I was stunned last year when MF Global, a major commodity futures and derivatives broker, declared bankruptcy. At the time, I had an account at the firm, and I wondered how I was going to manage a transfer of my assets to a new brokerage. I wasn’t too worried about my capital though; I had a rough understanding that Wall Street customer accounts are segregated from proprietary brokerage funds, and so I thought that while it might take some time to disentangle my own funds from the bankruptcy, there was a low risk of any kind of impairment of my capital.
Then came another stunning development: the bankruptcy trustee announced that a $1.2 billion shortfall had been discovered in customer accounts. I thought at first this might be a case of flat-out fraud, in which a manager at MF Global had dipped into customer funds in an attempt to save the firm. While it’s still unclear exactly what happened to the customer cash at MF Global, I stumbled on an astonishing alternative explanation that makes plenty of sense and is entirely within the law. As I learned, the segregation of client funds is not nearly so clear cut as I had thought, and in fact, customer funds are routinely posted as collateral by the big brokerage firms.
Before I get into the details how this takes place, let’s take a moment to discuss some terminology. “Hypothecation” is the practice of securing a loan with collateral. For instance a mortgage on a house is secured by the home itself. The house is “hypothetically” controlled by the bank; if the homeowner defaults on the loan the banker can step in and liquidate the house to satisfy the loan. On Wall Street, if you have a margin account, your brokerage firm can hypothecate your securities so that if you fail to meet a margin call the brokerage can liquidate your securities to keep your account solvent. Additionally, a brokerage firm can “re-hypothecate,” or re-pledge, your securities, using them as collateral for its own account.
This is entirely legal and is governed by Regulation T, which are the rules promulgated by the Federal Reserve that govern the provision of credit by brokers. Under Reg T rule 15c3-3, a broker can re-hypothecate up to 140% of the value of a client’s liability to a broker. In the event of the insolvency of the broker, the client then becomes an unsecured creditor as to that collateral. As Joel Greenblatt would say, “Hmmm…that’s interesting.”
On its face, this arrangement makes some sense to me. The broker has extended credit to the client, so it seems reasonable that the broker should then be able to collateralize that loan. I think that 140% of that liability, however, seems excessive to me. Doesn’t that mean that customer equity is implicitly at risk? Why does the broker get to pledge something beyond 100% of the credit it has extended? As I read about re-hypothecation and Reg T, I found that in order to understand how these large brokers use re-hypothecation, I had to brush up on the sale and repurchase agreement, or “repo,” market.
Traditionally, banks make money by taking deposits from individuals while paying, say, a 1% interest rate, and then lending the proceeds out at, say 3%. The bank gets to keep the 2% spread between the two rates. Depositors are looking for a safe place to park their cash and earn a little interest, they can withdraw the money any time, and they can write checks that will be honored. The repo market basically serves the same purpose for large institutions.
In a repo transaction, an institution deposits money with a broker, and in return receives collateral, typically in the form of bonds, that back the deposit. The broker might pay an interest rate (the “repo rate”) of, say, 1% to the institution, might be obligated to repurchase the bonds at some future date, and meanwhile would earn interest on the bonds of, say, 3%. So repo is just a spread game, as with traditional banking. If the broker goes bankrupt, the institution sells the bonds to make itself whole on the value of the deposit, and doesn’t have to get involved in any messy bankruptcy proceedings. Sometimes referred to as the “shadow banking system,” the repo market has exploded over the past decade, with some estimating its size to be about $12 trillion, which exceeds the assets of the entire U.S. banking system of approximately $10 trillion.
In essence, the repo market has become the lifeblood of the financial system, and re-hypothecation has become a significant portion of the repo market. Our good friends over at Reuters sifted through some SEC filings and came up with the following estimates for re-hypothecation totals for a number of firms (figures are in billions):
So re-hypothecation appears to be widespread and commonplace. The benefits to the financial system are that the pledging and re-pledging of assets essentially creates additional synthetic liquidity within the system. Think back to your economics 101 class, when you learned about fractional reserve lending. When a bank takes a $1.00 deposit, it is required to keep, say, 20% as a reserve requirement, and can lend out $0.80. This $0.80 is deposited, and after a 20% reserve requirement, another $0.64 is lent out. So there is a multiplier effect. The same money multiplier effect holds for the re-hypothecation process, in which collateral is pledged, and re-pledged, in a chain of credit.
Back to MF Global: what happened to the firm is that it got involved in a risky repo transaction. The firm bought European bonds, and then used them as collateral for new loans, using the proceeds to buy yet more European bonds. The transaction wasn’t risky in the sense that the firm had exposure to default risk – the European Financial Stability Facility guaranteed them – and they had also engineered it so that their repurchase obligation occurred at the same time the bonds matured, in a so-called “repo-to-maturity.” Instead MF Global was exposed to the risks involved in maintaining the trade. While the spread was profitable, there was the risk of margin calls and other frictional costs. If the bonds decreased in value, MF Global might have to post additional collateral. And in fact, that is what happened. As the European sovereign debt crisis grew, FINRA required MF Global to post more collateral. As MF Global’s leverage reached a reported 40-to-1 ratio, the costs to maintain the trade quickly swamped the firm’s capital base.
So it appears to me that MF Global was relying on the credit it extended to customers, in accordance with Reg T, as a financing mechanism in the repo market. This is just business as usual for any prime broker in the U.S. When the firm got overleveraged in its big repo bet, the firm’s counterparties merely did what any good capitalist does when a borrower defaults: they exercised their right to take possession of the collateral on the loan. But the problem for me was – that collateral call came out of my personal account and now I’m just another unsecured creditor!!!
At the end of the day, I have so far recovered about 70% of the liquidation value of my account at the time of the bankruptcy. I just sent in my paperwork to the bankruptcy trustee, so with any luck, maybe I’ll get some more back, but I’m not holding my breath. Perhaps I will get back everything up to 140% of the value of my margin debt balance? We’ll see. For me, it was not the end of world, since I didn’t have any risky positions open at the time of the bankruptcy that moved against me, and the account was fairly small. But that’s not true for everyone.
MF Global did a big business in commodity futures, and many of the firm’s clients were farmers trying to hedge their businesses. When the bankruptcy occurred, there was chaos in these accounts, with clients unable to trade, even as the commodity markets traded as usual, which I’m sure wreaked havoc on many accounts. How can these customers reasonably be expected to participate in the commodity markets, when these risks exist? And so while I’m not expert on Fed policy, there is something strange about a rule that allows this kind of thing to happen. I appreciate that re-hypothecation provides critical liquidity for global financial markets, but it doesn’t seem fair that a Midwestern farmer will get wiped out because his broker made an overly aggressive leveraged repo bet. So maybe the Fed needs to fix Reg T. But I’m not holding my breath for that either.