How an App Became a Bank for 17 Million People

Transcript

From the video captions, split by chapter. Timestamps link to that moment on YouTube. Corrections are marked inline.

2 passage(s) with price targets or valuation scenarios were removed from this transcript (see Methodology).

0:00 How did an app become a bank for 17 million people?

Seventeen point one million. That is how many members SoFi had at the end of September, up thirty-five percent in a year. In just three months, it added a record one point two million of them. A company that started out refinancing student loans now looks a lot like a bank that lives in your phone.

Today I want to ask why that could matter for the next decade. Now zoom out. In a late twenty twenty-five survey by Morning Consult for the American Bankers Association, fifty-four percent of US bank customers said a mobile app was the way they managed their account most often. Only nine percent said a branch.

Back in twenty seventeen, the same survey found just twenty-six percent for the app. That share has more than doubled. Among Gen Z it is sixty-three percent, and among millennials sixty-seven percent. For most people, the phone is now the front door of the bank.

This is Bet The Decade. Today: SoFi. How a student loan startup got a bank charter, how its product loop works, what the latest numbers say, the bear case, and my thesis. This is not financial advice.

1:11 How did a student loan startup become a bank?

SoFi, short for Social Finance, was founded in twenty eleven at Stanford's business school. The first idea was simple: alumni lending to students. In twenty twelve it added student loan refinancing, and that became its core product. By twenty nineteen, it was big enough to put its name on a stadium.

SoFi signed a twenty-year naming deal for the new home of the Los Angeles Rams and Chargers, reported at more than thirty million dollars a year. In twenty twenty, it agreed to buy Galileo, a payments technology company, for one point two billion dollars. And in June twenty twenty-one, SoFi became a public company on the Nasdaq. Then came the key step.

In January twenty twenty-two, US regulators approved SoFi's plan to become a national bank, by buying a small California lender called Golden Pacific Bank. Why does that matter? A bank charter lets SoFi take deposits from its own members, and lend that money out, instead of relying mainly on borrowed money from other lenders. That one change sits under almost everything that follows.

SoFi itself describes the journey like this. In twenty eighteen, it was a single-product student loan refinancing business. Today, it calls itself an everything financial app.

2:32 How does the SoFi loop work?

SoFi calls its strategy the Financial Services Productivity Loop. In plain words: win people over with one product, then offer them the next one, and the next. At the end of September, members held twenty-six point seven million SoFi products, up forty-four percent in a year. That is about one point five six products per member, up from one point four seven.

And fifty-one percent of new products were opened by people who were already members. Thirty-seven percent of members now use two or more products, up from thirty-three percent a year earlier. That is the loop, working in the data. The biggest pieces: SoFi Money, the checking and savings account, with eight point five million products.

Relay, which tracks credit and spending, with eight point seven million. SoFi Invest, with four point two million. And the credit card, growing fast, at about six hundred fifty-six thousand. Here is where the bank charter pays off.

Deposits grew by five point eight billion dollars in a single quarter, to fifty-one point three billion. On average, deposits made up ninety percent of SoFi's liabilities. Deposits are usually a cheaper way to fund loans than warehouse lines from other lenders. SoFi says it paid a much lower rate on deposits than on those facilities.

That helps explain a net interest margin of six point one four percent last quarter.

4:00 Is SoFi still a lender at heart?

Lending is still the engine. In the third quarter, SoFi originated a record sixteen point nine billion dollars of loans, up seventy-one percent. Personal loans were twelve point six billion of that. Student loans were two point nine billion, and home loans one point four billion.

Not all of it stays on SoFi's balance sheet. Through its Loan Platform Business, SoFi originates loans on behalf of outside partners, and earns fees. That business did about three point eight billion dollars of the quarter's originations. Why does that matter?

It is a capital-light model. SoFi can grow its lending, and its fees, without tying up as much of its own balance sheet. That is part of a bigger shift. Fee-based revenue reached about five hundred seventy-one million dollars, forty-two percent of total revenue.

Over time, management wants fees to be half of revenue. And there is a third segment, the Technology Platform, built around Galileo. It runs accounts for other financial companies, behind the scenes. It powers about one hundred forty million accounts.

But segment revenue fell twenty-four percent in a year, after a large client moved off the platform. Excluding that client, it grew eighteen percent.

5:19 What do the latest numbers say?

Now the scoreboard. Third-quarter adjusted net revenue was one point three seven billion dollars, up forty-four percent. Reuters reported that analysts had expected about one point two five billion. Net income was two hundred forty-one million dollars.

Adjusted EBITDA was four hundred seventy million, a thirty-four percent margin. SoFi says it was the twelfth profitable quarter in a row. There is also a simple test that many growth investors use, the Rule of forty: revenue growth plus profit margin should add up to at least forty. SoFi scored seventy-nine, with forty-four percent growth and a thirty-four percent EBITDA margin.

It says it has passed that test for twenty quarters in a row. SoFi raised its full-year revenue outlook to about five point zero four billion dollars. And it gave an early look at twenty twenty-seven: about six point four billion in revenue, roughly twenty-seven percent growth, with net income of about one point two eight billion. That would be a net margin of roughly twenty percent.

The long-term goal is even bigger: one hundred million products by the end of twenty thirty. That is roughly four times today's level. For context, Reuters put SoFi's market value at about eighteen point seven billion dollars around the report. [Sentence removed on this site: price target or valuation scenario. See Methodology.]

So the market is not pricing it like a mature bank, and not like a high-flying tech stock either. Whether that guidance is met is the real question.

6:56 What could go wrong?

Now, the other side of the trade. Bear one: credit. About three of every four dollars SoFi lent last quarter went into personal loans. These loans are unsecured, with no house or car behind them.

The annualized net charge-off rate on personal loans was about three point four percent. That has improved, but in a real recession, losses on loans like these can rise fast. Bear two: interest rates. SoFi holds many of its loans at fair value.

In the third quarter, higher discount rates caused more than eight hundred million dollars of fair value losses. Hedges offset most of that, about seven hundred thirty-four million, but rate swings can still hit results. Bear three: the tech platform. Losing one large client cut that segment's revenue by almost a quarter.

It shows how concentrated that business can be. Bear four: dilution. In twenty twenty-five, SoFi sold new stock twice, about one point five billion dollars each time, in July and in December. Management says the twenty twenty-seven plan is funded without raising more capital, but more shares mean each share owns a smaller slice.

Bear five: competition. Big banks have strong apps too, and so do other fintechs. SoFi has to keep winning members against both, while running many businesses at the same time.

8:23 Why am I betting the decade on it?

Fast-forward to twenty thirty-five. In the bull case, the phone is where most people bank, and SoFi is one of the main apps they choose. Members hold many products, deposits fund the loans, and fees keep growing. That is a scenario, not a forecast.

So here is my thesis. Banking is moving from the branch into the phone. SoFi has a national bank charter, a fast-growing deposit base, and a loop that sells more products to the same members. If that loop keeps working, I think SoFi could become one of the big consumer banks of the next decade.

The risks are credit, and above all, execution. What to watch: credit losses on personal loans, deposit growth, the tech platform's recovery, and progress toward one hundred million products. Here is my personal model, and it is not a forecast or advice. [Sentence removed on this site: price target or valuation scenario. See Methodology.]

The method: revenue at the end of the horizon, times margin, times an earnings multiple, compared with today's valuation, after dilution. It is my opinion, and it could be wrong. Not financial advice. [Personal holdings statement omitted on this site. The author may hold shares in companies discussed. See Disclosure.]

Do your own research. Would you trust one app with your whole financial life? Tell me in the comments. Next, watch my full Tesla film, linked on screen, for another company that wants to be much more than its first product.

If you like thinking in decades, not quarters, subscribe. See you in the next one.