Description

GSKY is a recent Fintech (May’18) IPO. Market cap is $3.8bn and the stock is fairly liquid and easy to borrow.

The bull case on GSKY is that it is unique fintech (profitable) growing 30-40%, with minimal credit risk, run by visionary CEO who is well incentivized as he owns 1/3 of the company. Just like LendingClub and OnDeck, GSKY is largely covered by technology analysts, who are mostly very bullish and think that a 30+x PE is reasonable because it grows faster than most payments companies and could be the ‘next Square’. As we will show below, the bull case is fundamentally flawed as GSKY retains significant credit risk and its current profitability is massively overstated due to quirky accounting.

Let’s start with the management. The CEO’s previous company was “technology enabled” bank called Rockbridge.

• Rockbridge was supposed to be a commercial lender with unique tech services for small businesses. You can easily find the FDIC Inspector General report from after when Rockbridge collapsed- FDIC took around $100mm in losses on $290mm in deposits.
• Inspector General report, as we understand it, says that they told banking regulators that commercial lending was going to be the majority of the business, but in reality, they put 2/3 of the balance sheet in much riskier construction and CRE loans. The government actually sued the GSKY CEO and his Rockbridge cofounders and they settled for around $5mm:

             https://www.fdic.gov/about/freedom/plsa/ga_rockbridgecommercialbankburrlehwald.pdf

The comparison with Square is also deeply flawed. Square IPO’d at a $4bn market cap- and SQ invested well over $1bn on technology and S&M in the preceding 3 years to earn this valuation- while GSKY IPO’d at a $4bn valuation and appears to have invested under $30mm in technology and S&M in the trailing 3 years.

The story that the street believe is a simple one-  that GSKY helps banks originate home improvement loans via their network of contractors for which they receive an upfront fee from the contractor and there are some minor expenses tied to origination. Since the loans sit with the partner bank, the perception is that GSKY does not take credit risk (or is, at worst, exposed up to the 1% of loan amount that is set aside in an escrow) and is a high ROIC and high margin (EBITDA margin of 44%) origination business. We believe that the reality is quite different.

First, let’s investigate the accounting of revenue and associated costs where we believe that a mismatch between the timing of revenue and expense recognition is boosting short-term profitability. If you carefully read through the S-1, you notice that the COGS line includes a counter-item called “Fair value change in FCR liability”, which in turn has a counter-item called “FCR related receipts”, which in turn is effectively equal to what the banks receive from the borrower, less a guaranteed interest to the bank. In 1Q18 the “Receipts” amount is $28mm while reported EBITDA is $27mm.

GreenSky books the entire upfront commission it receives from the contractor as revenue while it recognizes associated credit losses over the term of the loan. These losses flow through the FCR related receipts line of COGS. Since the originated loans can be up to 12 years in duration, the mismatch between the timing of recognizing revenue vs associated costs boosts current period profitability. Management is guiding to flat “EBITDA” margins from here, even though margins have compressed by 1,200 bps in the past 2 years- while revenue growth decelerated. As the stock of loans, on which the company must eventually absorb credit costs via P&L, builds up, we believe the EBITDA margin will compress dramatically.

While many bulls neglect credit risk altogether, we believe that GSKY appears to be to guaranteeing their partner banks a specific spread over floating interest rate benchmark - and every dollar of credit costs that happens on these loans flows into GreenSky’s P&L. Unlike most Financials we know of, GSKY neither provisions for future credit losses nor takes a credit-related NPV adjustment to revenue like in gain-on-sale accounting. In fact, you can search for the word “greensky” in earnings call transcripts of their partner banks, e.g., FITB, and you will see the bank partners saying they are happy taking GSKY loans because they provide ~500bps of first loss protection to the bank.

·         Several sell-side reports do consider a hypothetical credit sensitivity, which they calculate to be 1200-1400 bps of margin for 100 bps of credit deterioration- but seem to then assume that the max loss is capped at 100 bps and can only be triggered in a significant macro deterioration, and thus GreenSky is “almost” not sensitive to credit

·         We believe the “100bps max loss” number that bulls focus on is the amount of money Greensky puts into escrow to effectively serve as collateral against counterparty exposure by the partner bank to GreenSky- that is, as credit losses go up, Greensky takes losses into own P&L until earnings go negative, and then on top of that the partner banks are entitled to the “100 bps” escrow money as extra protection.

o   To over-simplify our understanding, the Greensky shareholders are on the hook for all of the credit losses until the company is out of business, and then Greensky creditors are on the hook for that 100bps extra payable to the bank.

The company appears to tout its promotional loans (e.g., no interest for 12 months if paid back before the end of the 12-month period) as its key flagship product and a big driver of its profitability- and the bulls appear to derive significant comfort in the idea that promo borrowers are not likely to default in that 12-month period if underwriting was done well and macro is good, so they imagine a nearly credit-risk free highly-profitable fee income stream. We will get to credit details later- although it is worth highlighting that even the super-prime borrowers with the best-underwritten loans in the best possible macro still are likely to have a floor of charge-offs well over 100bps due to factors like death and divorce, that is just the nature of unsecured consumer lending. But let’s ignore credit for now and look at the rest of the economics. You can find a number of sheets publicly available on the internet by searching for them on google- you will find both current and historic sheets from different segments of home improvement. In short, we believe the “flagship product” to be unable to cover their share of OpEx, even before any credit costs.

http://static1.squarespace.com/static/5763a7c359cc68887f3e84d3/t/579ffddf5016e13b82ad8d08/1470103007637/greensky-products-sheet-072016.pdf

https://media.egia.org/documents/geosmart/installmentPlus/installmentPlusRateSheet.pdf

·         In round numbers, we believe GSKY gets ~650bps of fee revenue, gives ~400bps to the bank, spends ~120bps on origination and servicing and another ~170bps in allocated OpEx- leaving nothing resembling a 44% EBITDA margin even before any credit costs are incurred.

So how does the company manage to report a 44% EBITDA margin? Here is how we look at the book. 

·         The majority of the book is, in fact, long-term unsecured consumer loans- per company disclosures only ~40% of the mix is promotional, and the remainder are fixed-APR 7-12 year term loans

o   We will skip the detailed math for brevity, but per discussion above, the vast majority of revenue is recognized upfront, and the book is growing 40+%, leading to an accounting “margin boost” that we believe is likely to compress dramatically as origination volume slows relative to balance outstanding, not to mention if they slow down originations in absolute dollars

·         Even if one takes the reported 44% EBITDA margin as a starting point here, and use sell-side math of ~12pts of margin compression per 100bps of credit- that means +370bps on charge-offs takes GSKY earnings to zero- which is a remarkably thin cushion in the context of unsecured consumer lending

o   For comparison, credit costs need to go up +450bps for, e.g., Discover earnings to go to zero- and we are yet to see a sell-side note claiming DFS has “minimal” credit risk, it trades at 9x EPS precisely because of the credit. So even on the bullish credit math GSKY is more credit sensitive than the credit cards- yet trades at a tech P/E

·         Bulls tend to point out that management says their borrowers are homeowners with 750-760 FICO, those people don’t default outside major recessions, and the charge-offs are ~280bps- i.e. GSKY has a highly differentiated borrower profile. We will skip the discussion of ‘average’ FICO vs FICO distribution (public documents suggest they go as low as 640)- but let’s again compare this with Discover:

o   DFS card book is 730 FICO, and their term loan book is 750-760 FICO, which they say behave roughly like 730 FICO revolving because of behavioral differences; their borrowers are 85% homeowner- they are even showing NCOs at 2.70 last year- it’s nearly identical borrower profile if you believe GSKY management. Yet DFS trades at 9x.

o   Even though we are not near a recession right now, DFS’ NCOs have been going up by 50bps every year for last 3 years, and that is happening in the best unemployment environment probably in history. We think DFS multiple correctly reflects that things aren’t likely to get any better for the unsecured borrower from here, and the charge-offs can move up 100+bps without a recession - and in bad recessions, they tend to go up +500bps, with corresponding impact on earnings

·         Sensitivity to credit conditions for GSKY is actually in many ways much worse than a credit card company- at least DFS or SYF have access to stable deposit funding, so as credit costs go up they still can make revenue to compensate. On the other hand, GSKY is at the mercy of the partner banks for every dollar of revenue every day, and the funding we believe is likely to be pulled at the exact time as credit begins to worsen, producing a “double whammy”

o   Financials analysts will immediately know what it means to be a wholesale funded lender who needs to pay ~200bps over LIBOR for funding and whose credit box is dictated by its partners

o   For those who aren’t deep in Financials, you can just look at this quote from S-1, and compare it to many bulls’ belief that GSKY has firm multi-year commitments with banks: “loan origination agreements with certain of our largest Bank Partners, entitle the Bank Partner to terminate the agreement for convenience”.

o   Thus, if at any time bank partners decides that a recession is somewhere on the distant horizon (recall loan term of 7-12 years), or if the bank regulators express discomfort with growing long-term unsecured loans late in the cycle, we believe most partners can stop originations almost immediately, taking GSKY revenue down to a fraction of current run rate. For their largest bank partners, these portfolios are only ~1% of the balance sheet- these banks can turn off GreenSky originations or pull back on them with no material implications for their overall bank revenue. LendingClub’s difficulties in 2016 are an interesting example of how this may play out.

• To the point above regarding funding sensitivity, many bulls seem to say something to the effect of ‘this is a hypothetical- what bank would want to give up a guaranteed spread- the banks will keep funding it more and more- look at the giant TAM’. Well, we actually just got precedents!
• Union Bank, ticker UBSH, reported July 18^th. They were a partner bank of GSKY until this quarter when they stopped originating new loans and sold their outstanding exposure. We will just quote the transcript:
• "it was very clear to us that we could replicate the 3.8% yield if we chose to do so in the securities portfolio in an investment-grade type with a comparable duration. The fact that we were able to sell it at par was fantastic from our standpoint, because, clearly, it's not that it was a good portfolio. We weren't particularly concerned about the risk, but it had a de-risking effort. It had a liquidity improvement aspect. And quite candidly, why would you invest in that when you could turn right around and replicate it in investment-grade securities and have that as dry powder."
• So here Is an actual bank partner essentially saying, as we understand it: even though what we have on the balance sheet right now is worth par- we have no desire to originate more because we want to de-risk and risk-adjusted the returns aren’t there.
• Regions, ticker RF, reported July 20^th. They were a major partner, have grown their GSKY book to 1.7bn, from $1.2bn as of 3Q17 and $700mm 2 years ago. To paraphrase our reading of the transcript - they are looking to cap the outstanding to $2bn max, and the cap will not be reached in 2018 but over a longer period of time. Details on this issue are also available in bank analysts’ notes referencing recent meetings with RF management.
• RF is not referencing origination volume but amount outstanding – and originations (which drive GSKY revenue) need to drop massively if a bank stops growing amount outstanding (which drives bank revenue).

• Bulls also tend to forget or minimize the embedded rate sensitivity. Bank funding is priced off a floating benchmark. Each 100bps of rate increase is ~1,200 bps of margin compression over time according to our calculations. We looked at a number of price sheets over time, and we believe GSKY’s APRs and merchant fees are nowhere keeping up with rate increases.
• We will skip for brevity the detailed discussion of why the reported EBITDA margin compressed by ~1,200 bps over past 2 years

• Bulls seem to believe GSKY has a unique model serving a neglected market niche, and thus possesses a nearly-unconstrained ability to raise prices. In reality, there are multiple players serving the exact same market niche, and not just traditional financial companies like WFC or SFC and niche lenders like Enerbank, but also ‘modern fintech’ companies like Service Finance.
• Internet search surfaces a number of price sheets of competitors, indicating pricing in a fairly narrow range as would be expected in a competitive commoditized product- and GSKY approaches the market with a cost-of-funds disadvantage relative to the ‘more traditional’ players of hundreds of bps.

Now, the above is a very detailed and complex discussion. We wanted to highlight the bulls’ disconnect in something that was easier to follow by those not interested in accounting nuances or bank earnings calls. We called up several bullish analysts and asked them what the key product is and what the unit economics look like.

• The response on product tends to be something like: this super prime consumer who doesn’t need to borrow money happens to be offered a no-interest-for-12-months financing by his HVAC guy, sees it as “free money”, takes it and feels good- while the contractor upsells them on some extra stuff. The consumer pays the loan off at the end of 12 months and is happy because it’s “free money”, contractor pays a 10% fee to Greensky and is happy because the consumer being happy helped them close the deal and maybe even upsell, and bank is happy because they made money with no risk.
• However, we believe the bulls to be meaningfully off:
• First, as we discussed, all promotional offers are only 40% of the mix, while the remainder we believe is  fixed (e.g., 7.99% APR) 7+ year term loans that are in many ways identical to what one can get at, e.g., LendingClub, which is currently lending $10k+ unsecured for 60 months to 720+ FICO borrowers at APR as low as 7.5%
• Second, internet search finds training materials for contractors that instruct them to “mark up” the cost of their projects, and then offer a discount to consumers who don’t need financing- so it’s not “free money” to consumer but ~5 “points upfront”, and thus there is very likely to be adverse selection as to which consumer ends up borrowing from GreenSky vs using other sources of funds
• But most importantly, as we discussed above, pricing sheets for GSKY are available online and instead of using hypothetical numbers like “10%” one can construct actuals- and as we discussed above, economics on this promotional product looks nothing like what the bullish analysts lay out
• To recap, we believe GSKY gets ~650bps of fee revenue, gives ~400bps to the bank, spends ~120bps on origination and servicing and another ~170bps in allocated OpEx, plus absorbs credit costs incurred.
• In order to understand why a company may write unprofitable product like this, we note that according to GSKY disclosures ~10% of the promo borrowers finish the promo period and, instead of paying off the remaining balance, choose to go into the non-promotional term. And these borrowers’ profitability we believe to be very different. We will illustrate our understanding with a borrower who purchases a $9,500 HVAC replacement:
• The borrower takes a $10,000 loan with a 12-month promo product and makes minimum payments
• At 12 months, that borrower has made ~$3,200 in monthly payments, and has a choice of either finding the remaining $6,800 in their bank account/ home equity line/ LendingClub etc., or…
• …they are charged not only 27% APR on their balance for the next 6 years, but their balance is stepped up by ~40% (ie back to $9,500) because Greensky charges them 27% APR retroactive to day 1.
• So in an environment where prime borrowers get 5-year loans at LendingClub with a stated use for “home improvement” at sub-8% APRs, these borrowers end up paying a 27% APR on a balance that steps up ~40% upfront (or effective APR of well over 40% for the remaining 6 years of loan term)
• We speculate these borrowers are not prime in their behavior, regardless of FICO at origination:
• Maybe something happened in the life of these borrowers since, and they no longer have any other access to credit,
• Or maybe the contractor has adversely selected them since the contractor has no stake in longer-term borrower performance,
• Or whatever other reason why a borrower would choose to pay effective APRs more commonly associated with deep subprime.
• This (as we call it) “subprime” population would then produce massive economics in their 1st year of paying the non-promotional APR, which we believe is the reason why GSKY accepts unprofitable borrowing from “remaining prime” borrowers
• However, unsecured subprime lending is not a new invention- in fact, investor return curves for LendingClub's grade F&G 60-month term loans are on LC website. The losses rates, unsurprisingly, are very high over the life of the loan.
• What we observe in LC curves (and what is known generally in consumer lending) is a massive yield is available in the initial year- i.e., even subprime borrowers try to pay their debts at first- but eventually “life catches up with them”, and by the time years 3, 4, 5 come by, losses consume more than 100% of all the proceeds.
• GSKY expenses losses as they come, there is no provisioning for the future- i.e., as long as the “subprime” book is not seasoning the initial-year profitability overwhelms the later year losses.
• However, that profitability requires a near-exponential growth curve- the moment that near-exponential growth doesn’t happen and the book begins to season, we believe margins are likely to collapse.
• Again, LendingClub is a good example to consider- at this point, LC has stopped writing grades F&G altogether- as the book seasoned, we believe the investors realized these types of loans produced a negative yield over the life of the loan, and LC was forced to refocus on better credit tiers.

Valuation

• Their main competitor, Service Finance, sold in a competitive transaction in September of last year, to a Canadian outfit called ECN Capital, and you can see their numbers in ECN's disclosures. They are growing 2x as fast and yet sold for 1/4 the valuation in a competitive bid.
• Using that as a benchmark, he stock is worth 10x consensus '19 EPS, which is $9 or -50% lower- and if you use ECN’s current market valuation, it appears to be worth even less.
• Given the credit and funding sensitivities, in the medium term (i.e., some point prior to the next recession) the stock is worth $0, as no capital has been set aside to survive a funding pullback- in fact, the company levered up before the IPO
• S-1 contains details regarding the use of pre-IPO term loan and IPO proceeds to the benefit of insiders that we will not discuss here

Risks

• The main risk we see is a short-term headline risk causing the stock to spike temporarily
• As an example, the company has recently announced a partnership with American Express, which got the market excited. The announcement is short on details, especially around the economics, so we won’t speculate on those here, but here are some thoughts on why we believe it’s an example of a short-term headline risk and not likely to be a story changer:
• Amex already has the technology to write both off-card term loans in a fully-online process, and more importantly an in-app functionality to convert specific card charges into low-APR term loans (try it yourself- it is the "Plan It" button you see in the app next to transactions over $100. There is very little that Greensky can add to them from a technology standpoint
• Amex has the cardholder and the merchant relationships- they aren't getting those from Greensky, Greensky is getting those from Amex. If Greensky has to pay for access, e.g., by giving up the interchange they would normally receive in a MasterCard-based transaction to AXP, they would lose very material economics (interchange accounts for ~1/3 of reported profit) which as we discussed above already is highly inadequate
• Amex is actively trying to grow its own lending book- that is a key element of their growth story- and has access to deposit funding that is ~200bps cheaper vs what GSKY has to pay its partner banks...they absolutely can originate these loans onto own balance sheet themselves at a cost advantage
• We suspect that AXP is using this partnership with Greensky to experiment with loan terms that are much longer vs what they currently offer- and if they loans perform, Amex is likely to take future volumes in-house, and if the loans don’t perform- it’s GSKY shareholders' problem...and we believe that AXP management appears to emphasize the "pilot" nature of this partnership for a good reason
• The company has, in fact, promised multiple upcoming bank partner announcements to the bulls during the roadshow, so in a way, some level of announcements is ‘priced in’
• We are not sure what adding bank partners can prove at this point, as management already claims 'commitments' of >$8bn from existing partners- which of course tend to be cancel-able on a short notice
• ‘Bigger TAM’ narrative taking hold- the company is talking about expanding TAM beyond home improvement into segments like elective healthcare, and the bulls may get excited about reports of early traction in these segments
• However, the elective healthcare space has been served by, e.g., Synchrony CareCredit for a long time, and we believe the main hurdle to wider adoption has not been  lack of convenient options to the consumer but very high realized credit losses- something that we believe the GSKY product does not address in any new way
• We note that at the time of IPO GKY reported $67mm of “R&D Receivables” on GSKY’s own balance sheet (i.e., receivables not transferred to bank partners due to “insufficient experience with particular products or industry verticals” per S-1), while cume originations of Elective Healthcare loans stood at $91mm

Disclaimer:

I do not hold a position with the issuer such as employment, directorship, or consultancy. 

I and/ or others I advise have a (long or short) investment in the issuer’s securities.  I may also sell (if long) or cover (if short) at any time without notification.

Catalyst

Investors realizing that GreenSky is reporting higher margins vs Lending Club primarily because of its unusual accounting and retention of tail credit risk and not because of a unique business model.