%        $14.97    100%

18%        $16.12    115%

19%        $17.28    130%

As mentioned earlier, book value is heavily distorted by CECL.  Ex-CECL reserves TBV is ~$7.38.  The DCF implied multiple of 2.3x is fairly consistent with its historical trading multiple seen below (the black line denotes the adoption of CECL which depressed reported TBV).  Using a multiple of only 2.0x we arrive at a fair value of $14.76, or 97% upside.  For those that want to use GAAP book value including the CECL adjustment, recognize that GAAP book value will be growing at an extremely high rate for several years due to the CECL-induced accounting distortions.    

Share Buyback

SLM recently implemented a new strategy of selling approximately $3B of loans per year and using the gain-on-sale and the associated released capital to fund a large share repurchase program.  This should keep the loan book flat at around $20B but lead to significant accretion to the economic value of the company.  They are currently executing a $525MM ASR which is expected to run through year-end.  Taking advantage of today’s low price via a large stock buyback is rare in Corporate America today and speaks to their confidence in the quality of their assets and capital strength.  They expect to retire ~17% of shares this year assuming current prices and have already repurchased 11%.

Three-year guidance given at year-end 2019 was as follows:

• Loan sales $3B per year
• Balance sheet growth flat due to loan sales
• Stock buyback Up to $1.4B cumulatively with loan sale proceeds

Acquisition Candidate

SLM would make an excellent acquisition for a bank.  It would provide high yielding assets with good credit quality, an opportunity to cut funding costs in half, and drive material expense savings.  We find that monoline lenders such as SLM are often acquired by larger diversified lenders.  ValueAct owns 8.6% of SLM and is the largest non-passive shareholder.  In our opinion, their presence further increases the likelihood SLM is acquired. 

Risks

Democratic Sweep

The Dems support for making college more affordable for more people is well known.  Biden has largely adopted Bernie’s proposals which include “free” college and student debt forgiveness.  Short of a Democratic sweep though, they don’t stand much of a chance considering their high cost.  Maybe a more realistic threat would be an increase in the lending cap.  A significant increase in FDLP’s lifetime lending cap would reduce originations proportionately and crimp growth.  As noted earlier, it hasn’t been increased since 2008.  Since then the cost of a college degree has increased ~4% annually.  Historically increases in the lending cap have only kept up with inflation, although recently it hasn’t even done that.  The government doesn’t make it easy though – high default rates and poor collection efforts on the $1.5T in government student loans make it difficult to argue for expanding the program.  

Here is what management has to say about the matter on their 2Q20 call:

“During the previous Presidential campaign, both Senators Clinton and Sanders campaigned on this approach and it has been adopted by the Biden campaign. Interestingly, it has already been implemented in several states with New York's Excelsior program being the most prominent example. Under this gold standard program, New York State residents who earn less than $125,000 receive free tuition if they enroll full time, maintain a minimum GPA, and make post-graduation New York residency commitments. It is important to point out that statewide programs already exist in 19 states and 18 additional states have programs with variations on this theme. In the first year of the New York program, our originations in the SUNY system went down 3% and have grown every year since then. We also recognize that borrowers sometimes get overextended and need relief. To that end, federal debt forgiveness and bankruptcy reforms are two other common proposals. The cost of forgiving all federal student loans is approximately $1.5 trillion, making it hard to imagine that such a policy would have a life beyond the campaign trail. This is especially true given other Democratic priorities such as clean energy, infrastructure, and healthcare access. However, one might envision programs more focused on those individuals experiencing deep financial difficulty. These targeted proposals would likely have little impact on our business as they would be directed at higher risk customers who are already struggling to make their payments and likely expected to default. As for bankruptcy, we have long been supportive of prospectively allowing the discharge of student loan debt in bankruptcy provided there is some reasonable period of post-graduation payment to guard against the moral hazard of declaring bankruptcy simply to discharge student loans.”

Wells Fargo recently on the subject:

“We take a look at an interesting study by the American Enterprise Institute (AEI) that looks at the impact of free tuition proposals on student debt. We think the topic is quite timely as “free college” has become a favorite talking point for politicians and the media. The analysis focuses on the impact of proposals of free tuition at public colleges for in-state students. In summary the AEI analysis suggests that free tuition proposals would only reduce total federal student debt by a mere -15%.

…The AEI study focuses only on the federal student loan program, not private student loans. Our industry contacts suggest that students at private institutions are about twice as likely to have private student loans as in-state public students and the amounts they borrow are 3x larger. So, the reduction for the private student loan market would likely be pretty minimal under the free college proposals.”

Our base case is gridlock and little movement.

Competition from Consolidation Loans 

Outside of the government-backed FDLP, perhaps the greatest competition comes from the loan consolidation market.  When the loan is in repayment status competitors frequently move in to refi student loans at a lower rate.  Sofi and Earnest have been aggressive in taking share in this market.  SLM lost $1.5B of its loan book to consolidatoins last year which is relatively significant on a ~$20B loan book.  The company has been working hard over the last couple of years to identify the most at-risk loans and have put measures in place to improve loan retention.  The pandemic appears to have slowed consolidation activity - a short-term positive. 

Part of the answer to this issue appears to be to sell part of their loan book each year, which we favor.  This allows SLM to recognize the fair value for their assets, take capital out of the business, and avoid losing the loan to refi lenders down the road. 

Interest Rates 

The balance sheet is modestly asset sensitive.  Lower rates could further pressure NIMs, as well as increase consolidation activity.  Although “lower for longer” seems firmly consensus today, thus leaving only upside if rates rise.  With interest rates the lowest in the history of the world and the Fed at the zero-bound, we’ll take our chances they don’t go lower.

Credit Quality

There could always be surprises.  Loan losses will track the economy.

Catalyst

Normalizing NIM and credit costs.

Acquisition.

Value.