Not Everything Will Be a Bank

Ryan Floyd

November 23, 2021

Comments on owning faster-growing businesses

A company that grows topline for five years by 35% above inflation versus one that grows by 20% justifies a very different valuation, although 20% is excellent by any standard. As the world digests the probability of these different future paths, stock prices can swing around a lot. I realize this sounds obvious.

In many ways, I am doing some of the best due diligence work of my life, triangulating data points from different perspectives and seemingly understanding a business as well as anyone. However, I often think about the lessons from reading Benoit Mandelbrot on fractals: reality is not necessarily clearer at different scale when one has a stronger magnifying glass. Nevertheless, I still like having those extra lenses.

Though still naïve, I enjoy having discounted cash flows on rapidly growing companies, even though these models provide only a rough shape of the future. In addition, I have enjoyed backing into the implied long-term growth rates by current valuation metrics too. This helps as a gut check for reasonableness.

Will everything really be a bank?

An idea currently circulating says that “everything will become a bank” or, at least, that all digital businesses will be banks; that is, digital businesses know their customers so well that they can underwrite loans with lower defaults than traditional banks, who see only a fraction of a customer’s purchases. Let’s compare QuickBooks to First Republic Bank. Intuit can see its customers’ full financial statements and each of its transactions from different banks and credit cards. In comparison, a bank is only one of several financial institutions that a small business will use, so it must be more conservative in its lending approach. Using this idea, Intuit’s loans should default less frequently than that of a bank. As a result, they argue, customer-facing software or marketplaces should lend extensively to their customers to help grow revenues.

I am skeptical about this. I don’t mind that the e-commerce has developed their own payments platform. I can see the logic in Intuit selling receivables to financial institutions. But I scratch my head a bit when companies get really excited about extending credit to their customers on their own balance sheet and calling it “fintech.” Call me old school, but in my world, “extending credit to customers” just means higher receivables, more net working capital, less outright cash flow, and potentially more volatility in earnings as bad debts materialize. Likewise, “Buy Now Pay Later,” sounds a lot like the installment loans with which my grandparents’ generation bought their 1947 bedroom set. Even Sears Roebuck in the early 1900s tried to become a bank but found too many regulatory hurdles and balance sheet risks so shelved the project.

In mathematical terms, companies that extend a lot of credit and take balance sheet risk on receivables and loans trade at lower multiples because their earnings are more volatile, and they have a higher probability of imploding. Underwriting loans is a specific kind of muscle. Good companies that underwrite well, like some banks, have institutionalized memory around the default cycle and pass on the tales of the hundred-year flood to new employees. But in quant parlance, they have a higher “beta.”

I find this reasonable. Somehow, investors are trying to own stocks with more receivables on their balance sheet, take the risk of default, but still trade at valuations of companies with lower beta. They can’t have it both ways.