My former student writes,

I just read the post–thanks. But I now have more questions. Am I slow, or is the repo market really hard to understand?
When you talk about the original intent of repo, I have a few questions:

1) I understood from Metrick in class that the repo market functions as a type of banking system. Dealers take “deposits” and provide securities as collateral; these deposits are returned at some point with interest. Is that different than what you say the repo system used to be, or is that the same?

2) You say that Gorton/Metrick make no distinction with what the repo market used to do, and what it now does. That investment banks used to keep inventory and sell it, but now they have trading and investment portfolios. What does this mean with regard to the repo market? I.e. how does this change what’s happening?

3) Metrick did say that the securitization of real estate was what created new “safe” securities that people with assets they didn’t want to put at risk demanded. Is this the thing you find inherently dangerous? The idea of manufacturing new, relatively “safe”, liquid assets?

Sorry if these questions are stupid–or even embed inaccuracies/misunderstandings. I think I understand the type of transactions that takes place in the repo market, but I’m not at all sure who are the counterparties on opposite sides of the deal, why they want these deals in the first place for such brief periods, and why this market is as large as it is.

Here are my answers:

In the old days, the investment banks were bond dealers. The difference between a broker and a dealer is that a broker brings the buyer and seller together, while a dealer stands in the middle. For used goods, eBay is a broker. A used-book store is a dealer. It buys books from people and sells them to other people. The books that it has to sell are its inventory.

Suppose that a used-book dealer needs a loan to finance an increase in its inventory. That is, it is going to buy lots of books with the hope of selling them later. It could go to the bank and say, “I want a loan, using the books in my inventory as collateral. That is, if I default on the loan, you can take possession of the books and sell them.”

Bear with me, and assume that the bank is really comfortable with the idea of having books as collateral. One way to structure the loan would be as follows:

1. The bank buys books. It lends cash to the book-dealer. Say it lends $10K to the book dealer. The dealer uses this $10K to buy more books for inventory.

2. One week later, by pre-arrangement, the dealer buys back the books at a price of $10K + ten dollars. The ten dollars is interest that the bank earns for lending the cash.

This transaction is a repo. The book dealer sold the books and repurchased them from the bank. We say the bank made a repo loan of cash to the book dealer. We say that the book dealer lent the books to the bank. We say that the book dealer funded itself using repo.

Now, for the books, substitute bonds. For the bank, substitute any institution with cash to put to work. It could be a corporate money manager, a money market fund, whatever. It could be the Fed, trying to inject more money into the system. For the book dealer, substitute a Wall Street bond dealer. That’s your traditional repo market.

Two things gradually changed over the last thirty years. Both are documented pretty colorfully in Michael Lewis’ first book Liar’s Poker, which I heartily recommend. One thing is that Wall Street firms decided that being mere dealers in bonds was not as much fun as taking positions in bonds. Think of a hedge fund, which buys securities that it thinks will do well and shorts securities that it thinks will do poorly. The big Wall Street firms turned themselves into bond hedge funds.

The second thing is that Wall Street took over the mortgage market. The economics of how this happened are explained in Bob Van Order’s “Duelling Charters” paper that I referred to earlier. The politics of how this happened are at least as important. Lewis talks about it, and so does the more recent book All the Devils are Here by Bethany MacLean and Joe Nocera.

Put thing one and thing two together, and you have Wall Street firms that became giant mortgage securities hedge funds. You also have a subplot of “The Cat in the Hat,” which kind of fits, actually.

Like hedge funds, the Wall Street firms used a lot of leverage. That is, they did not put up much equity capital relative to their long and short positions in bonds and mortgage securities. A lot of their funding came from short-term loans, which were repo loans.

In the old world, a bank issued a mortgage, held the mortgage, and funded it with deposits, which it hoped would sit in the bank for a long time (at least on average). In the 21st century, a mortgage was pooled with other mortgages into a mortgage-baked security, then carved into tranches, and a lot of the tranches were like hot potatoes. An investment bank would hold the hot potato for a week, financed with a repo loan. The maybe it would do it again for a week, or sell it to another investment bank who held it for a week. If the rate of return on the mortgage security tranche thingy were higher than the cost of the repo loan, the investment bank would come out ahead.

So, one could say that these Wall Street dealers-turned-hedge-funds were like banks, with repo loans as their “deposits” and mortgage security tranche thingies as their assets.

A lot of the tranche thingies were supposed to be perfectly safe. That was under the assumption that we would never see house price declines in more than a few isolated locations. As long as they were taken to be safe, they served as good repo collateral, so the repo market could become really, really, big. Much bigger than in the old days.

But, when the subprime mortgage crisis hit the fan, the mortgage security tranche things started to look a lot more like used books than like AAA-rated bonds. That is, your typical money manager with cash to deploy didn’t know how valuable the used books (subprime mortgage security tranche thingies) really were, or whether he could sell them if his counterparty, say, Lehman, could not buy them back to close out the repo.

So as confidence in the underlying security thingies collapsed, and along with it confidence in some of the big repo borrowers collapsed, you had a whole bunch of money managers decide that they did not want to be the ones caught with their pants down in the repo market. Hence the “run on repo.”

What was dangerous about the idea of creating safe securities is that they were not safe, because they could not survive the home price scenario that actually took place.

But where I really part ways with Gorton and Metrick is in my view about the evolution of bond dealers into hedge funds. I think it was bad for the financial system. My reasons have to do with political economy. I think capitalism works when firms can fail. But Wall Street firms are immune from failure, because they are the Fed’s “primary dealers,” and the Fed is not going to let its primary dealers fail. Lehman is the first and last instance of that. So now we have what amounts to to-big-to-fail hedge funds. And that is a recipe for private profits and socialized risk.

Your questions have been good so far. Feel free to probe further.