My Senate Homework from "SEFs – A Progress Report"

By | August 12, 2011

Also published at

On June 29, I testified before the Senate Committee on Banking, Housing and Urban Affairs Subcommittee on Securities, Insurance, and Investment. Chairman Tim Johnson (D-S.D.) followed up with some questions. Following are Johnson’s questions and my answers formally submitted for the record.

1. The reporting requirements that the CFTC/SEC has proposed for all SEFs transactions will require virtual real-time reporting of key transaction data. Won’t liquidity providers (i.e., dealers) increase their bid/ask spreads to reflect the increased risks associated with communicating key data to the marketplace (since dealers will not be able to hedge these positions before they are front-run)? In order to justify this risk, won’t liquidity providers necessarily pass these increased costs to end-users? In your view, does moving to SEFs justify these increased costs and reduced liquidity?

In the most liquid products, reporting requirements will be of little long-term consequence. To those in the market already, pricing data is in fact quite transparent and so additional dissemination will have little impact on spreads. For less liquid products however, it is very likely that a risk premium will now be embedded in the quoted price. However, even in today’s market, brokers often hedge new positions using other instruments such as futures and bonds to avoid being “picked off” by other market participants who are aware a trade just took place. This approach will become more prevalent in the new world.

It is important to note however that as the market becomes more electronic and more efficient, new liquidity providers will emerge to keep the prices between futures, bonds and swaps very closely aligned. This will only make it easier to hedge a new swap position elsewhere with little market impact.

2. The CFTC’s proposed SEFs rules would require that market participants put out a minimum of five Requests for Quotes before they complete a transaction. Given that most of the OTC market currently trades in a non-standardized form, wouldn’t this requirement to garner five RFQs cause participants to share important information to the marketplace, which the market could use against that participant? In other words, wouldn’t this requirement to trade with the RFQ model increase bid/ask spread for end-users and potentially increase volatility?

It is first important to note that the majority of trading in interest rate swaps and credit default swaps occurs on standardized contracts. Vanilla U.S. dollar interest rate swaps of standard durations (2-yr, 5-yr, 7-yr, 10-yr, 30-yr) and index credit default swaps are in fact viewed as quite liquid by market participants. TABB Group’s conversations with buy side traders, bulge bracket swaps dealers and mid-sized swaps dealers confirm this point.

That said, these same market participants all believe that requiring an RFQ to be sent to five market participants would in fact widen spreads, decrease liquidity and drive trading to other products that did not have the same requirement. TABB Group believes that five is an arbitrary number and one that is not supported by historical precedent in this or any other financial market. Yet although we firmly believe a principles-based approach to SEF regulation, one in which they are free to compete with each other based solely on their merits is best, in keeping with the goals of Dodd-Frank changing the RFQ requirement to read “more than one” would act as a reasonable compromise that would not impact the majority of RFQ trades done today.

3. If the CFTC defines the size of a “block trade” too narrowly, then very few trades will be permitted off the SEFs. Given that most of the interest rate and credit default swaps trade in blocks too small to qualify as “blocks” under the new rules, wouldn’t a phased-in approach be more appropriate than a cold-turkey move to the various SEFs rules? With regard to the CFTC’s block rules, does the CFTC’s one-size-fits-all approach make sense? Not all swaps have the same risk characteristics and lumping all interest rate swaps into one bucket for blocks (and similarly for CDS) does not seem consistent with market convention.

Setting the block trade size as a multiple of the current average trade size is unreasonable. The majority of swaps trades done today are, in fact, block trades. The average size of an interest rate swap is $129 million – but that is because much of the trading in this market is done by financial and commercial end users hedging real positions. That is in stark contrast to the highly electronic futures market where most market participants are looking for short-term exposure to a particular reference entity. One can reasonably conclude that once the vanilla interest rate swap market is centrally cleared and traded electronically, the average trade size could decrease by a factor of 10. That said, block sizes must be forward looking and take into account how these products are used and by whom.

4. Do you envision that block trades will be treated differently by SEFs versus DCMs? If so, how and why?

Block trade rules are and should be focused on reporting and not on method of execution. As stated in our testimony, TABB Group strongly believes a principals-based approach is best for the swap execution space allowing SEFs and DCMs alike to compete for liquidity based on trading mechanisms provided, price, technology and other competitive factors. That said, the time delay for reporting a block trade as well as the size of a block trade must be consistent regardless of where a trade is done. If one venue sets the block size lower than another, we will quickly see liquidity move to the venue with the lower threshold. So ultimately, execution method should be left open to the venue but the block definition must be consistent system wide.

5. The margin calculation for SEFs (which requires a minimum five-day liquidation period) v. DCMs (one-day liquidation period) has a significant impact on required margin. Why were these arbitrary liquidation periods established?

Market convention uses liquidation periods of between one and 10 days. Five-day liquidation is not uncommon. The liquidation period used to calculate margin is influenced by liquidity. The lower expected or perceived liquidity of an instrument, the wider the liquidation period (up to 10 days). But since margin can be changed often, it is not critical to fixate on a particular look-back period as long as it is in the acceptable range. Furthermore, it makes sense that margin levels for newly-clearable products would start out at conservatively.

The longer the duration of a contract, the greater the risk to the clearinghouse. As swaps tend to be of much longer duration (the 10-year interest rate swap is one of the most common) as compared to futures (often 3-6 months), the risk – and hence the margin requirements – are greater for swaps. TABB Group research has found that on the short end of the curve, margin levels are in fact quite similar for swaps and futures. But as duration increases the gap widens considerably.

6. We understand the CFTC is considering a different segregation regime for customer margin for SEFs v. DCMs. Why? What is the benefit?

Independent of the vernacular, there are two margin segregation schemes being contemplated. One is like the futures markets in which customer funds are comingled in an omnibus account of the clearing member. The problem with this structure is that customers do not want to have exposure to one another for OTC derivative trades. The other segregation method is described as legally separated but operationally comingled. This format is intended to provide the margin benefits of the futures model without the exposure to defaulting parties.

In all cases, the benefit of pooling more margin funds is that it gives the clearinghouse the potential to offer margin offsets between more products, such futures and swaps (which are often used to hedge one another). In short, fund segregation regimes can determine the level of margin offsets (offered to spread products), and margin offsets are the primary key to lowering and ultimately minimizing the oncoming burdens of initial margin requirements for OTC derivatives.

7. Why do the proposed SEF rules not allow for derivatives voice trading?

Yes, and they should. Swap transactions are often complex and very large. Following TABB Group conversations with real-money buy side accounts, it became quite clear that the ability to speak with a broker is critical in the trading process for many. As a case in point, there is a reason why your average retail investor is still willing to pay $50 per trade to call in an order even though trading online is available for under $10 per trade. That said, we firmly believe that even for transactions discussed over the phone, prompt entry into an execution platform for reporting purposes is critical to the transparency and ultimate success of the swaps market going forward.

Have something to say about what I wrote? Please comment...