Keynote Address of Commissioner Brian D. Quintenz before the Smart Financial Regulation Roundtable
A Better List
November 2, 2017
Thank you for that very kind introduction and a big thanks to the Mercatus Center at George Mason University and the Institute for Financial Markets for hosting such an amazing conference. It’s an honor to be here with you all tonight. Before I begin, let me quickly say that the views contained in this speech are my own and do not represent the views of the Commission.
In thinking about what I wanted to touch on tonight, I was struck by how much we can learn about smart regulation from a children’s story that I recently read to my two children. You may be familiar with a popular series of children’s books called Frog and Toad Together. The books describe the adventures of two best friends, Frog and Toad. In one particular story, Frog and Toad Together: A List, Toad wakes up one morning and decides to write a list of things to do for the day — wake up, eat breakfast, get dressed, go to Frog’s house, and so on.1 After dutifully completing his first couple of items, Toad goes over to Frog’s house, and asks Frog to go for a walk with him, as that is the next item on his list. Just as Toad is crossing off “Take walk with Frog” from his list, a gust of wind blows the list right out of Toad’s hands. Frog shouts to Toad to hurry and run with him to catch the list. But, Toad protests, explaining that he cannot possibly run after the list because running after his list was not one of the things that he had written on his list of things to do! With the list lost and its contents forgotten, Toad insists that he and Frog sit and do nothing until night falls.
Toad’s response seems nonsensical. If the purpose of Toad’s list is to ensure he accomplishes each item on it, then of course he should run after it. In many respects, the CFTC under the prior administration was not unlike Toad. That CFTC could have focused on any number of logical reforms designed to strengthen and revitalize our financial markets under the mandates of Dodd-Frank. But, in reflecting back on those seven years following Dodd-Frank’s passage, those types of policy objectives weren’t on the CFTC’s list of things to do.
The CFTC rushed to finalize and implement many rules, sometimes forsaking the objectives those well-crafted rules would have addressed. I will take a different approach during my tenure. I have been consistent and clear that my priority as a Commissioner is to ensure the CFTC focuses on risk by identifying risk appropriately and targeting that risk specifically. Smart regulation, which prioritizes a thoughtful analysis of policy goals, regulatory costs, well-calibrated thresholds, and corresponding impacts on incentives, will best accomplish this.
Goal-oriented or risk-mitigating public policy often involves formulating thresholds. If such thresholds are met, a certain population is assigned to bear the obligations and regulations necessary to affect the stated goal or mitigate the targeted risk.
For example, to ring-fence risk, regulators must: 1) understand, assess and specify risks, 2) determine if a regulatory solution is needed, 3) calibrate the regulations, costs, and obligations to target those risks, and 4) identify the types of entities potentially responsible for those risks that should bear those obligations. To effect this, regulators must create a threshold or metric to capture that narrowly-defined population exclusively. This helps ensure that regulatory burdens are not imposed on a broader population.
Trouble occurs when a threshold captures a broad population unassociated with an identified risk or unable to bear the cost of the regulation, or both. Regulations that capture entities that pose little risk and regulate them like high-risk entities penalize and disincentivize low-risk activity. This imbalance is often the result of a one-size-fits-all policy approach.
The CFTC’s de minimis threshold for swap dealer registration is Exhibit A of the prior CFTC leaderships’ one-size-fits-all philosophy. It imposes broad costs and regulatory burdens on generic activity without regard for achieving stated policy goals and mitigating identified risks. In its current form, the de minimis threshold does not serve the end-users for whom these marketplaces are designed and meant to benefit. Rather, it acts in direct opposition to their interests and well-being.
But first, a little background. Title VII of the Dodd-Frank Act enacted a number of reforms designed to reduce risk, increase transparency, and promote market integrity. One of the most significant Title VII provisions calls for the registration of entities dealing swaps. Once an entity is designated as a swap dealer, it becomes subject to the extensive costs of swap dealer regulations. Recognizing that not all swap dealing activity warranted such extensive regulation, Congress crafted an exception to the law excluding entities with de minimis swap dealing activity.
In 2012, the CFTC issued a final rule further defining when an entity must register as a swap dealer and the de minimis exception. The final rule established a five-year phase-in period during which the de minimis threshold was $8 billion or less in gross notional value of swap dealing activity measured over the prior year. At the end of the phase-in period, the de minimis threshold automatically decreased to $3 billion, unless the CFTC took prior action.2 The CFTC has twice delayed that reduction, and for good reason. Notional value thresholds of $8 billion and $3 billion neither mitigate systemic risk nor justify the costs of swap dealer registration.
Some have described the current de minimis threshold of $8 billion as a “loop hole.”3 In reality, the threshold’s reduction to $3 billion would create a “black hole,” sucking in community banks and end-users who pose zero systemic risk. At the center of that black hole lies an enormous set of costs: up to $1 million of annual membership fees with the NFA4; $20–$100 million in minimum capital requirements5; clearing and margin costs; IT costs for trade processing, reporting, confirmation, and reconciliation activities; costs to create and send clients daily valuation reports; costs for recordkeeping obligations; third party audit expenses; legal fees to develop and implement business conduct rules and many, many more. If that sounds like a big bill, it is. A prominent economic research firm estimated the cost of swap dealer compliance over ten years at $390 million … per firm.6
Small banks and businesses cannot withstand such costs. Compared to the costs of swap dealer registration, the revenue from $3–$8 billion of swap dealing activity is immaterial—swap transactions are not a major source of revenue, and in fact, fees usually cover only the cost of the service.7 And yet, a firm’s single alternative to registration is to limit client-driven, risk-hedging swap transactions in order to avoid exceeding the de minimis threshold, thereby losing business to larger competitors. The government is putting these smaller client-serving institutions between Scylla and Charybdis: choose between saving your business or helping your customers.
Why have we found ourselves in this Cornelian dilemma? Put simply, poor public policy choices have burdened too many market participants with costs disproportionate to the risks they pose and the possible benefits of their registration.
Dodd-Frank advanced three main policy objectives for swap dealer registration: systemic risk reduction, counterparty protection, and enhanced swap market transparency and efficiency.8 Those policy objectives must also be considered in conjunction with the policy objectives of an appropriately calibrated de minimis exception: providing regulatory certainty to market participants, allowing limited ancillary swap dealing in connection with other client services, encouraging new entrants into the swap markets (potentially fostering competition, lowering costs for end-users, and diversifying risk among a greater number of market participants), and regulatory efficiency.9
When the current de minimis threshold policy was adopted, the CFTC provided no evidence that either an $8 billion or $3 billion notional value threshold would advance those policy objectives10— an unsurprising fact given the thresholds were finalized without the benefit of any swap data.11 At that time, the costs and burdens of swap dealer regulation could also not be weighed against any concomitant policy benefits because many of the substantive regulations had yet to be proposed.
As I describe in further detail below, I believe the Commission must better align the criteria for the de minimis threshold with the costs of swap dealer registration, specifically the largest costs targeting systemic risk. It is my hope that with the benefit of updated and improved data as well as five years of swap dealer regulatory experience, this can be accomplished quickly.
Limitations of Relying Solely Upon Notional Value
Notional value’s deficiencies as a metric make any registration threshold based solely upon it unlikely to advance the policy goals of swap dealer registration. Notional value is an incredibly poor measure of activity and an almost meaningless measure of risk.12 In fact, $8 billion in notional value of swaps can represent a very small amount of risk—in some cases almost none—which is exactly why smaller banks and businesses might easily exceed a $3 billion or even $8 billion threshold. The following over-simplified examples involving interest rate swaps (IRS) illustrate the disconnection between notional value and risk.13
In an IRS, one side pays a mutually-agreed fixed interest rate while the other side pays the market’s floating interest rate over the life of the swap.14 The party which pays the floating rate faces risk from rising interest rates during the swap’s life (the swap’s tenor).
In our example, let’s say Party A wants to hedge its interest rate risk by entering into a one-year, $100 million (notional value) IRS with Bank X. Over the life of the IRS, Bank X will be responsible for payment of a floating rate to Party A and Party A will be responsible for paying a fixed rate to Bank X. So, if interest rates rise, Bank X may lose money due to the increased floating rate it will owe to Party A (whose rate is fixed).15 What is Bank X’s risk from this transaction? Is it $100 million? Not even close.
Bank X’s market risk is determined primarily by how much interest rates rise during the swap’s life. Let’s assume that immediately after the IRS’s execution, interest rates rise by 1% and then remain unchanged over the life of the one-year swap.16 Because the floating rate that Bank X must pay Party A has risen by 1%, Bank X faces a loss. But, how much of a loss? Roughly speaking, on the one-year, $100 million IRS, Bank X loses $1 million—the additional 1% in interest Bank X owes to Party A—a far, far cry from the massive-sounding $100 million notional value figure of the swap.
To extrapolate notional value’s deficiencies, imagine if Party A wants to enter into a $100 million interest rate swap, but only for one week.17 Let’s assume the same immediate 1% rise in interest rates (which, again, remain unchanged during the swap’s life). Bank X is worse off by the amount of extra interest it must pay to Party A, but not for one year—only for one week. To roughly estimate this, we can take the extra interest associated with the one-year swap from above ($1 million) and divide it by 52 (number of weeks in a year).The total would be about $19,000—significantly less than the first example, but with an identical contribution toward a notional value threshold. This is also a far, far cry from the massive-sounding $100 million notional value figure.
The disconnect between notional value and risk is even more obvious if we assume Bank X has both long and short IRS positions. For example, let’s say Bank X is long a $100 million IRS but also has an offsetting short $100 million IRS position. Bank X’s market risk is essentially zero, yet its gross notional calculation for purposes of de minimis under this scenario has just doubled to $200 million.
Paradoxically, the scenario in which Bank X has the smallest potential risk—the two offsetting $100 million swaps—results in the highest notional value level ($200 million), whereas the transaction with the highest potential risk—the one-year, $100 million IRS—results in the lowest notional value. Importantly, even in the scenario with the greatest risk, that risk is a fraction of the very large notional value figure.
From the examples above, it should be apparent that notional value, on its own, is a poor measure of activity and completely disconnected from risk. The current $8 billion notional threshold, and certainly a lower $3 billion threshold, could capture firms that pose no systemic risk and whose client-driven activity constitutes a very small percentage of the overall market.
If firms do not pose systemic risk, it is hard to argue that they should be captured as swap dealers and subjected to the same immense costs as the world’s largest, most interconnected financial institutions. Smaller entities may exceed a notional threshold but have a relatively minor footprint in the market place. Their registration would not meaningfully advance the goals of customer protection and market transparency, but it would still saddle them with significant costs. In the past, the Commission has agreed in principle with using a metric that would exclude activity “sufficiently modest in light of the total size, concentration and other attributes of the applicable markets” and that “the imposition of regulatory burdens on [such] entities…would not be expected to contribute significantly to advancing the customer protection, market efficiency and transparency objectives of dealer regulation.”18 Further, systemic risk posed by non-registered banks is limited: every bank’s derivative activity, whether it is registered as a swap dealer or not, is already heavily scrutinized during regular exams by federal bank regulators and publically disclosed on quarterly federal Call Reports.19 What’s more, there are currently 102 registered swap dealers, 51 of which are subject to prudential regulation.20 Where is the gaping hole in the systemic risk spectrum which an overly strict swap dealer registration threshold would plug or a more flexible standard would expose? I see none.
Establishing a Rational Metric for Registration
A better, more thoughtful registration metric is required which better aligns costs with risk and assesses that risk against the backdrop of the entire swaps regulatory framework.
The Commission is not establishing the de minimis exception in a vacuum. Subsequent to the adoption of the swap dealer definition, other regulatory requirements have gone into effect which also advance the goals of swap dealer registration, such as mandatory clearing, SEF trading, swap data and real-time reporting, and margin for uncleared swaps. For example, regardless of whether an entity is registered as a swap dealer, its swap activity is transparent to the Commission because of the swap data and real-time reporting requirements that apply to all market participants.
Also, any new metric must be better aligned with the costs of swap dealer registration. Those costs have two primary components: (i) risk-based costs (like capital and margin requirements), which target the reduction of systemic risk, and (ii) activity-based costs (like data reporting, trade confirmation, and reconciliation), which are intended to promote customer protection and market transparency. The largest costs of registration, namely capital and margin requirements,21 are designed to mitigate systemic risk, a threat that entities hovering around the current $8 billion de minimis threshold likely do not pose.22 In addition, it is unclear that an entity engaging in swap dealing activity around the $8 billion level has a sufficient number of transactions and counterparty relationships such that their registration would have an appreciable impact on market transparency and counterparty protections.23
For these reasons, I believe the de minimis threshold should be based on metrics more closely correlated with risk, as opposed to the current activity-based notional value metric. Examples of risk-based measures could include the amount of capital a firm holds against its swaps book, the economic exposure of a firm’s swaps book, or some measure of a firm’s uncleared swaps transactions. I look forward to exploring those potential metrics as we wait for new data on the swaps market.
However, if the Commission is intent on using only an activity-based metric to impose both risk-based and activity-based costs, it is my hope that we can use the next few months to develop a more robust measure of activity than notional value alone. One idea would be to combine notional value with the number of a firm’s swap dealing transactions and counterparties. While each of those metrics has its own limitations in measuring actual dealing activity,24 a de minimis threshold comprised of several metrics would mitigate false positives from any one of the others.25
Improvements to the De Minimis Exception Policy
The de minimis exception can be improved beyond just the notional value threshold. Specifically, the agency should explore three possible modifications: (i) broadening the Insured Depository Institution (IDI) exclusion to align it with current lending practices, (ii) excluding swaps that are cleared, and (iii) treating all hedging swaps consistently and excluding them from the threshold.
The IDI exclusion allows an insured depository institution to exclude swaps with customers connected to the origination of a loan from its de minimis count. Congress created that statutory exclusion to ensure that banks were not discouraged from assisting their borrowers in hedging risks associated with a loan, such as interest rate or currency risks.26 However, the Commission has implemented that statutory provision narrowly and thus, in my view, imposed unnecessary restrictions which thwart the Congressional intent behind the exclusion.
I believe the current exclusion for IDI lending activity should be broadened to reflect the dynamics of borrowers’ needs and demands. First, a loan is not the only form of credit extension that should qualify for the exclusion. Letters of credit, lines of credit, leases, revolving credit facilities, credit-enhanced bonds, and other forms of typical credit extension should also be eligible.27 Swaps tied to all of these different transactions should be considered within the exclusion for IDI lending activity.
In addition, the IDI exclusion’s restriction of swap hedging transactions only within a narrow window of time surrounding loan origination does not currently allow for hedging transactions around loan amendments or re-financings, nor through the life of a borrowing transaction.28 The artificial time constraints imposed on client-driven hedging activity, i.e., that it must occur between 90 days before a loan agreement up until 180 days after, do not reflect the economic realities of how banks or other firms currently serve their customers.29 The Office of the Comptroller of the Currency expressed similar views directly to the CFTC, stating that the IDI exclusion should be “tailored to allow for ongoing hedging that is connected to an extension of credit,” and that the statutory language of the IDI exclusion “does not limit the loan exclusion to swaps that are connected to the financial terms of a loan, nor does it require that the swap be entered into contemporaneously with loan origination.”30
Secondly, currently an entity must include cleared swaps in its de minimis calculation. However, central clearing addresses many of the systemic risk concerns that the Dodd-Frank Act and swap dealer registration were designed to mitigate. Once a swap is cleared, the relationship between the counterparties is extinguished and the clearinghouse manages the risk of the swap, including by requiring the posting of margin. Accordingly, I do not see the regulatory necessity of requiring such swaps to count toward an entity’s registration threshold.
Lastly, the Commission should provide market participants with regulatory certainty regarding the exclusion for hedging activity.31 For example, it should be clear that swaps entered into exclusively for purposes of hedging, which do not otherwise meet a prong of the swap dealer definition, should be excluded from an entity’s de minimis calculation. I believe “hedging” should be broadly interpreted to allow banks to continue managing their own risks as opposed to being dictated by Washington how to do so.
The de minimis exception is only one example of how the CFTC has adopted a one-size-fits-all approach that fails to identify, target, and regulate risk efficiently. You could host this conference every week for months with me keynoting and I wouldn’t run out of examples. We could talk about no-action letters: since 2013, the CFTC has issued over 400 staff letters providing no-action or exemptive relief to market participants from various regulatory requirements. This astonishingly high number is indicative of the number of regulations that have been issued with unrealistic compliance dates, that have adverse consequences, or that simply apply to entities or transactions that were not meant to be captured. But actually changing the rules wasn’t on the list of things to do.
Or, we could talk about all of the entity determinations which end-users must make to understand whether they are included in various CFTC rules or not.32 Why so many? Because defining a “commercial end-user,” which is perhaps the most important term in Title VII of Dodd-Frank given that the statute specifically excludes entities hedging commercial risk from many of its mandates, wasn’t on their list of things to do.
Or, we could discuss the nauseating web of our cross-border policies, which, when visually diagramed, would make Jackson Pollock blush.
We should be making it easier for market participants to understand their regulatory obligations, not harder. We should be incentivizing hedging, not penalizing it. We should be enhancing competition, not creating regulatory economies of scale. We should strive for simplicity and risk-targeting rules, not one-size-fits-all mandates.
The prior CFTC had their list, which was most notable for what wasn’t included.
This CFTC has our own list, which, I believe, is most notable for what is.
We are dedicating our thought and energy into accomplishing good, meaningful, and smart regulatory reform.
I look forward to contributing to those accomplishments and sharing them with you in the near future.
Thank you very much your patience and attention, and I am so grateful to Mercatus and IFM for putting on such a timely and important conference.
1 Arnold Lobel, Frog and Toad Together: A List (Harper & Row)(1972).
2 12 C.F.R § 1.3(ggg)(4).
3 See Ben Protess, Regulators to Ease a Rule on Derivatives Dealers, N.Y. Times, April 17, 2012, https://dealbook.nytimes.com/2012/04/17/regulators-to-ease-a-rule-on-derivatives-dealers/.
4 The NFA annual membership fee for a Tier 1 swap dealer is $1 million. See National Futures Association, https://www.nfa.futures.org/registration-membership/dues-revenue-structure.html.
5 See Capital Requirements of Swap Dealers and Major Swap Participants, 81 Fed. Reg. 91252, 91297-98 (Dec. 16, 2016); Proposed Regulation 23.101(a)(1). Depending on the activities of the firm, its actual capital requirements may be significantly higher than these minimum amounts.
6 See National Economic Research Associates, Cost-Benefit Analysis of the CFTC’s Proposed Swap Dealer Definition 1(Dec. 20, 2011) (“NERA Report”), http://www.nera.com/content/dam/nera/publications/archive2/PUB_SwapDealer_1211.pdf. It is difficult to estimate the initial and incremental, ongoing costs of swap dealer regulation. NERA’s report regarding the costs of registration for non-financial energy firms remains one of the only comprehensive analyses produced.
7 See letter from American Bankers Association dated January 13, 2016, regarding Swap Dealer De Minimis Exception Preliminary Report (“ABA Letter”).
8 See Swap Dealer De Minimis Exception Preliminary Report (“Preliminary Report”) at 35-36, http://www.cftc.gov/idc/groups/public/@swaps/documents/file/dfreport_sddeminis_1115.pdf.
9 Preliminary Report at 36-38.
10 Without the benefit of swap data reported to SDRs, the Commission established the $3 billion threshold partly by examining limited index credit default swap (CDS) data provided by the Securities and Exchange Commission (SEC), which represents a very narrow part of the overall swap market, and accepting the recommendation of several commenters to set the threshold at an amount equal to .001 percent of the overall domestic market for swaps. See Further Definition of “Swap Dealer,” “Security-Based Swap Dealer,” “Major Swap Participant,” “Major Security-Based Swap Participant,” and “Eligible Contract Participant,” 77 Fed. Reg. 30596, 30632-33 (May 23, 2012).
11 See Hearing to Review the 2016 Agenda of the Commodity Futures Trading Commission Before the H. Comm. on Agric., 114th Cong. 17 (2016) (response of Timothy Massad, former CFTC Chairman, to question posed by Congressman David Scott (D-GA)), https://agriculture.house.gov/uploadedfiles/114-40_-_98680.pdf.
12 See letter from Financial Services Roundtable dated January 19, 2016 (“We do not see a benefit to requiring an entity that enters into a small number of swaps with a large notional amount but little exposure to choose between exiting the market or registering as a swap dealer, nor should entities that are taking on very large exposures without crossing a notional threshold, or a trade or counterparty count metric, be unregulated because they have concentrated risk in a small number of trades.”).
13 An interest rate swap was chosen because IRS constitute approximately 76% of outstanding notional amounts in the global OTC swaps market. See Bank for International Settlements, Semiannual OTC Derivatives Statistics (September 17, 2017), at Table D5, https://www.bis.org/statistics/derstats.htm.
14 The fixed rate is determined such that the market value of the two legs is the same.
15 More precisely, this is the loss relative to interest rates staying the same over the life of the swap.
16 The timing of the interest rate increase is important. For example, if interest rates rose by 1% six months into the swap’s tenor, then Bank X’s loss would only be approximately $500,000, because it would only pay the higher rate for six months instead of a year.
17 The one-week tenor was chosen for ease of comparison. Although one-week interest rate swaps are not common, many interest rate products have short tenors (e.g., three month FRAs).
18 77 Fed. Reg. at 30629 (“We believe that factors that exclude entities whose dealing activity is sufficiently modest in light of the total size, concentration and other attributes of the applicable markets can be useful in avoiding the imposition of regulatory burdens on those entities for which dealer regulation would not be expected to contribute significantly to advancing the customer protection, market efficiency and transparency objectives of dealer regulation.”).
19 See ABA Letter.
20 For a list of provisionally registered swap dealers, see http://www.cftc.gov/LawRegulation/DoddFrankAct/registerswapdealer.
21 See NERA Report at 9.
22 See ABA Letter.
23 For example, in the Swap Dealer De Minimis Exception Final Staff Report (“Final Report”), staff estimated that if the de minimis threshold was increased to $15 billion, almost double its current amount, in IRS and CDS, less than 1% of notional activity, swap transactions and unique counterparties, respectively, would no longer be covered by swap dealer regulation as compared to the current $8 billion threshold. Final Report at 21, available at http://www.cftc.gov/idc/groups/public/@swaps/documents/file/dfreport_sddeminis081516.pdf.
24 See letter from Financial Services Roundtable dated January 19, 2016.
25 See letter from Institute of International Bankers dated January 19, 2016.
26 156 Cong. Rec. S5922 (daily ed. July 15, 2010)(statement of Sen. Lincoln)(“In addition, we made it clear that a bank that originates a loan with a customer and offers a swap in connection with that loan shouldn’t be viewed as a swap dealer.”).
27 See letter from M&T Bank dated January 13, 2016 (M&T Bank Letter).
29 Letter from ABA dated November 3, 2011, submitted in response to the Further Definition of “Swap Dealer,” “Security-Based Swap Dealer,” “Major Swap Participant,” “Major Security-Based Swap Participant” and “Eligible Contract Participant”, 75 Fed. Reg. 80174 (Dec. 21, 2010) (Proposed Rule). For example, with respect to loans to commodity swap customers, risk-mitigating commodity swaps are frequently not entered into at the inception of the loan, but are instead driven by the realities of the customer’s underlying physical business, for example, drilling or acquiring a new well. See ABA Letter, letter from Bank of Oklahoma, NA dated January 19, 2015.
30 Letter from Acting Comptroller of the Currency John Walsh to CFTC Chairman Gary Gensler (June 30, 2011) (regarding OCC Staff Comments on CFTC Dodd-Frank Act Proposed Rules).
31 See, e.g., 77 Fed. Reg. at 30611(“In general, entering into a swap for the purpose of hedging is inconsistent with swap dealing.”); 12 C.F.R. § 1.3(ggg)(6)(iii)(exclusion for swaps entered into to hedge physical positions).
32 See letter from National Rural Electric Cooperative Association and American Public Power Association dated September 28, 2017, regarding Project KISS “Miscellaneous”.
Last Updated: November 3, 2017