Monday, October 10, 2011

On the Leaked Volcker Rule

The American Banker leaked part of a draft of the regulators’ proposed Volcker Rule (pdf) last week, which has caused quite a stir. The first thing to note is that the American Banker did not leak the most important part: the text of the proposed rule. Instead, they leaked the “Supplementary Information” (which I call just the “Supplement”), the core of which is a lengthy, section-by-section analysis of the proposed rule. In addition, the leaked portion does not include the Appendices to the proposed rule, which, from reading the Supplement, appear to be very important — Appendix B, for example, contains a “detailed commentary regarding how the Agencies propose to identify permitted market making-related activities,” which is a core issue.

First I’ll give some general thoughts on the proposed Volcker Rule, and then, because I’m such a generous guy, I’ll go ahead and highlight some of the most important pressure points in the proposed rule.

General Thoughts on the Proposed Volcker Rule

In general, the proposed Volcker Rule appears to be very good: it’s a serious effort by a group of smart, market-savvy people to draw a workable distinction between market-making and proprietary trading. The regulators recognize the importance of both market-making and hedging, but they also recognize (most of) the places where market-making and hedging can bleed into proprietary trading. And in those situations, the regulators realize that any effort to distinguish impermissible prop trading from permissible market-making or hedging will — quite appropriately — require a very fact-intensive inquiry. That said, I’m still going to have to withhold my final judgment until I see the actual text of the proposed rule.

Also, even though the WSJ keeps trying to gin up controversy over the proposed Volcker Rule allowing hedging on a portfolio basis, the regulators note in the Supplement that allowing hedging on a portfolio basis is “consistent with the statutory reference to mitigating risks of individual or aggregated positions” (emphasis in original). I explained this a couple of weeks ago; it is a faux-controversy. Moreover, prohibiting banks from hedging on a portfolio basis is a monumentally stupid idea in the first place — it would make risk managers’ jobs 100 times harder, introduce all sorts of new risks into banks’ books (counterparty risk would skyrocket), and dramatically raise hedging costs. This is one thing that the statutory text of the Volcker Rule actually got right.

Some Pressure Points in the Proposed Volcker Rule

Now, to the nitty-gritty of the proposed rule. Here some of the major pressure points in the proposed Volcker Rule that I see:

1. Hedges must be “reasonably correlated” to the underlying risk: In defining “risk-mitigating hedging activity,” the rule requires that the hedge be “reasonably correlated” to the underlying risk(s). The Supplement implies that the correlation must be reasonable at the outset of the hedging transaction, which is absolutely appropriate — a lot of times, you think a trade will be a good hedge when you put the trade on, but because of circumstances beyond your control, it turns out not to be a very good hedge (e.g., liquidity may unexpectedly dry up in either the underlying or the hedge, screwing up the normal correlation).

The real question here is: in normal market conditions, when does the correlation between the hedge and the underlying become “unreasonable”? In other words, how much leeway will banks have in determining how to hedge their books? Say a bank enters into a swap that only hedges 50% of the DV01 of the underlying bond. Would that be considered “reasonably correlated” to the underlying risk? (Obviously, I’m simplifying my examples for illustrative purposes.) After reading the Supplement, I strongly suspect that I know what the regulators’ answer would be: it depends on the particular facts and circumstances. If, for example, the swap was coupled with another transaction that hedged the remainder of the DV01 of the underlying bond, then both transactions would be permitted, because when viewed together, they were both part of a legitimate hedging strategy.

The Supplement also hints at what regulators would NOT consider to be “reasonably correlated” — it states that “[a] transaction that is only tangentially related to the risks that it purportedly mitigates would appear to be indicative of prohibited proprietary trading” (emphasis mine). I think it would be a stretch to say that regulators intend to consider any transaction that’s more than “tangentially related” to be a “reasonably correlated” hedge. But this at least indicates that regulators won’t simply accept a hand-waving, “trust me, they’re related” response to inquiries about the appropriateness of a hedge.

In the end, what we know is that a permissible hedge must be less than “fully correlated” but more than “tangentially related” to the underlying risk, and that if the appropriateness of a hedge is questioned, it will be a very fact-intensive inquiry. Which, by the way, is the way it should be.

2. “Additional significant exposures”: The proposed rule prohibits hedges that themselves introduce significant, unhedged exposures. However, the Supplement also states that:

“[T]he proposal also recognizes that any hedging transaction will inevitably give rise to certain types of new risk, such as counterparty credit risk or basis risk reflecting the differences between the hedge position and the related position; the proposed criterion only prohibits the introduction of additional significant exposures through the hedging transaction.”
There are actually three potential flashpoints in this prong. The first flashpoint is what constitutes “additional significant exposures.” How significant does the new risk that the hedging transaction is introducing have to be before regulators will require it to be hedged as well?

The second flashpoint is what constitutes mere “basis risk,” and what constitutes an impermissible residual risk. The Supplement says that a hedge that merely introduces counterparty or basis risk is permissible. Here’s what I would tell our trading desks if I were still working at an investment bank: start calling every residual risk a “basis risk.” Basis risk evidently doesn’t need to be hedged under the proposed Volcker Rule, regardless of how significant the exposure is. So if you want to profit from the price movement in a certain risk, then just partially hedge the risk with another transaction and call the residual risk “basis risk.”

The third flashpoint has to do with when the “additional significant exposure” must be hedged. The Supplement states that if a hedge introduces a significant new exposure, then the exposure must be hedged “in a contemporaneous transaction.” Assuming that regulators will allow banks some time to hedge the new exposure, the question becomes how much time will they have to hedge the new exposure? An hour? A day? A week? I strongly suspect that the regulators’ answer will be that banks will have to hedge the new exposure “as fast as humanly possible” (not in those words, obviously — the legislative language will probably be something like “as quickly as technologically practicable.”)

3. “Bona fide liquidity management”: This is where I would go first if I was trying to circumvent the Volcker Rule. The statutory text of the Volcker Rule defines proprietary trading in a very roundabout way, such that the real definition of proprietary trading is in the definition of a “trading account.” However, the proposed rule provides an exclusion from the definition of a “trading account” for accounts that are use “to acquire or take a position for the purpose of bona fide liquidity management, so long as [five] important criteria are met.”

The reason I would go here first if I was trying to circumvent the Volcker Rule is that if a trade could fit under the “bona fide liquidity management” exclusion, there would be no need to bother with any of the more complicated “permitted activities” exceptions, and evidently, no need to report nearly as much, if any, quantitative trading data to regulators.

The proposed rule requires that trades done under the liquidity management exclusion be done according to a “documented liquidity management plan” that meets five criteria. But none of the five criteria in the proposed rule appear to me to be prohibitive if a bank wanted to use the liquidity management exclusion for prop trades. The plan has to “specifically contemplate and authorize any particular instrument used for liquidity management purposes” — fine, just write a liquidity management plan that contemplates the use of a (very) wide range of instruments (a lot of instruments have reasonably liquid markets in normal times). The second criterion basically requires that an instrument used for liquidity management not be used “principally” for prop trading purposes, which is easy, since prop trading is prohibited regardless of whether it’s the “principal” purpose of the instrument.

The third criterion requires the liquidity management plan to be “limited to financial instruments the market, credit and other risks of which are not expected to give rise to appreciable profits or losses as a result of short-term price movements.” This criterion simply can’t be enforced terribly stringently — even Treasuries, which are the core of any serious liquidity pool, often experience significant short-term price movements. Fourth, the plan would have to limit liquidity management positions to “an amount that is consistent with the banking entity’s near-term funding needs.” This also can’t be seriously enforced, because it would directly conflict with Basel III’s new Liquidity Coverage Ratio (LCR), and cautious liquidity management in general. Finally, the plan would have to be “consistent with the relevant Agency’s supervisory requirements ... regarding liquidity management.” Seeing as the new liquidity rules set a floor on a bank’s liquidity management, and not a ceiling, using instruments that don’t qualify for the LCR in a broader liquidity management plan would certainly still be “consistent with” the regulators’ liquidity requirements.

4. “Near term” / “Short term”: The statutory text of the Volcker Rule effectively defines a proprietary trade as any trade done “principally for the purpose of selling in the near term.” While the Supplement doesn’t provide much detail on what constitutes “near term,” it does hint at an answer: 60 days or less. The proposed rule will apparently include a rebuttable presumption that any account used to take a position that is held for less than 60 days will be considered a “trading account.” Therefore, it stands to reason that accounts which are used (exclusively) to take positions that are held for longer than 60 days will not normally be considered “trading accounts,” and thus not subject to the Volcker Rule. But, of course, I strongly suspect that the regulators will say that this determination is ultimately going to be based on the particular facts and circumstances of the trade.

Anyway, there are a few more pressure points like this in the Supplement, but that’s all I have time for right now.