- You want a loan.
- You got to Lending Club's website and fill out a form with, like, your name and how much money you want and why.1
- I have some extra money.
- I go to Lending Club's website and open an account.
- I browse the borrowers and think you look like a likely candidate.
- I lend you money.
- Lending Club sets up the loan, collects payments from you, keeps an eye on things generally, and takes a fee for its trouble.
None of that is exactly wrong, but here is a more accurate picture: 2
- You want a loan, fill out the form, etc.
- I open an account, think you look likely, etc.
- A bank gives you the loan.
- The bank sells the loan to Lending Club.
- Lending Club keeps the loan on its balance sheet.
- Lending Club creates an unsecured structured note referencing that loan.
- It sells that note to me.
- Then you make payments to Lending Club, and it makes payments on the note to me.
- You never owe me money: You owe Lending Club money, and Lending Club owes me money.
These things are economically similar enough (enough for what?), but the vagaries of law and accounting and convenience and so forth mean that Lending Club ends up standing between you and me not just as an agent -- setting up my loan to you -- but as a principal. I lend money to Lending Club and can look only to Lending Club for repayment, while Lending Club lends the money to you in a separate-though-obviously-related transaction.3
This has some pretty goofy consequences. For instance, Lending Club's loan investors -- who tend to be institutional investors, not actually schlubs like me, so "peer-to-peer" is a bit of an inaccurate description -- don't own their loans, so they're subject to Lending Club's credit risk as well as to the borrowers' risk.4 And Lending Club is, in the eyes of U.S. generally accepted accounting principles, a very heavily indebted company. Here's a dumb thing I did:5
If Lending Club was a bank, it would look like a pretty risky bank!
But of course it's not a bank. There are a lot of ways in which it is not a bank, but the big one is that basically all (95.6 percent) of its liabilities are "notes and certificates," that is, just unsecured structured notes tied directly to specific underlying loans. Banks, on the other hand, are funded mostly by deposits and repo and other short-term senior borrowing. So:
- Lending Club's assets and liabilities are perfectly matched in duration: Those notes and certificates mature when the corresponding loans mature. A bank, on the other hand, is in the business of borrowing short to lend long.
- Lending Club's assets and liabilities are perfectly matched in loss bearing: Every dollar that a borrower doesn't pay back to Lending Club is a dollar that Lending Club doesn't pay back to note holders. The note holders know going in that they bear the entire risk of loss on the underlying loans. A bank depositor expects to get her money back even if the bank makes some bad mortgage loans.
This takes, in round numbers, all of the risk out of Lending Club's balance sheet. I mean, its business has risks: It can make dumb investments or competitors can steal all its customers or whatever. And its debt -- those notes and certificates -- is risky, in that the borrowers might not pay it back. But all of the things that make banks scary don't apply. A run on Lending Club is not possible; nobody can pull their money out of notes or certificates. And if a lot of loans go bad, that will hurt the investors in those notes, but Lending Club as an entity won't be insolvent or even have any losses at all.6
So that dumb leverage math I did above, though absolutely correct as a matter of GAAP accounting, is entirely wrong as a matter of business reality. In any useful sense, Lending Club is a 100 percent equity-funded bank: Every dollar that it lends comes from long-term, loss-bearing investors. Lending Club is the perfect bank under the Anat Admati school of thought, which wants banks to have vastly more equity capital and vastly less leverage. Even though it's 19 times levered.
This is all obvious, but it makes Lending Club a lovely model for illustrating points about What Banks Are. Here are some.
First: You wouldn't want all banks to be like Lending Club. People like having checking accounts. Even beyond checking accounts, people like having liquid risk-free money-like claims. Banks exist to turn risky investments -- loans to people and businesses! -- into risk-free money-like claims, whether those are checking accounts or certificates of deposit or repo agreements. That is a dirty business, because turning risky investments into risk-free claims requires making the risk disappear, which is just a magic trick, relying on diversification and deposit insurance and capital requirements but also some amount of good old sleight of hand. The point of banking is to conceal risk.7
Lending Club does not conceal risk -- it gives its investors all the information it has -- and so it cannot serve the social function of banking. It doesn't claim to, and there's no reason that it should. But while it is a perfect equity-funded bank, it does show the limits of equity-funded banks. Banks -- real banks -- are in the business of producing loans, sure, but they're really in the business of producing money-like claims. Equity-funded banks can't do that.
There's another point that's a flip side of this one, which is: Equity-funded banks are great at lending. Lending Club is perfectly able to make loans, and apparently at cheaper rates than banks. So while I think Lending Club does not make an unambiguous case that equity funding is the way to go for banks, it is a good data point for the proposition that increasing bank capital would not hurt lending.8 Lending Club, with 100 percent capital (ish), is growing its lending business -- although that business is still much smaller than the lending business of any big bank.
Against that, though: Banks couldn't do what Lending Club does. Imagine that JPMorgan said "We're going to make some new loans and fund them by selling unsecured structured notes that reference those loans." What would happen? Well, JPMorgan would do, say, $100 billion of new loans, and sell $100 billion of precisely offsetting notes. That business, considered as a unit, would be no more or less risky than Lending Club's business -- which is to say, not risky at all, to JPMorgan's capital or depositors or whatever.
But that business would also make JPMorgan's capital and leverage ratios much worse: It would gross up its balance sheet by $100 billion of assets, and fund those new assets entirely with debt. I mean, with a thing that I called "100 percent equity funding" a few paragraphs ago, but nonetheless with a thing that accountants and bank capital regulators call "debt." (Because it's debt.) JPMorgan would have to raise an extra, I don't know, $8 billion of common equity9 to do the business that Lending Club can do with pretty much zero equity if it wants to.
This is sort of fanciful, but it has more realistic implications too. There's a general consensus that new leverage-ratio rules have reduced banks' matched-book repo business. Matched-book repo looks not entirely unlike what Lending Club does: A bank lends against collateral from a hedge fund, borrows against that same collateral from a money market fund, and clips a fee for standing in the middle. The bank is not really taking risk, but it is using balance sheet, and the rules for banks make it expensive to grow their balance sheets.10 Because no one trusts banks: When they grow their balance sheets to do risk-free activities, people just assume that they're hiding risks somewhere.
But Lending Club can grow its balance sheet all it wants. Lending Club is not a bank. So it's not subject to banking regulation, which means that it can do a core function of banking much more efficiently than an actual bank can.
Here it is:
I mean, there's probably a longer form if you actually want the loan. Also there's a cartoon guy's elbow in the the bottom left corner of that picture.
2 This picture comes from the IPO registration statement but also from the Lending Club notes prospectus, which it uses to issue some of those unsecured notes. Also I should say that Lending Club does sell whole loans to some (presumably large) investors -- moving them off its balance sheet -- and that's a growing business, with almost a billion dollars of sold whole loans at the end of June. But the notes-and-certificates route remains bigger.
Laugh at "watress," and the blank loan description, if you want. But consider how much better this disclosure is than the borrower-level disclosure typically provided to investors in mortgage-backed securities deals. The Securities and Exchange Commission just today issued final rules requiring standardized disclosure of borrower-level information. But here you pretty much know everything that Lending Club knows -- and you know that they don't know anything else. Like, you're not gonna read that description and say "oh but I'm sure some underwriter knows the purpose of this loan and underwrote it according to its customary standards." You're gonna read that and say, meh, it's $8,000 at 6.69 percent for three years, seems like a fine risk.
There is not perfect information. But there is no information asymmetry. Lending Club doesn't know anything you don't, and it's not selling you the garbage while keeping the good loans. (It's not keeping any loans.) You're making your own decisions.
3 And really Lending Club isn't even lending the money to you -- it needs a bank to do that -- though the bank is more or less an agent and sells the loan on to Lending Club immediately. Lending Club steps into the bank's shoes and really is your lender, in a way that I am not.
4 Here's a risk factor from the notes prospectus:
In a bankruptcy or similar proceeding of us there may be uncertainty regarding whether a holder of a Note has any priority right to payment from the corresponding Loan. The Notes are unsecured and holders of the Notes do not have a security interest in the corresponding Loan or the proceeds of the corresponding Loan. Accordingly, the holder of a Note may be required to share the proceeds of the corresponding Loan with any other creditor of ours that has rights in those proceeds. If such sharing of proceeds is deemed appropriate, those proceeds that are either held by us in the clearing account at the time of the bankruptcy or similar proceeding of ours, or not yet received by us from borrowers at the time of the commencement of the bankruptcy or similar proceeding, may be at greater risk than those proceeds that are already held by us in the “in trust for,” or ITF, account at the time of the bankruptcy or similar proceeding. To the extent that proceeds of the corresponding Loan would be shared with other creditors of ours, any secured or priority rights of such other creditors may cause the proceeds to be distributed to such other creditors before, or ratably with, any distribution made to you on your Note.
Lending Club doesn't have a ton of other creditors, though it has some. There's a term loan led by Morgan Stanley to fund an acquisition, for instance.
Also keep in mind that the notes are not general unsecured obligations of Lending Club: It only pays back a note to the extent it gets a payment from the borrower underlying the note. It's not "credit risk of Lending Club instead of credit risk of borrower"; it's "credit risk of both borrower and Lending Club."
5 DON'T TAKE THIS SERIOUSLY! This is not "leverage ratio" in the bank capital sense, it's just dumb GAAP accounting division. Numbers are from the Lending Club S-1, JPMorgan 10-Q, Lehman 10-Q. Also, this is not pro forma for Lending Club's proposed capital raise, which is supposedly $500 million, even though the use of proceeds is pretty much "give us money, and we'll figure out what to do with it later." Acquisitions presumably. But I guess it would de-lever them.
6 I've done no rigorous work on the accounting for that, I just mean it as an economic matter. Presumably Lending Club would have a loss on the loans and an exactly offsetting gain from the reduction in its liabilities.
Obviously if all the loans go south, then there'll probably be fewer loans in the future, but that's a business risk, not a balance-sheet one.
This glib line of thinking draws on much more sophisticated thinking from Arnold Kling, Steve Randy Waldman, Gary Gorton and his co-authors, etc.
8 This is a central claim of Anat Admati's that I more or less believe.
9 I sometimes imagine that capital requirements are 8 percent. The reality -- with stress tests and risk weighting and G-SIFI surcharges -- is way more complicated than is worth discussing here.
10 A few more quick points:
- Regulatory arbitrage. Lending Club is putting a banking business in a format that is not regulated as a bank, which gets a bad name as "shadow banking" and "regulatory arbitrage." But here it seems pretty benign, right?
- Skin in the game. There's been a big focus on the bank originate-to-distribute model as a problem, because banks make loans with no "skin in the game." So regulators want banks to have to retain a certain amount of the risk of their loans. Lending Club, of course, retains no risk on its loans; it passes the risk to investors. So it has no "skin in the game." What it has is a business: If it does a bad job of underwriting loans, people will stop buying loans from it, and its fee income will dry up. It's interesting to ponder (1) if this is enough to induce Lending Club to do a good job, and (2) why it's not enough to induce banks to do a good job. The answer likely has to do with the difference between regulatorily protected incumbents and startup interlopers.
To contact the writer of this article: Matt Levine at firstname.lastname@example.org.
To contact the editor responsible for this article: Kirsten Salyer at email@example.com.