Paul Volcker had a simple idea: get the government out of the hedge fund business. From his simple idea was born a simple proposal: ban proprietary trading and investments in hedge funds at government-backstopped banks. Congress agreed, to a point, and passed the “Volcker Rule” as section 619 of the Dodd-Frank Act.
The draft of the Volcker Rule, which grew from a three-page proposal to a 300+ page behemoth, was released by the regulatory agencies this October. The draft rule grants a number of exemptions from the proprietary trading restrictions. One of our major concerns is the blanket exemption for repurchase agreements (“repos”). The exemption isn’t mentioned in the statute, and for reasons discussed below it seems to defy the intent of the rule. In our eyes, the presence of such an overbroad exemption is profoundly disappointing. Whose interests are the regulators serving?
Repo lending is best described as the financial equivalent of a visit to the pawnshop. An asset is deposited with a lender in exchange for cash, with an agreement that at some point in the future a slightly larger amount of cash will be repaid and the initial goods returned. The pawnshop provides a way to exchange a valuable-but-illiquid asset (a grandfather’s watch) for a source of short-term liquidity (next month’s rent).
Repo provides much the same function for banks—except that instead of a watch, one deposits bonds or other securities, and instead of next month’s rent, it is used to fund today’s trading activity. Repo financing is an integral component in global financial markets. It is the major source of funding for trading positions in global Bond and Derivatives markets. The regulators argue that repos are simply secured-lending arrangements, and that the securities underlying repo transactions should be excluded from the definition of proprietary trading assets. If this were the case, why grant such a broad exemption?
In the Spring 2009 issue of Stanford Lawyer, Berkshire Hathaway Chairman Charles T. Munger described repos’ role in terms sharply divergent from those implied by the regulators presumption of repo as a benign banking activity in defense of the exclusion:
Munger’s comment should serve as sharp reminder that repo has hardly been a harmless bystander in recent systemic shocks. The fall 2008 ‘liquidity’ crisis was largely the result of a breakdown in the repo market. Yet the draft rule sees fit to license any and all repo trades in the name of innocuous ‘securitized lending’. These trades are hardly innocuous. Repos are intimately involved with prop trading books, overleveraging, and regulatory evasion, and are exactly the kind of mischievous and systemically risky transactions the Volcker Rule was meant to restrict. Starving the beast, rather than keeping it well-fed with regulatory exemptions, was the intent. That intent is damaged, if not entirely compromised, by the proposed exemption.
Repos are not simply secured lending anymore
If the last twenty years of financial history have taught us anything, it is that every instrument can be restructured into something barely resembling the core asset it purports to be based on. MF Global’s recent implosion—replete with the apparently-irretrievable loss of still-unknown amounts of shareholder money—offers an abject lesson.
“MF Global used a version of the off-balance-sheet move called ‘repo-to-maturity.’ The firm offered billions of dollars in sovereign debt as collateral on a series of loans designed to expire at the same time as the collateral itself. With the collateral and the loans coming due simultaneously, MF Global might never take possession of that debt again. That entitled the firm to count those as sales, and moved $16.5 billion off its balance sheet, most of it debt from Italy, Spain, Belgium, Portugal and Ireland.
…To top it all off, the accounting for these deals added $124 million in financing payments to the firm’s revenue over the last four quarters, according to SEC filings, firm documents and people close to the firm.”
This case illustrates one of the principal problems with the proposed rule’s interpretation of repos as secured loans: this interpretation fails to account for the fact that repos often take on the characteristics of proprietary trading. In the case of MF Global, the ambiguity around ownership of the underlying asset was exploited to destabilizing effect. Plain vanilla repo is unambigous about ownership of the asset during the secured lending period.
We would hope that regulators might have learned their lesson about assuming that all repo was plain-vanilla repo. Although there is a standard contact for repos similar to the ISDA master agreement, dealers are free to be extremely creative in their contracts. This potential has been abused, notably, with the Lehman Repo 105 and the MF Global reverse-repo account ploy.
At this point, little data is available on the number and nature of repo-to-maturity and other repo-as-prop-trade transactions in the marketplace, or on the specific counterparties involved. That said, in the same Reuters article it is noted that Merrill Lynch has “disclosed that they use the [repo-to-maturity] structure.” We feel it is very likely that other US banks have repo trades of this type on their books, making this transaction and the concerns it raises a potentially significant issue in both the Volcker Rule’s implementation and our future systemic stability. This is known to the regulators–Mary Schapiro said on December 1st that she is in talks with the FASB about whether the repo-to-maturity loophole MF Global used requires further disclosure, and the FASB has a new rule closing the Lehman 105 loophole. This seems to amount—as so frequently the case—to a closing of the stable doors after the horses have bolted.
These tardy responses to long-standing abuses suggest that the regulators had the authority but failed to use it to shut down repo abuses that resulted from reliance on varying interpretations of accounting rules. The SEC could have shut down the Repo 105 abuse by rule long before Lehman collapsed. Perhaps as damning, regulators are fully aware that a panoply of bad behavior by the banks has been carried out under the guise of repo. If past is prologue, Volcker’s repo exclusion perpetuates rather than limits the ability of firms to game the regulatory and accounting rules.
For all these reasons, the short-sighted reasoning the regulators present in support of the blanket exclusion should give us pause. They are aware—as indicated by Schapiro’s comment—that some portion of the repo market is structured, and some of those structures likely contain elements that would cause them to qualify as trading assets. Furthermore, they are aware that repos have been used multiple times to hide risky proprietary positions (or losses) that might otherwise be troubling to the marketplace. Taken in sum, this suggests a bizarre and seemingly willful amnesia on the part of the drafters.
How to run a Proprietary Trading book with Repo
Even if repo had not been abused in the past, we would be alarmed. The blanket exemption opens up enormous loopholes where banks could use repo to structure prop trades. Here are just a few examples of ways to do so:
- Shorting: A bank enters a reverse repo with Counterparty X using bonds as collateral. The bank immediately sells the bond, anticipating that the price of the bond will decline. When it is time to return the bonds to X, the bank buys them from the open market, hoping to benefit from price depreciation in the bond. This is essentially a short position on the bond, wrapped up in a repo.
- Basis Trades: A bank enters a reverse repo with Counterparty X, using securities as collateral. Later, the bank (the repo lender) returns “substantially equivalent” securities instead of the original securities. Since the proposed rule uses the broadly-interpretable ‘stated asset’ in the definition of repo, it seems the rule would allow the bank to return a “similar” asset instead of the original one. The bank is essentially going long the initial security it takes in as collateral, and short the “substantially equivalent” security that it will eventually return to Counterparty X.
- Put Options: A bank repos some securities in exchange for cash. The repo lender takes the securities. Later, the repo lender fails to return the securities, either due to an outright default or pursuant to an embedded right to refuse delivery. The bank has essentially sold the securities. The CFTC has actually highlighted this possibility. They said: “under new bank capital standards, a sale of securities subject to a repurchase agreement with a unilateral right in the transferee to refuse to return them could be construed to be the granting of a put from the perspective of the original ‘seller.’ This would attract a capital charge.”
- Interest Rate trades: A standard repo trade is a rates trade at its core, as the repo rate is effectively the interest on a collateralized loan. Booking a repo looks like three separate trades:
- a sale of securities
- a future purchase of the same securities, and
- a swap, the cashflows of which are the repo rate.
The purchase and the sale of the securities net out, leaving a (proprietary) directional swap.
- CDS: A bank wants to speculate on the failure of a Counterparty X, so it enters into a repo transaction with X with a significant haircut. The bank lends X some cash, and demands collateral with significantly higher value than the cash. If X defaults, the bank keeps the collateral and locks in a huge profit. (This is functionally a CLN with X, referencing X)
Alongside this menu of desired proprietary exposures that can be smuggled under the rule as basic “repos”, the advent of financial engineering has led to repos that are designed to house many more types of risk. Below is a list of some of the major structured repo categories that contain elements of proprietary trading:
- Cross Currency Repo: By accepting collateral denominated in a different currency than that of the cash exchanged for it, a bank can embed almost any desired FX exposure into a Repo.
- Callable Repo: By including an early termination option for the repo lender, any repo swap can be made to include an option on that swap. If rates go up, the repo lender can exercise its option, recall the collateral, and re-repo at a higher rate.
- Total Return Swap: A more generic way to structure a CDS into a repo, the repo rate in this structure is typically some spread to LIBOR, where the spread is determined primarily by the credit risk of the collateral at the time of the trade. In essence, the Bank is lending money in exchange for collateral AND gaining exposure to the credit risk of the collateral.
It is not difficult to see how banks can package almost any kind of risk into a repo by modifying the conditions of the ‘repo rate’ within them.
Repos are already excluded under the Liquidity Management exemption
The blanket exemption given to repo agreements is particularly puzzling—and perhaps sinister—in light of Section __.3(b)(2)(iii)(C) of the Proposed Rule, which states that a bank’s trading account is not subject to the Volcker Rule if that account contains financial positions undertaken for ”the bona fide purpose of liquidity management and in accordance with a documented liquidity management plan.” That section goes on to define five criteria the liquidity management plan must meet. These include requirements that transactions be for managing liquidity and not be prop trading, that positions taken are “highly liquid” and not expected to earn “appreciable P&L”, and that the amount spent on liquidity management be “consistent with the banking entity’s near-term funding needs, including deviations from normal operations.”
A classic plain-vanilla repo agreement would certainly meet the above criteria. The regulators—or those lobbying them—must be concerned that there are more exotic cases where the use of repos by banking entities do not adhere to the criteria. This makes the blanket repo exclusion all the more inexplicable and troubling.
Now’s the time to speak up
There is, however, some good news here. This draft of the Volcker Rule is, well, a draft! The public has the right to submit comments on this draft up until January 13th, 2012, and these comments will become a part of the public record.
Our objections boil down to:
- As evidenced by Lehman’s “Repo 105” and MF Global’s “repo-to-maturity,” repos can and have been used to obscure and misstate exposures, thus hiding growing insolvency.
- Repos can be structured to hide proprietary trades, thus subverting the rule and violating the law. There is no mention of special status for repos in the statute.
- Typical repos would be already covered under the exemption of accounts used for Liquidity Management, so there is no reason to carve out a specific exclusion for repos.
- The repo exclusion in the proposed Volcker rules impacts non-bank repo counterparties, primarily Money Market funds. The exclusion has a systemic effect on markets in general and should be an issue of public concern.
One of the questions the regulators have asked for public comment on in the Proposed Rule is about repos:
“Question 30. Are the proposed clarifying exclusions for positions under certain repurchase and reverse repurchase arrangements and securities lending transactions over- or under-inclusive and could they have unintended consequences? Is there an alternative approach to these clarifying exclusions that would be more effective? Are the proposed clarifying exclusions broad enough to include bona fide arrangements that operate in economic substance as secured loans and are not based on expected or anticipated movements in asset prices? Are there other types of arrangements, such as open dated repurchase arrangements, that should be excluded for clarity and, if so, how should the proposed rule be revised? Alternatively, are the proposed clarifying exclusions narrow enough to not inadvertently exclude from coverage any similar arrangements or transactions that do not have these characteristics?
If you have additional arguments against repos as they are included in the Volcker Rule, we invite you to leave them in the comments. Though we’d far prefer you’d address Question 30 in a comment letter of your own.
If you’d like to join Occupy the SEC in their efforts to sniff out and highlight other loopholes in the Volcker Rule, you can find out more at http://occupythesec.org.