We Need to Talk about SoFi’s Business Model

On this week’s Slate Money podcast, I argued that SoFi’s frat house culture isn’t the only problem plaguing the marketplace lender. SoFi may claim that its disruptive business model is as bulletproof as its members’ resumes. But, if economic conditions change, this model could turn out to be about as stable as a drunk couple balancing on an office john.

SoFi likes to remind everyone that they don’t have customers: they have members. But, I’d like to remind everyone, that SoFi’s borrowers aren’t really customers or members: they’re the product. And this could become a problem if Sofi’s true customers — the hedge funds and institutional investors purchasing their loans and loan-backed securities — stop buying.

To understand the flaws in SoFi’s business model, it’s important to understand how this model differs from a traditional banking model. So let me give you a Cliff’s Notes version of banking finance. I know this isn’t as exciting as tales of sex in office parking lots, but please bear with me.

Why old-fashioned banks are magic

Traditional commercial banks take deposits and make loans. If you imagine a balance sheet with assets on one side and liabilities on the other (plus shareholder equity, but we won’t worry about that for now), deposits represent the liabilities: banks borrow depositors’ money, which is why banks must pay interest and ultimately return the cash. Loans represent the bulk of a bank’s assets: banks use deposits to make loans that pay the bank interest. Banks are only required to keep a fraction of their deposits either on hand to satisfy withdrawals or with the central bank to satisfy reserve requirements. The remainder can be lent out. Boring, perhaps. But this is how banking enables the economy to thrive. I loan the bank $100; the bank loans someone else $90. My $100 just became $190, hooray! Fractional reserve banking is a beautiful thing.

Now, in reality, banks have other funding sources. There’s public equity and debt financing: stocks and bonds. Large banks can borrow short-term from other large banks, giving us such delightful terms as LIBOR and the federal funds rate. NPLs can be offloaded. Available-for-sale securities can be sold. Structured securities can be made. Suffice it to say. Banks have options. Which is important because banks have short-term liabilities that must be met. Which is why banks are also required to maintain sufficient equity or equity-like capital so that they won’t have to impair senior bondholders or depositors — or require government assistance — during an economic crisis when the bank’s market value declines, access to liquidity disappears, and depositors demand to be paid.

Okay, so, if you’re still with me, you might ask yourself. How does a bank make money? While many assume it’s from the overdraft and credit card fees everyone hates, the bulk of a bank’s profits actually come from a bank’s net interest margin — the difference between the interest banks have to pay their depositors and the interest they receive on the loans they make with the depositors’ money.

Which brings me to SoFi. They have none of this. Okay, that’s not exactly true. They have raised over $2 billion in private debt financing and venture capital. But what they don’t have are deposits, or, for that matter, many loans.

Lack of predictable funding

SoFi loves to remind everyone that they’re not a bank. And this is true. They are a marketplace lender (MPL). So they can’t rely on deposits to fund their loans, and they don’t have reserve or capital ratio requirements. They must constantly offload their balance sheet to make room for more loans, so they can generate more fee income. SoFi may have raised over $2 billion in capital, but they’ve originated over $10 billion in loans.

SoFi’s original plan was to rely on the kindness of university alumni: wealthy investors would fund the loans and thereby help recent graduates from their alma maters, making a nice profit on a can’t-lose investment. Everybody wins. But SoFi quickly discovered the limits of school spirit. Which meant they had to find new buyers. So they began to sell the loans and loan-backed securities to hedge funds and other institutional investors who were attracted by higher yields and eager to give money to an innovative fintech pioneer.

So what’s the problem?

The problem is that if appetite for these loans and securitized products declines, the company can’t simply rely on stable deposits and interest income. As Todd Baker of Broadmoor Consulting has noted, the company must keep making and selling new loans, or the model doesn’t work. While Baker admitted that SoFi is in a slightly safer position compared to other MPLs because of its $2 billion capital position and because they hold some of their loans, he has also pointed out that this “safety” is marginal. Cagney himself has admitted that the industry is facing “balance sheet problems,” with too little money chasing too many loans. And, even more frighteningly, he’s been quoted as saying, in reference to a loan product offered to investors at a rich yield: “In normal circumstances, we wouldn’t have brought a deal into the market, but we have to lend. It’s a problem in our space.” When a financial professional is forced to engage in market activity despite the investment climate and the metrics of the deal, it’s time to worry.

Investor interest in consumer-focused fintech has declined over the past few years as investors have expressed concerns about the sector’s ability to scale and sustain rapid rates of growth. While no one can predict whether investor appetite will increase or decrease tomorrow, that’s just the point. It’s unpredictable. Unlike a traditional bank that has deposits, holds reserves, collects interest, and let’s be honest — has the de facto support of the federal government — online lenders are dependent on their ability to attract borrowers and buyers. And they can sometimes get ahead of themselves. In 2012, SoFi lied to investors about securing $90 million in financing for a loan product. As the NYT reported, they subsequently bought the product back from investors, and paid out the promised returns. They claimed no investors were harmed, but this is not the behavior of a mature, ethical company. This is the behavior of a company that will do anything to survive — even when this behavior borders on illegal.

Lack of funding diversity

Let me repeat: SoFi’s borrowers aren’t the customers; the investors are the customers. So you might think that SoFi would avoid directly competing with its customers— i.e., its lifeblood. But this is Silicon Valley. And they don’t play by the rules. In 2016, SoFi set up the SoFi Credit Opportunities Fund, a hedge fund intended to purchase SoFi’s bonds, as well as those of other lenders. SoFi also discussed setting up a REIT (real estate investment trust) to buy SoFi’s mortgages. Why did they do this? Because, in theory, they could raise money from investors who weren’t interested in buying their loans and securities directly, but who would be interested in investing in a fund that invested in loan-based products. In theory, this could diversify their funding sources. In theory, they could trade in derivatives to offset their rate risk. And, in theory, they could raise $1 billion. But, as other commentators have noted, there are a few problems with this theory: (1) investors might be concerned that SoFi, who has an obvious informational advantage, might select the best loans for its own fund; (2) although SoFi was targeting a raise of close to $1 billion, the fund had extremely modest beginnings ($15mm); and (3) if an economic correction led to a slowdown causing investors to shun direct exposure to these products, it would likely also cause investors to shun indirect exposure. They’re the same investments, and likely the same class of investors. Calling this diversification is a stretch.

Underwriting problems

SoFi might argue that its products will perform better during a downturn because of the strength of their underwriting model. But this is only true if they don’t degrade this model to accommodate faster loan growth. Up until now, SoFi has seemingly only lent money to the most credit-worthy investors. But, again, SoFi is an MPL, and MPLs must keep making more and bigger loans if they want to survive. SoFi’s recent movement into mortgages — which are, on average, significantly larger than student loans — has already proven challenging. The NYT reported that Sofi was approving mortgages without including appraisals or verifying incomes. They also reportedly began to allow customer service representatives to approve loans. While SoFi claims they have stopped these practices, these growing pains are more than a little concerning when the lender’s entire raison d’être is the quality of its loans.

Prepayment risk

When you’re dealing in loans or bonds backed by loans, you not only have to worry about default risk; you also have to worry about prepayment risk. If I buy a bond that will pay me interest and maintain its value based on the cash generated from an underlying package of loans expected to pay interest over a set number of years, I will be pretty pissed if all of the borrowers suddenly decide to pay off their loans immediately. There goes my cash flow (and much of the net present value of my investment). Prepayment can refer to literally paying off a loan faster than required (i.e., paying more than you owe each month or paying off the entire loan early), or it can refer to refinancing your loan. When SoFi refinances a student loan, it is prepaying the original loan with the new loan it just made to the graduate.

While SoFi loves to tout their low default rates, I’ve heard little about their prepayment rates. Millenials are known to be more debt averse than previous generations, and millennial graduates from elite universities are reported to prepay loans at higher rates than average borrowers. I myself am a millennial who graduated from an elite university, and I paid off my undergraduate student loans within one month of graduation (I was lucky; I know). So SoFi is lending to the most attractive borrowers, who are both those most apt to be in the position to prepay their loans and those most apt to receive refinancing offers from other lenders.

SoFi can obviously mitigate this risk by securitizing loans and offering higher returns to investors who purchase tranches of loans with higher prepayment risks. But, as with any structured product, the risks are only properly accounted for if the underlying assets behave as expected. The Wall Street Journal reported that borrowers of SoFi’s initial crop of loans were paying off 30% more than required, but prepayment rates had only been anticipated to be 10%. While this rate was not large enough to significantly impair the bonds, the WSJ also noted that some of the bonds backed by these early loans had fallen in value as of 2016, likely because of prepayments.

Excessive risk tolerance

Going back to the sex scandal for a moment, let’s remember that Cagney allegedly sexually harassed his assistant after he hired his wife to work at the same office. This suggests that his tolerance for risk is, how shall we say, high. Now, all entrepreneurs need to have a fairly high tolerance for risk. Otherwise, we’d all still be talking on rotary phones. But there’s risk, and then there’s running-your-company-like-Animal-House-while-seeking-to-IPO risk. And Cagney appeared to have a fondness for the latter. In 2016, about one month before the Brexit vote, Cagney told the Financial Times that he would be happy to originate mortgages in London at 90% loan-to-value because of the value of London real estate – which certainly would never decline significantly. I bet the company is now happy that he couldn’t legally make these loans, as the value of London real estate has, in fact, declined, with the most significant declines occurring at the higher end of the market. Investors who don’t understand that high prices can, it turns out, come down tend to become cautionary tales.

Now, Cagney is out, and a more reasonable CEO is likely to be appointed. SoFi is currently at little risk of failing. Investor demand may have slowed, but it hasn’t disappeared. Default rates have increased, but remain relatively low. And SoFi is still valued at $4 billion and will likely IPO once the dust settles. But its business model — as well as its business practices— remain questionable.

And no one knows how SoFi or other MPLs will perform if central banks become more hawkish, if the U.S. economy slips into recession, or if Chinese growth slows significantly. What seems like a can’t-lose proposition during an economic upswing can often become a must-lose reality during a downturn.

It’s telling that before the scandal, SoFi was seeking a charter that would allow them to start acting more like a real bank, including enabling them to accept deposits. Which suggests that the bad boy of fintech is coming to terms with the limits of disruption.