Sunday, 13 April 2008

Anatomy of a Crisis: The "Credit Woes" of the Summer of '07 – (9)

The “structure of capital and financial markets” is a standard course in every business school. It has a set agenda, designed around descriptions and definitions such as: what is a stock exchange, how corporations raise money and the difference between money markets and capital markets. (The students, like their teachers, rarely appreciate the last point, which is why later in the position of power in financial institutions they think nothing of funding capital market instruments with a money market product such as commercial paper. I mentioned this earlier and will return to it again in some detail because the point needs emphasizing.)

These technical descriptions are at some level important. But they are irrelevant to us. In the context of the discussion of a crisis centered around liquidity and credit risk, the most critical thing we must know about the structure of markets is the function of the tri-party repo market. For that, we need to know the role of broker-dealers and their clearing banks.

Broker-dealers are large retailers of financial products. When you call your broker to purchase a security, say 1000 shares of IBM or $50,000 face value of a 2-year bond, your order is filled from the broker-dealer’s inventory. That is how these retailers make money, by buying and holding inventories and then selling them at a slightly higher price to you. In that regard, they are no different from any large retailer, say a supermarket or a car dealership, only in their case the truck is now a bond; a can of soup, a stock.

Like all retailers, broker-dealers must finance their inventory. Depending on its size, a broker-dealer would hold from tens of millions to hundreds of billions of dollars of securities in inventory. It would simply be impossible to pay to for those securities in full, much less run an economically viable business.

There are thousands of BDs in the US. The largest five – now four – are household names. They are: Merrill Lynch, Goldman Sachs, Morgan Stanley, Lehman Brothers and the late Bear Stearns.

This “description” of the broker-dealers is accurate and factual in every sense. But it, too, suffers from a flaw in that it is not “pointed” enough. It is benign – neutered, if you will. It is a description put together by a clueless finance professor or an ignorant CEO of a large sinking broker-dealer. That is because it takes the focus away from the critical aspect of the business of broker-dealers and, by virtue of that fact, is misleading. To really understand the broker-dealer business, we need to go inside the business. The best way of doing so is to start one. All we need is a telephone and a clearing bank and we are in business. (I am exaggerating only slightly. In addition to an easily obtained license, we would also need some nominal “start up” money, which could be as little as $250,000)

Broker-dealers make money by buying and selling securities. So we immediately get to the business. The rules permit us to buy $10 million worth of US Treasuries with our current equity of $250,000. That is what we instruct our clearing bank to do. The trade is immediately filled. It will settle T+1, i.e., the next day.

We must now turn our attention to the large elephant in the room that we managed to ignore in our initial excitement: we just purchased $10 million worth of securities for which we have no money. We need to come up with the money before the next day settlement.

Our clearing bank will be more than willing to lend us money and take the securities as collateral. It suggests to us, however, that we could more cheaply borrow from an investor. These “investors”, in the parlance of the Wall Street, are large mutual funds, pension funds and insurance companies. They hold daily excess cash for redemptions. They also receive a constant flow of money from new subscribers. Instead of leaving the cash idle, they like to invest it overnight; every bit helps. So they are more than glad to take our securities as collateral and lend us $10 million. The arrangement is no different than buying a house where the purchase is financed by pledging the house as the collateral. Only when the collateral is a security instead of a house, the transaction is called repo. And when a clearing bank manages the transfer of money and collateral, in fact acting as the third party between us and the investor, the transaction is called tri-party repo.

Repo financing is mostly overnight. This means that the loan starts at about 4:30pm on the day of transaction and ends the next day at 8:30am, when the Fed securities wire opens.

Why do we and the investor both insist on an overnight loan? For the investors, they might need the cash the next day. Also, in the overnight lending, the credit exposure to us is minimal; something going wrong with our finances between 4:30 pm and 8:30am of the next day (when the markets are closed) is less likely. (All these references are to the US market. But the principle applies everywhere, as we will see.)

As for us, for the transaction to make economic sense, the interest we are paying for the financing must be less than the interest we are getting from the bond. In a positively-sloped yield curve environment which generally – but by no means always – reigns, the shorter the term of the loan the lower the interest rate. So the overnight borrowing is the cheapest.

That is why tri-party repos are typically overnight. The term of deals might extend longer, up to one month, but not much longer.

Having thus secured the financing, we go home to have a good night's sleep.

Overnight, the technicalities grow in importance.

I will shortly return with the 10th and final part of the series.