Reviewing the Risks In WeWork’s $500 Million Bond Deal

This article first appeared in Data Sheet, Fortune’s daily newsletter on the top tech news. To get it delivered daily to your in-box, sign up here.

Real estate machinations have been a regular feature of a career in media over the past decade. Aaron in for Adam today, thinking about my current base of operations in a Boston WeWork.

The appeal of my WeWork space is fairly obvious: Someone else maintains the free coffee, waters the plants, and tidies up the sun dappled work spaces. Plenty of add-ons are on offer, too. Pay extra for use of a big conference room, printing, VoIP phone service, or snacks. (Unlike working at a tech giant, the food isn’t free.) And in Boston, the five locations are spread throughout all the hottest neighborhoods.

The hip office sharing startup is back in the news today as it seeks to borrow at least $ 500 million in the junk bond market. Interest rates have been ticking up this year, you may have noticed, but for WeWork they are especially high. Netflix (nflx) just borrowed $ 1.9 billion for 10 years at a rate below 6%, and Uber’s $ 1.25 billion seven-year loan was reported last month to be below 5%. WeWork is also burning through cash to fuel rapid growth just like those two, but the less established startup will have to pay around 8% on its bonds.

So far it’s all working quite nicely to spur strong growth at WeWork. Revenue doubled last year to $ 822 million, according to the bond offering documents—but so did the net loss to an impressive $ 933 million. Of course, WeWork, like many startups before it, has its own special metric, dubbed “Adjusted EBITDA before Growth Investments,” which takes out all those niggly expenses like taxes, interest, depreciation, and amortization plus vast amounts of other stuff like almost all of its sales and marketing costs. On that basis, it generated $ 49 million.

But there’s an even deeper concern about WeWork’s basic business model, and it’s reminiscent of several banking crises of the past. WeWork itself has to enter into long-term leases on its office space, typically for 10- or 20-year terms. At the same time, its startup and gig economy fueled clientele are renting month to month. Banks have gotten into trouble because they relied on funding from savings and checking account depositors that could be yanked daily, while the loans they made with that money didn’t have to be paid back for years. In bad times, depositors wanted their money back all on one day, but the banks didn’t have it, creating the classic run on the bank disaster scenario.

The timing mismatch feels equally dangerous for WeWork. When the economy finally hits a recession, it will still need to pay back billions: $ 5 billion through 2022, and $ 13 billion more in 2023 and later on its existing lease commitments, to be precise. The company has structured the leases through subsidiaries that could be closed independently, but customers could be canceling their monthly leases in droves. Even at a juicy 8% yield, it may not be a risk worth taking.