[Federal Register Volume 85, Number 249 (Tuesday, December 29, 2020)]
[Notices]
[Pages 85788-85802]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2020-28662]


-----------------------------------------------------------------------

SECURITIES AND EXCHANGE COMMISSION

[Release No. 34-90763; File No. SR-OCC-2020-016]


Self-Regulatory Organizations; The Options Clearing Corporation; 
Notice of Filing of Proposed Rule Change Concerning The Options 
Clearing Corporation's System for Theoretical Analysis and Numerical 
Simulation (``STANS'') Methodology Documentation

December 21, 2020.
    Pursuant to Section 19(b)(1) of the Securities Exchange Act of 1934 
(``Act'' or ``Exchange Act''),\1\ and Rule 19b-4 thereunder,\2\ notice 
is hereby given that on December 9, 2020, the Options Clearing 
Corporation (``OCC'') filed with the Securities and Exchange Commission 
(``Commission'') the proposed rule change as described in Items I, II, 
and III below, which Items have been prepared by OCC. The Commission is 
publishing this notice to solicit comments on the proposed rule change 
from interested persons.
---------------------------------------------------------------------------

    \1\ 15 U.S.C. 78s(b)(1).
    \2\ 17 CFR 240.19b-4.
---------------------------------------------------------------------------

I. Clearing Agency's Statement of the Terms of Substance of the 
Proposed Rule Change

    This proposed rule change by OCC would adopt a new document 
describing OCC's System for Theoretical Analysis and Numerical 
Simulation (``STANS''), which OCC uses to calculate daily and intra-day 
margin requirements for its Clearing Members (such document being the 
``STANS Methodology Description''). OCC also proposes to make 
conforming and other non-substantive changes to its Margin Policy.
    The proposed STANS Methodology Description is submitted without 
marking in confidential Exhibit 5A to SR-OCC-2020-016 because this 
document is being submitted in its entirety as new rule text. The 
proposed changes to OCC's current rule text related to the STANS 
methodology, known as the Margins Methodology, are contained in 
confidential Exhibit 5B to SR-OCC-2020-016. Material proposed to be 
added to the current rule text is

[[Page 85789]]

marked by underlining and material proposed to be deleted is marked by 
strikethrough text. The proposed changes to the Margin Policy are 
contained in confidential Exhibit 5C to SR-OCC-2020-016.\3\ Material 
proposed to be added to the Margin Policy is marked by underlining and 
material proposed to be deleted is marked by strikethrough text. The 
proposed rule change does not require any changes to the text of OCC's 
By-Laws or Rules. All terms with initial capitalization that are not 
otherwise defined herein have the same meaning as set forth in OCC's 
By-Laws and Rules.\4\
---------------------------------------------------------------------------

    \3\ OCC has filed a proposed rule change with the Commission to 
adopt a new Third-Party Risk Management Framework (``TPRMF''), which 
would replace the Counterparty Credit Risk Management Policy and 
provide an overview of OCC's approach to third-party risk 
management. That proposed rule change also includes conforming 
changes to OCC's Margin Policy. See Securities Exchange Act Release 
No. 90406 (November 12, 2020), 85 FR 73582 (November 18, 2020) (SR-
OCC-2020-014). The proposed changes to the Margin Policy currently 
pending Commission review in SR-OCC-2020-014 are marked in double 
underlining and double strikethrough text.
    \4\ OCC's By-Laws and Rules can be found on OCC's public 
website: https://www.theocc.com/Company-Information/Documents-and-Archives/By-Laws-and-Rules.
---------------------------------------------------------------------------

II. Clearing Agency's Statement of the Purpose of, and Statutory Basis 
for, the Proposed Rule Change

    In its filing with the Commission, OCC included statements 
concerning the purpose of and basis for the proposed rule change and 
discussed any comments it received on the proposed rule change. The 
text of these statements may be examined at the places specified in 
Item IV below. OCC has prepared summaries, set forth in sections (A), 
(B), and (C) below, of the most significant aspects of these 
statements.

(A) Clearing Agency's Statement of the Purpose of, and Statutory Basis 
for, the Proposed Rule Change

(1) Purpose
Background
    The STANS methodology is OCC's proprietary risk management system 
for calculating Clearing Member margin requirements.\5\ In general, 
STANS utilizes large-scale Monte Carlo simulations to forecast price 
and volatility movements in determining a Clearing Member's margin 
requirement.\6\ The STANS margin requirement is calculated at the 
portfolio level of Clearing Member accounts with positions in 
marginable securities. The STANS margin requirement consists of an 
estimate of a 99% expected shortfall (``ES'') \7\ over a two-day time 
horizon with additional charges for model risk, stress tests, 
liquidation costs, and various add-ons. The STANS methodology is used 
to measure the exposure of portfolios of options, futures, and cash 
instruments cleared by OCC.\8\
---------------------------------------------------------------------------

    \5\ See Securities Exchange Act Release No. 53322 (February 15, 
2006), 71 FR 9403 (February 23, 2006) (SR-OCC-2004-20).
    \6\ See OCC Rule 601.
    \7\ The ES component is established as the estimated average of 
potential losses higher than the value-at-risk (``VaR'') threshold. 
VaR refers to a statistical technique that is used in risk 
management to measure the potential risk of loss for a given set of 
assets over a particular time horizon.
    \8\ Pursuant to OCC Rule 601(e)(1), OCC also calculates initial 
margin requirements for segregated futures accounts on a gross basis 
using the Standard Portfolio Analysis of Risk Margin Calculation 
System (``SPAN''). SPAN is separate from STANS and is therefore not 
described in the STANS Methodology Description.
---------------------------------------------------------------------------

    OCC maintains technical documentation that describes in detail how 
the various quantitative components of STANS were developed and 
operate, including the various parameters and assumptions contained 
within those components \9\ and the mathematical theories underlying 
the selection of those quantitative methods (``Model Whitepapers''). 
The Model Whitepapers are currently synthesized in a single document, 
the Margins Methodology, describing how STANS operates from end to end. 
The Margins Methodology includes material aspects of OCC's risk-based 
margin system, which OCC must establish as a covered clearing agency 
under the Exchange Act and the rules promulgated thereunder, and which 
must be reasonably designed to, in part ``(i) [produce] margin levels 
commensurate with [the] risks and particular attributes of each 
relevant product, portfolio, and market; (ii) [mark] participant 
positions to market and [collect] margin, including variation margin or 
equivalent charges if relevant, at least daily . . . ; (iii) 
[calculate] margin sufficient to cover its potential future exposure to 
participants in the interval between the last margin collection and the 
close out of positions following a participant default; (iv) [use] 
reliable sources of timely price data and [use] procedures and sound 
valuation models for addressing circumstances in which pricing data are 
not readily available or reliable; [and] (v) [use] an appropriate 
method for measuring credit exposure that accounts for relevant product 
risk factors and portfolio effects across products . . .'' \10\ The 
Margins Methodology also includes information that would not be 
considered material aspects of OCC's methodology, such as internal 
procedural and administrative details, or details that could be 
reasonably and fairly implied by OCC's existing rules or other 
information contained in the document.
---------------------------------------------------------------------------

    \9\ See Securities Exchange Act Release No. 82473 (January 9, 
2018), 83 FR 2271 (January 16, 2018) (SR-OCC-2017-011), which 
describes how OCC periodically reviews the parameters and 
assumptions used by STANS pursuant to its Model Risk Management 
Policy and in accordance with 17 CFR 240.17Ad-22(e)(6).
    \10\ 17 CFR 240.17Ad-22(e)(6).
---------------------------------------------------------------------------

    Over time, OCC has filed sections of the Margins Methodology with 
the Commission as proposed rule changes under Section 19(b)(1) of the 
Exchange Act and Rule 19b-4 thereunder to effect changes to individual 
components of STANS.\11\ Thus, those chapters of the Margins 
Methodology have become codified as OCC rule text under Section 
19(b)(1) of the Exchange Act and Rule 19b-4. However, OCC now proposes 
to adopt a new STANS Methodology Description, which would replace the 
Margins Methodology document and codify the STANS methodology in its 
entirety under Section 19(b)(1) of the Exchange Act and Rule 19b-4. 
After adoption of the STANS Methodology Description, OCC would no 
longer maintain the Margins Methodology, neither as an OCC rule nor as 
an internal document.
---------------------------------------------------------------------------

    \11\ See Securities Exchange Act Release No. 74966 (May 14, 
2015), 80 FR 29784 (May 22, 2015) (SR- OCC-2015-010); Securities 
Exchange Act Release No. 76128 (December 28, 2015), 81 FR 135 
(January 4, 2016) (SR-OCC-2015-016); Securities Exchange Act Release 
No. 79818 (January 18, 2017), 82 FR 8455 (January 25, 2017) (SR-OCC-
2017-001); Securities Exchange Act Release No. 82161 (November 28, 
2017), 82 FR 57306 (December 4, 2017) (SR-OCC-2017-022); Securities 
Exchange Act Release No. 84524 (November 2, 2018), 83 FR 55918 
(November 8, 2018) (SR-OCC-2018-014); Securities Exchange Act 
Release No. 85440 (March 28, 2019), 84 FR 13082 (April 3, 2019) (SR-
OCC-2019-002); Securities Exchange Act Release No. 85755 (April 30, 
2019), 87 FR 19815 (May 6, 2019) (SR-OCC-2019-004); Securities 
Exchange Act Release No. 86296 (July 3, 2019), 84 FR 32816 (July 9, 
2019) (SR-OCC-2019-005); Securities Exchange Act Release No. 87387 
(October 23, 2019), 84 FR 57890 (October 29, 2019) (SR-OCC-2019-
010); Securities Exchange Act Release No. 89392 (July 24, 2020), 85 
FR 45938 (July 30,2020) (SR-OCC-2020-007); Securities Exchange Act 
Release No. 90139 (October 8, 2020), 85 FR 65886 (October 16, 2020) 
(SR-OCC-2020-012).
---------------------------------------------------------------------------

    In connection with this proposed rule change, OCC would also retire 
as rule text any chapters of the Margins Methodology previously filed 
with the Commission, as the proposed STANS Methodology Description is 
intended to cover the material aspects of the STANS methodology. Those 
chapters of the Margins Methodology that OCC has previously filed under 
Section 19(b)(1) of the Exchange Act and Rule 19b-4 \12\ would be 
superseded in their entireties by new corresponding sections in the

[[Page 85790]]

STANS Methodology Description, as described in further detail herein.
---------------------------------------------------------------------------

    \12\ Id.
---------------------------------------------------------------------------

    The current text of the Margins Methodology includes various 
details that would no longer be OCC rule text following the adoption of 
the proposed STANS Methodology Description. While the details that OCC 
proposes to remove are described in further detail herein, 
thematically, they consist of the following:
     Details on OCC's historical modeling practices and 
potential future enhancements, which do not describe how a model 
currently functions;
     Details on the exact set of current products applied to 
each STANS component, which will change from time to time as OCC-
cleared products are listed and delisted;
     Various configuration choices made by OCC, such as data 
sources, model parameters, and model performance monitoring, that are 
not inherent to model selection or design and that do not materially 
impact a model's results, which OCC may from time to time determine it 
should modify based on current market conditions and business 
practices;
     Extensive, detailed testing results and explanatory 
rationale supporting a model;
     Recitations of standard mathematical and economic 
theories/techniques that are well-known in quantitative finance, 
readily found in public sources, and do not include OCC-specific 
modifications or applications;
     Redundant descriptions of a model component appearing in 
multiple chapters;
     Details on OCC's implementation of a model in its internal 
technology systems and application of model results in operational 
procedures that are not inherent to a model and that OCC could change 
from time to time without affecting a model's results; and
     Manual margin adjustments and add-ons OCC employs pursuant 
to OCC rules, policies, and/or procedures outside the STANS 
methodology.
    The proposed STANS Methodology Description is intended to be a 
comprehensive description of STANS that is made available to Clearing 
Members and enable an informed reader to understand OCC's modeling 
choices and the interconnectedness of STANS model components in 
producing OCC margin requirements. Therefore, OCC believes the details 
summarized above and described herein are extraneous to this purpose. 
Rather, OCC believes these types of details are more appropriately 
covered--to the extent these details are specific to an OCC model--in 
other OCC rules and policies, Model Whitepapers, or other internal OCC 
documentation.
    OCC also believes, as described in Item II.A.2 below, these details 
do not need to maintained as OCC ``rules'' as defined by Section 
19(b)(1) of the Exchange Act and Rule 19b-4.\13\ These internal 
procedural and administrative details used by OCC's model developers 
and model validators would: (1) Be reasonably and fairly implied by the 
proposed STANS Methodology Description, OCC's Margin Policy,\14\ OCC's 
Model Risk Management Policy,\15\ and other OCC rules; and/or (2) 
otherwise not be deemed to be material aspects of OCC's margin setting-
related operations. While OCC would not maintain these details in the 
STANS Methodology Description, OCC would continue to maintain and 
update these details when necessary in the Model Whitepapers and other 
internal OCC documentation, where these details are also currently 
found.\16\
---------------------------------------------------------------------------

    \13\ Section 19(b)(1) of the Exchange Act requires a self-
regulatory organization (``SRO'') such as OCC to file with the 
Commission any proposed rule or any proposed change in, addition to, 
or deletion from the rules of such SRO. See 15 U.S.C. 78s(b)(1). 
Section 3(a)(27) of the Exchange Act defines ``rules of a clearing 
agency'' to mean its (1) constitution, (2) articles of 
incorporation, (3) bylaws, (4) rules, (5) instruments corresponding 
to the foregoing and (6) such ``stated policies, practices and 
interpretations'' (``SPPI'') as the Commission may determine by 
rule. See 15 U.S.C. 78c(a)(27). Exchange Act Rule 19b-4(a)(6) 
defines the term ``SPPI'' to include (i) any material aspect of the 
operation of the facilities of an SRO and (ii) statements made 
generally available to membership of, to all participants in, or to 
persons having or seeking access to facilities of an SRO that 
establishes or changes certain standards, limits, or guidelines. See 
17 CFR 240.19b-4(a)(6). Rule 19b-4(c) provides, however, that an 
SPPI may not be deemed to be a proposed rule change if it is: (i) 
Reasonably and fairly implied by an existing rule of the SRO or (ii) 
concerned solely with the administration of the SRO and is not an 
SPPI with respect to the meaning, administration, or enforcement of 
an existing rule the SRO. See 17 CFR 240.19b-4(c).
    \14\ See Securities Exchange Act Release No. 82355 (December 19, 
2017), 82 FR 61058 (December 26, 2017) (SR-OCC-2017-007).
    \15\ See Securities Exchange Act Release No. 82473 (January 9, 
2018), 83 FR 2271 (January 16, 2018) (SR-OCC-2017-011).
    \16\ OCC's Model Risk Management Policy establishes detailed 
standards for Model Whitepapers and governance to adopt or make 
changes to a Model Whitepaper. See id.
---------------------------------------------------------------------------

STANS Methodology Description
    The proposed STANS Methodology Description would describe the 
material aspects of OCC's margin methodology. Specifically, the STANS 
Methodology Description would include (i) an executive summary; (ii) 
descriptions of the quantitative model components of STANS; and (iii) 
``model utilities'' intended to enhance the quality of model results. 
Each of these sections is described in further detail below.\17\
---------------------------------------------------------------------------

    \17\ The proposed STANS Methodology Description would also 
include the following non-substantive sections: (i) A table of 
contents; (ii) a list of references to academic and technical 
documents, both public and internal to OCC; and (iii) a table of 
defined terms used in the STANS Methodology Description.
---------------------------------------------------------------------------

Executive Summary
    The STANS Methodology Description would provide an executive 
summary of STANS. This executive summary would describe how the purpose 
of STANS is to determine margin requirements for OCC's Clearing Members 
(as described below), and in doing so meet various risk management 
goals and regulatory requirements for OCC. The executive summary would 
then describe the types of positions and collateral modeled through 
STANS, which include (i) valued securities and stock loans; (ii) 
equity, index, and futures options; (iii) Flexible Exchange (``FLEX'') 
options; (iv) equity and index futures; (v) volatility futures; and 
(vi) commodity futures. The executive summary would then provide an 
overview of the STANS methodology, which includes (i) econometric 
calibration; (ii) copula estimation and Monte Carlo simulation; (iii) 
volatility forecasting; (iv) theoretical underlying price generation; 
(v) theoretical derivatives price generation; and (vi) aggregation and 
margin calculation. These components are described in further detail 
below. The executive summary would then describe OCC's model monitoring 
activities, which include (i) daily backtesting and (ii) ongoing 
parameter monitoring pursuant to monitoring plans established by OCC's 
Model Risk Working Group (``MRWG'').\18\ The executive summary would 
then describe that STANS relies on price feeds of real-time market data 
to generate theoretical values in calculating margin requirements, and 
how OCC staff may use price editing techniques to improve the quality 
of pricing data for input into STANS.\19\ Lastly, the executive summary 
would briefly outline the organization of the sections of the STANS 
Methodology Description that substantively describe the core components 
of the STANS methodology

[[Page 85791]]

and the related data processing utilities used by STANS.
---------------------------------------------------------------------------

    \18\ OCC's Margin Policy and Model Risk Management Policy 
provide more detail on OCC's model monitoring activities. See supra 
notes 14 and 15.
    \19\ OCC's Collateral Risk Management Policy and Margin Policy 
provide more detail on the function of OCC's Pricing & Margins 
department. See Securities Exchange Act Release No. 82009 (November 
3, 2017), 82 FR 52079 (November 9, 2017) (SR-OCC-2017-008) and supra 
note 14.
---------------------------------------------------------------------------

    The proposed text of this executive summary would replace current 
OCC rule text from the Margins Methodology's introductory section. The 
current text, in addition to summarizing the STANS methodology as would 
the proposed text described above, includes descriptions of the 
following:
     OCC's historical modeling practices: OCC does not believe 
this historical information is needed to understand how the model 
functions.
     Redundant details of the STANS methodology also found in 
the main body of both the Margins Methodology and the proposed STANS 
Methodology Description: This information, would already be detailed in 
the main body of the STANS Methodology Description, and OCC does not 
believe repeating it here is needed to understand how STANS functions.
     A ``documentation guide'' describing what information can 
be found within various sections of the Margins Methodology: OCC does 
not believe this documentation guide is needed to understand how STANS 
functions, or to understand the organization of the proposed STANS 
Methodology Description.
    For the reasons stated above, OCC proposes to delete this rule text 
in its entirety without adding new, corresponding rule text in the 
STANS Methodology Description.
STANS Methodology Components
    The STANS Methodology Description would next describe the 
components of OCC's risk-based margin methodology, which OCC uses to 
cover the credit exposures presented by Clearing Members in accordance 
with Rule 17Ad-22(e)(6). In particular, the STANS Methodology 
Description would describe the (i) calibration of various parameters 
and price data inputs used by OCC's econometric and pricing models to 
create risk factors; (ii) construction of a copula from the risk 
factors that identifies correlations among simulated changes in the 
various risk factors; (iii) application of the simulated risk factor 
changes and correlations to actual data through Monte Carlo simulations 
that estimate 10,000 possible scenarios for each risk factor, then 
estimation of theoretical prices for securities, derivatives, and 
futures using these theoretical scenarios; and (iv) application of the 
theoretical prices to actual Clearing Member positions to calculate 
margin requirements.
i. Model and Econometric Calibration
    The STANS Methodology Description would describe how the 
quantitative models used by STANS incorporate various historical price 
data and econometric parameter inputs, which are used to estimate and 
simulate the risk for an associated product. These inputs consist of 
(i) returns on equity securities; (ii) implied volatilities; (iii) 
energy and commodity futures; (iv) treasury securities; (v) variance 
futures; and (vi) volatility futures. In total, there are currently 
approximately 40,000 of these inputs. The exact number of inputs is 
subject to change based on the types of products that OCC clears and 
OCC's research on what risk factors correlate with prices changes in 
these products. Historical price data comes from OCC's Pricing & 
Margins department, which obtains the data from external vendors and 
then validates it for use within STANS.\20\ STANS uses several models, 
described below, to calibrate this historical data and then transform 
the historical data into theoretical values that, along with 
specialized volatility forecast and marginal distribution parameters 
constructed by other OCC models described below, are used to construct 
a copula, described in the next step.
---------------------------------------------------------------------------

    \20\ See supra note 14.
---------------------------------------------------------------------------

Equity Returns
    STANS uses returns on equity securities that are based on current 
market prices. STANS first calibrates this data by simply creating a 
time series of logarithmic returns based on the closing, and in some 
cases opening, prices. This transformation does not require a separate 
model. The data is used to create econometric parameters and for 
pricing as described further below.
Implied Volatility
    STANS uses implied volatility risk factors to measure the expected 
future volatility of an option's underlying security at expiration, 
which is reflected in the current option premium in the market. To 
address variations in implied volatility, OCC models a volatility 
surface for options by incorporating into the econometric models 
underlying STANS certain risk factors called ``pivot points.'' These 
pivot points are chosen such that their combination allows STANS to 
identify changes in the level, skew, convexity, and term structure of 
the implied volatility surface. STANS generates a value for each of the 
nine pivot points for each eligible underlying asset and for each 
business day in the historical data period. To calibrate this data, for 
each of the nine pivot points STANS performs a kernel smoothing 
technique \21\ on the historical data. Application of these pivot 
points enables STANS to simulate implied volatility scenarios, which 
are then used to create the specialized volatility forecast and 
marginal distribution parameters described below, and in the pricing of 
options through OCC's option pricing models described further 
below.\22\
---------------------------------------------------------------------------

    \21\ ``Kernel smoothing'' is a statistical process by which data 
points are better fitted to an expected function using weighted 
averages and a ``smoothing parameter.''
    \22\ See Securities Exchange Act Release No. 76128 and 
Securities Exchange Act Release No. 84524 for more information on 
the function and application of the implied volatility model.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on implied volatility. The current rule 
text also includes other information related to the implied volatility 
model. Specifically, the current rule text includes descriptions of the 
following:
     Products eligible for implied volatility scenarios 
modeling in STANS: OCC does not believe the exact list of products to 
which this model is applied is needed to understand how the model 
functions, and this list may change from time to time as OCC-cleared 
products are listed and delisted.
     Data sources used by STANS to perform the kernel smoothing 
technique: These data sources are configuration choices made by OCC 
that are not inherent to the model's selection or design and that OCC 
could change from time to time without affecting the model's results.
     Rationale for the assumptions underlying implied 
volatility modeling of longer-tenor options: OCC does not believe that 
the justification for these model assumptions is needed to understand 
how the model currently functions.
     Historical background on OCC's decision to incorporate 
implied volatility modeling into STANS: OCC does not believe that this 
historical information is needed to understand how the model currently 
functions.
     Model testing and validation results for the implied 
volatility model: OCC does not believe that the internal testing and 
validation performed to verify the model is fit for use is needed to 
understand how the model currently functions.
    OCC believes that this information is more appropriately covered in 
the Implied Volatility Scenarios Model Whitepaper and other internal 
OCC documentation rather than in OCC's

[[Page 85792]]

rules for the reasons listed above. Therefore, OCC proposes to delete 
this rule text in its entirety without adding new, corresponding rule 
text in the STANS Methodology Description.
Treasury Securities
    STANS prices treasury securities \23\ using a Nelson-Siegel 
framework,\24\ a commonly used polynomial model for constructing the 
term structure of an interest rate and modeling changes in a yield 
curve.\25\ STANS constructs a theoretical yield curve using current and 
historical settlement prices for treasury securities.
---------------------------------------------------------------------------

    \23\ While OCC does not clear treasury securities or derivatives 
on such products, OCC permits Clearing Members to deposit treasury 
securities as margin collateral.
    \24\ See Nelson, C.R. and Siegel, A.F., ``Parsimonious Modeling 
of Yield Curves,'' 60 The J. of Bus. 4, 473-489 (1987) (describing 
the Nelson-Siegel model).
    \25\ In addition to pricing treasury securities, STANS uses a 
Nelson-Siegel framework to simulate potential changes in interest 
rates. Refer to the below description of the interest rate curve 
model utility.
---------------------------------------------------------------------------

    STANS calibrates this data by transforming the market prices into a 
time series of unobservable factors that represents the yield curve. 
STANS fits these Nelson-Siegel parameters using observed bond prices. 
In simulation, STANS creates ``shocks'' on theoretical Nelson-Siegel 
parameters \26\ to create theoretical interest rate curves, which are 
in turn used to price treasury securities.
---------------------------------------------------------------------------

    \26\ STANS also introduces extra ``noise'' into the bond prices 
to account for the bonds' idiosyncratic behaviors and prevent the 
resulting treasury securities price movements from being perfectly 
correlated.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on U.S. Treasury bills and fixed rate 
notes, bonds, and strips. The current rule text also includes other 
information related to the treasury securities and interest rate model. 
Specifically, the current rule text includes the following:
     Summary and introduction sections that describe OCC's need 
to model treasury securities and interest rates and provide an overview 
of the U.S. Treasury securities market: OCC does not believe these 
background descriptions of the macroeconomic environment, found in 
public sources, are needed to understand how the model currently 
functions.
     Restatements of mathematical definitions and equations 
describing the relationship between the forward and yield curves, and 
the payoff function for bonds used to describe all interest rate 
curves: This information, while relevant to understanding how the model 
functions, is foundational information commonly understood in 
quantitative finance and readily found in public academic sources. To 
the extent the text does not describe OCC's application of these 
theories, OCC does not believe this information needs to be maintained 
in OCC's rules.
     Details on how OCC implemented the model in its technology 
systems: These implementation details relate to OCC's internal 
administration of its technology systems and are not needed to 
understand how the model currently functions. Because these details are 
not inherent to the model's selection or design, OCC could also change 
them from time to time without affecting the model's results.
     Redundant description of the copula constructed by STANS: 
This information, described further below, would already be detailed in 
the STANS Methodology Description section related to the construction 
of a copula, and OCC does not believe repeating it here is needed to 
understand how the model currently functions.
    OCC believes that this information is more appropriately covered in 
the Nominal Treasury Securities Model Whitepaper and other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
Generic Futures
    Relying on current futures prices and time series of spot prices as 
inputs, STANS uses a generic futures model to price linear derivatives 
with limited term structures. Using basic economic assumptions that the 
relationship of spot prices to forward prices does not allow for 
arbitrage and that futures prices equal forward prices, or that any 
deviations from this are adequately addressed through costs implicit in 
carrying such positions,\27\ the model estimates and applies 
theoretical discount factors to the futures prices. These discount 
factors are based on a ratio of estimated spot prices on the underlying 
securities to the futures prices.
---------------------------------------------------------------------------

    \27\ As described previously, pursuant to OCC's Model Risk 
Management Policy OCC periodically reviews all parameters and 
assumptions used in STANS and they are subject to change.
---------------------------------------------------------------------------

Variance Futures
    STANS uses a specialized factor model to price variance futures, 
which uses historical data for both variance futures products and the 
Standard and Poor's 500 Index (``SPX''). This model relies on basic 
assumptions that the short-term volatility of variance futures prices 
tends to revert towards a mean (i.e., volatility remains relatively 
close to an average value), but the long-term volatility is itself 
stochastic. Using these assumptions, STANS fits current values of the 
volatility and volatility mean reversion level, in addition to 
parameters describing the dynamics, to the current term structure of 
variance futures prices. Modeling variance futures prices based on 
these assumptions allows the theoretical prices to move in a realistic 
fashion.
    The model is first calibrated with historical data on variance 
futures prices and their recent dynamics. It then simulates prices for 
variance futures using two sets of random variables: (i) SPX returns; 
and (ii) changes in the long-term volatility level, represented by 
normal random numbers that STANS generates daily for use only with 
variance futures and that have no correlation with other theoretical 
numbers generated by STANS. Both random variables are used to simulate 
scenarios for prices of the variance futures tenors.
Synthetic Futures
    Using logarithmic daily returns of active futures and various other 
securities, STANS uses a ``synthetic futures'' model to estimate prices 
of certain products such as volatility index-based futures (e.g., VIX 
futures). In general, the synthetic futures model creates an artificial 
(or ``synthetic'') time series of price innovations for actual futures 
contracts with approximately the same tenor as the actively-traded 
futures.\28\ This synthetic time series then serves as a uniform 
substitute for a time series of daily settlement prices for the actual 
futures contracts, which persists over many expiration cycles and thus 
can be used as a basis for econometric analysis. STANS performs this 
analysis by fitting the synthetic time series with associated 
volatility forecast and marginal distribution parameters, which are 
described below.
---------------------------------------------------------------------------

    \28\ See Securities Exchange Act Release No. 85440 for further 
information on OCC's synthetic futures model as applied to 
volatility index-based products. OCC notes that the synthetic 
futures model can also be used for other futures products, such as 
interest rate futures. See e.g., Securities Exchange Act Release No. 
89392 and Securities Exchange Act Release No. 90139.
---------------------------------------------------------------------------

    The traded futures contracts are then mapped to the simulated 
return scenarios of the corresponding synthetic contracts to produce 
theoretical prices. The first synthetic contract in the series contains 
returns of the front contract on any given day. STANS switches the

[[Page 85793]]

front contract of the series to the next one out on the day following 
the expiration date of the front contract. While the synthetic time 
series contain returns from different contracts, a return on any given 
date is constructed from prices of the same contract. Using a synthetic 
time series allows STANS to better approximate correlations between 
futures contracts of different tenors by creating more price data 
points and their margin offsets. These synthetic time series are mapped 
to the underlying futures product they are intended to represent.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on synthetic futures. The current rule 
text also includes other information related to the synthetic futures 
model. Specifically, the current rule text includes descriptions of the 
following:
     Rationale for making changes to the model in 2019 \29\ and 
other historical information: OCC does not believe that this rationale 
and historical information is needed to understand how the model 
currently functions.
---------------------------------------------------------------------------

    \29\ See id.
---------------------------------------------------------------------------

     Equations for standard GARCH provided for introductory 
purposes: A description of OCC's GARCH model, described further below, 
would already be detailed in the STANS Methodology Description section 
related to GARCH parameters, and OCC does not believe repeating it here 
is needed to understand how the model functions. Furthermore, the GARCH 
equations as implemented in STANS are modified from the standard GARCH 
equations provided here, and OCC believes this text could create 
confusion around the exact GARCH equations used in STANS.\30\
---------------------------------------------------------------------------

    \30\ See infra note 36.
---------------------------------------------------------------------------

    OCC believes that this information is more appropriately covered in 
the Synthetic Futures Model Whitepaper and other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
GARCH and NRIG Parameters
    STANS utilizes econometric parameters related to volatility 
forecasts and marginal distributions, and calibrates these parameters 
using ten-year histories of the data inputs described above. For both 
volatility forecasts and marginal distributions, STANS utilizes a 
generalized autoregressive conditional heteroskedasticity (``GARCH'') 
model. GARCH is a common statistical model for, in a time series of 
data, comparing the variance of one point in the time series to the 
previous point in the series rather than an arithmetic average of all 
the points in the series. This is particularly useful when the value of 
volatility at one point in a time series is known to be correlated with 
the volatility at previous points in the series. STANS estimates these 
GARCH parameters through a maximum likelihood estimation method. By 
fitting these GARCH parameters to a time series of risk factor 
innovations, STANS is able to remove the effects of volatility from--or 
``devolatilize''--the risk factor time series so that the copula 
described below can estimate the correlations among the risk factors 
irrespective of their individual volatilities.
    To model volatility forecast parameters, STANS fits the time series 
of implied volatility pivot points (described above) into a Student's 
t-distribution, a continuous probability distribution that is commonly 
used to estimate the mean of a population with a relatively small 
sample size and unknown standard deviation. To determine the 
appropriate degrees of freedom for a particular distribution, STANS 
applies an Anderson-Darling test.
    To model marginal distribution parameters, STANS utilizes a normal 
reciprocal inverse Gaussian (``NRIG'') distribution, a special case of 
the generalized hyperbolic distribution.\31\ The returns \32\ of each 
risk factor used in STANS are assumed to exhibit returns in the shape 
of a symmetric NRIG distribution.\33\ Consequently, STANS calibrates 
NRIG parameters that are designed to describe the shape of every risk 
factor individually.
---------------------------------------------------------------------------

    \31\ The generalized hyperbolic distribution is a special type 
of continuous probability distribution. See Barndorff-Nielsen, O., 
``Exponentially decreasing distributions for the logarithm of 
particle size,'' 353 Proc. of the Royal Soc'y of London. Series A, 
Mathematical and Physical Sci. 1674, 401-419 (1977) (explaining the 
generalized hyperbolic distribution).
    \32\ ``Return'' refers generally to changes in a risk factor's 
value over a time interval. Returns could take the form of simple 
differences, log returns, or other forms.
    \33\ Except for (i) Chicago Volatility Index (``VIX'') futures, 
which are assumed to follow an asymmetric NRIG distribution, and 
(ii) implied volatility, which is assumed to follow a Student's t-
distribution.
---------------------------------------------------------------------------

    As described previously, STANS constructs these GARCH and NRIG 
parameters from the historical price data and econometric parameter 
inputs that are first calibrated by the models described above. These 
historical price data and econometric parameters, and the resulting 
GARCH and NRIG parameters, are the foundational data elements used by 
the copula and pricing models described in the proceeding steps.
    The STANS Methodology Description would also describe the controls 
that may be placed on the GJR-GARCH parameters after their initial 
calibration. GARCH volatility forecasting models can be very reactive 
in certain market environments. As a result, OCC may implement 
parameter controls for risk factors and classes of risk factors, which 
are subject to periodic review and approval by the MRWG.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on GARCH forecasts. OCC notes that the 
current rule text describes the standard NRIG cumulative distribution 
function that is widely available in public academic texts. The 
proposed rule text would describe the same function in a re-
parameterized form that is proprietary to OCC. While the two forms are 
mathematically equivalent, the re-parameterized form is used in the 
Econometric Model for Risk Factors in STANS Model Whitepaper and the 
proposed text would therefore be made consistent with the Model 
Whitepaper. The proposed rule text would also include a citation to an 
academic paper describing the rationale for the re-parameterization.
    The current rule text also includes other information related to 
OCC's GARCH model. Specifically, the current rule text includes 
descriptions of the following:
     Introductory language describing the standard Glosten-
Jagannathan-Runkle GARCH model and the use of a Student's t-
distribution: This information, while relevant to understanding how the 
model functions, is foundational information commonly understood in 
quantitative finance and readily found in public academic sources. To 
the extent this text does not describe OCC's application of GARCH and 
the Student's t-distribution, OCC does not believe this information 
needs to be maintained in OCC's rules.
     Details on variance forecasting (i.e., considering how 
securities volatility tends to clusters during certain periods) as 
rationale for model selection: OCC believes this information is 
extraneous to understanding how the GARCH model currently functions in 
STANS.
     Variance forecasting as applied to the One-Day and Two-Day 
Scenarios model utility: This information, described further below, 
would already be detailed in the STANS Methodology Description section 
related to the One-Day and Two-Day Scenarios model

[[Page 85794]]

utility, and OCC does not believe repeating it here is needed to 
understand how the model utility currently functions.
     Mathematical rationale for the cumulative distribution 
function,\34\ inverse cumulative distribution function, and degrees of 
freedom for the Student's t-distribution used by the GARCH model for 
implied volatility risk factors: OCC believes this information is 
extraneous to understanding how the GARCH model currently functions in 
STANS. This information is also foundational information commonly 
understood in quantitative finance and readily found in public academic 
literature. To the extent this text does not describe OCC's application 
of these functions and the Student's t-distribution, OCC does not 
believe this information needs to be maintained in OCC's rules.
---------------------------------------------------------------------------

    \34\ In probability theory, the cumulative distribution function 
of a random variable is the probability that the variable will be 
less than or equal to a set value.
---------------------------------------------------------------------------

     Explanatory mathematical formulas for variance forecasting 
of implied volatility risk factors and a likelihood function \35\ and 
equations related to the Anderson-Darling test,\36\ including the 
Student's t cumulative distribution function for integer values of n: 
These details relate to implementation of the GARCH model in OCC's 
internal technology systems, are not inherent to the model's selection 
or design, and are not needed to understand how the model currently 
functions.
---------------------------------------------------------------------------

    \35\ A likelihood function is a tool used to measure the 
goodness of fit of a statistical model to sample data.
    \36\ The Anderson-Darling test is a statistical test of whether 
a given sample of data is drawn from a population of data with a 
specific probability distribution.
---------------------------------------------------------------------------

     Expressions for the Gamma and Beta functions: \37\ This 
information, while relevant to understanding how the model functions, 
is foundational information commonly understood in quantitative finance 
and readily found in public academic literature. To the extent the text 
does not describe OCC's application of Gamma and Beta functions in the 
model, OCC does not believe this information needs to be maintained in 
OCC's rules.
---------------------------------------------------------------------------

    \37\ Gamma and Beta functions, respectively, are related one and 
two-variable functions that serve as foundations for various 
mathematical applications.
---------------------------------------------------------------------------

    OCC believes that this information is more appropriately covered in 
the underlying GARCH Model Whitepaper and other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
ii. Copula Construction
    The STANS Methodology Description would describe how a copula is 
used to quantify the joint behavior and dependence structure of the 
risk factors used by STANS. A copula is a mathematical construct used 
in probability theory to calculate the cumulative distribution of a set 
of random variables. The fitted copula can then be used by STANS to 
perform Monte Carlo simulations of theoretical prices for underlying 
securities and associated derivatives, which will be used in the 
aggregation step during which margin requirements are calculated.
    To estimate the copula, STANS first transforms two years of 
historical data for the risk factors produced by the models described 
above into a data set described by the Student's t-distribution with 
four degrees for freedom.\38\ STANS then performs a singular value 
decomposition of this data set to obtain the eigenvector decomposition 
\39\ of the correlation matrix. This means the resulting fitted copula 
is a Student's t copula with four degrees of freedom.
---------------------------------------------------------------------------

    \38\ Based on OCC's research, four degrees of freedom is in the 
conservative end of a range of degrees of freedom that are generally 
suitable fits for univariate distributions and is therefore 
appropriate for use in constructing the copula.
    \39\ In the context of linear transformations, an Eigenvector is 
a value that does not change direction when the transformation is 
applied to it, but rather changes in scale based on the application 
of a scalar factor, called an Eigenvalue. Eigenvectors and 
Eigenvalues are used to analyze the characteristics of linear 
transformations, including correlation/covariance matrices, and 
generate random variables with the equivalent correlation.
---------------------------------------------------------------------------

    Before the copula is estimated, STANS performs an ``alignment'' 
step on the time series to identify and separately process risk factors 
with incomplete data sets that lack sufficient data to estimate the 
copula. Specifically, for pricing data/models for underlyings OCC 
extracts data on the previous two years (i.e., 500 business days) and 
ensures (i) the data has no more than 100 missing returns as compared 
to baseline dates and (ii) the five latest returns are not missing as 
compared to baseline dates. If a risk factor's data set does not meet 
each of these three criteria, it is subject to a conditional or default 
simulation, described below.
    To simulate price movements, STANS draws random samples from the 
multivariate Student's t-distribution described by the copula. These 
random draws are abstract values that correspond to correlated, 
uniform, normalized shocks in the risk factors. STANS then 
reincorporates the individual volatility and marginal distribution of 
the risk factors to create appropriate return scenarios. STANS next 
applies these theoretical returns to current market prices to generate 
potential price scenarios for underlying securities. STANS essentially 
performs the reverse of the function that was performed to fit the 
econometrics of the risk factors.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on the Student-t Copula model. The 
current rule text also includes other information related to the 
construction and simulation of a copula in STANS. Specifically, the 
current rule text includes a mathematical justification for using a 
copula generally, and introductory text describing the general 
properties of a Student's t copula. OCC believes this information is 
extraneous to understanding how the Student-t Copula model currently 
functions in STANS. This information is also foundational information 
commonly understood in quantitative finance and readily found in public 
academic literature. To the extent this text does not describe OCC's 
application of a mathematical copula, OCC does not believe this 
information needs to be maintained in OCC's rules. Instead, OCC 
believes that this information is more appropriately covered in the 
underlying Student-t Copula Model Whitepaper and other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
Conditional and Default Simulations
    For risk factors with data sets that have only recently become 
available, or that have experienced drastic changes in their return 
characteristics, and do not meet one or more of the criteria in the 
alignment step, there may be too small of a sample size to reliably 
estimate correlations among the data. In such cases, these risk factors 
are excluded from the copula simulation in STANS and OCC applies 
conditional or default simulation.
    OCC applies a conditional simulation when it believes that a risk 
factor that has been identified during the alignment step does not meet 
the data quality criteria but has an appreciable correlation with 
another risk factor that has a more robust dataset. OCC uses that more 
robust risk factor's data as a proxy for the identified risk factor. 
The identified risk factor is assumed to

[[Page 85795]]

exhibit simulated results that follow an NRIG distribution of specified 
mean, variance, and shape parameters, and any variation from the proxy 
data is assumed to be purely idiosyncratic. Pursuant to OCC's Margin 
Policy, OCC periodically reviews whether applying a conditional 
simulation to a particular risk factor continues to be appropriate.
    OCC applies a default simulation when it does not believe an 
identified risk factor has any obvious proxy and has no view on its 
prospective volatility, or when a risk factor is identified by STANS 
during nightly margin processing and OCC has not already selected it to 
undergo a conditional simulation. In a default simulation, movements in 
the risk factor are assumed to be entirely idiosyncratic and have a 
volatility that is typical of highly volatile stocks.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on default, derived, and conditional 
factors. The current rule text also includes other information related 
to conditional and default simulations. Specifically, the current rule 
text includes the following:
     Introductory text restating the use of time series in 
STANS: This information would already be described elsewhere in the 
STANS Methodology Description where applicable, and OCC does not 
believe repeating it here is needed to understand how the model 
functions.
     A description of ``derived scenarios,'' a special case of 
conditional simulations related to exchange rate risk factors: This 
special case is applied pursuant to internal OCC procedures, and occurs 
outside of the STANS methodology. Therefore, OCC does not believe this 
information is needed to understand how the model currently functions.
     A description of the how OCC operationally applies 
conditional simulations: These operational details relate to OCC's 
application of the model's results in operational procedures and are 
not inherent to the model's selection or design, and that OCC could 
change from time to time without affecting the model's results.
     Details on how OCC implemented default scenarios in its 
internal technology systems: These implementation details relate to 
OCC's internal administration of its technology systems and are not 
inherent to the model's selection or design, and that OCC could change 
from time-to-time without affecting the model's results.
    OCC believes that this information is more appropriately covered in 
the Student-t Copula Model Whitepaper or other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
iii. Implied Volatility Smoothing and Options Pricing
    The STANS Methodology Description would next describe how STANS 
utilizes the inputs and outputs described above to (i) perform implied 
volatility smoothing, (ii) price European and American options, (iii) 
price Asian FLEX options, and (iv) price Cliquet options.
Implied Volatility Smoothing
    STANS employs an Implied Volatility Smoothing model to estimate 
fair prices of listed option contracts based on their bid and ask price 
quotes. This model supports pricing of the following types of options: 
(i) European and American options on equity products with a dividend 
yield or with discrete cash dividends; (ii) European and American 
options on futures on equity indices, currencies, and commodities; 
(iii) options on volatility indices for which volatility futures trade 
(e.g., VIX options \40\); (iv) forward start options; and (v) Asian 
FLEX options.
---------------------------------------------------------------------------

    \40\ VIX options are treated as options on VIX futures and thus 
represent a type of option on futures that is also supported by the 
implied volatility smoothing.
---------------------------------------------------------------------------

    The model is essentially an advanced data filtering and pre-
processing technique to improve data quality to support option pricing 
during the calibration and simulation phases of the STANS methodology. 
It makes use of the same theory that underpins OCC's Vanilla Options 
model, described below. The most important stages of the Implied 
Volatility Smoothing model are: (i) A preprocessing procedure, to 
filter out ``bad'' price quotes; (ii) an implied forward price 
calculation using prices from near-the-money options on the same 
securities at all tenors or expiration dates; (iii) the smoothing, in 
which prices are generated for all plain vanilla listed options at all 
strikes by using corresponding bid and ask price quotes and forward 
prices (from step two); and (iv) construction of a volatility surface 
based on linear interpolation of total variance among the smoothed 
prices and performing any necessary post-processing. When applied to 
prices estimated by the option pricing models described below, the 
model functions to (i) makes data points comprising the volatility 
surface more consistent with the actual bid-ask spreads found in 
current market prices and (ii) correct data that would create arbitrage 
opportunities by not having monotonicity and convexity with respect to 
the strike and/or not satisfying put-call parity.\41\
---------------------------------------------------------------------------

    \41\ See Securities Exchange Act Release No. 86296 for further 
information on the smoothing algorithm used in STANS.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on the Implied Volatility Smoothing 
model. The current rule text also includes other information related to 
the model. Specifically, the current rule text includes the following:
     A description of the use of target prices based on model 
parameters: This represents configuration choices made by OCC that are 
not inherent to the model's selection or design and that do not 
materially impact the model's results, which OCC may from time to time 
determine it should modify based on current market conditions and 
business practices.
     Economic rationale for various features of the model: OCC 
does not believe that this economic rationale is needed to understand 
how the model currently functions.
     A discussion of the model's performance in deriving 
theoretical spot prices from underlying futures and indices, and 
specific ``if/then'' conditions the model applies to bid and ask prices 
to filter out poor quality data based on certain control parameters: 
These data filtering parameters are configuration choices made by OCC 
that are not inherent to the model's selection or design and that do 
not materially impact the model's results, which OCC may from time to 
time determine it should modify based on current market conditions and 
business practices.
     Mathematical rationale for the method by which the 
smoothing algorithm calculates implied forward prices: OCC does not 
believe that the rationale for the model's configuration is needed to 
understand how the model currently functions.
     A detailed description of the Vega-weighted least squares 
calculation performed during the first round of optimization to produce 
arbitrage-free options prices for European options: This information, 
while relevant to understanding how the model functions, is 
foundational information commonly understood in quantitative finance 
and readily found in public academic sources. To the extent the text 
does not describe OCC's application of a Vega-weighted least squares 
calculation, OCC does not believe this

[[Page 85796]]

information needs to be maintained in OCC's rules.
     Operational details on (1) how the model's results are 
applied to other models for pricing European and American options, 
options on futures, and long-dated \42\ volatilities; (2) price 
smoothing for contracts that are otherwise missing smoothed prices for 
various reasons, FLEX options, and over-the-counter options; and (3) 
detailed steps for a linear interpolation/extrapolation used to 
construct a volatility surface from smoothed volatilities: These 
details relate to configuration choices made by OCC to applying a model 
overlay in certain cases where there is insufficient data, that are not 
inherent to the model's selection or design, and that do not materially 
impact the model's results, which OCC may from time to time determine 
it should modify based on current market conditions and business 
practices.
---------------------------------------------------------------------------

    \42\ In the context of volatility smoothing, ``long-dated'' 
refers to expirations beyond the listed expiration date of standard 
exchange-traded options.
---------------------------------------------------------------------------

    OCC believes that this information is more appropriately covered in 
the Implied Volatility Smoothing Model Whitepaper and other internal 
OCC documentation rather than in OCC's rules for the reasons listed 
above. Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
European and American Options
    The Vanilla Options model is used by STANS to price European and 
American options. This model is comprised of several modules that (i) 
calculate theoretical option prices, (ii) calculate risk sensitivities 
of the option prices with respect to the market variables (the 
``Greeks''), and (iii) calculate implied volatilities from option 
prices. The model prices European options using a modified Black-
Scholes formula and American options using a Leisen-Reimer binomial 
tree.\43\
---------------------------------------------------------------------------

    \43\ See Securities Exchange Act Release No. 86296 for further 
information on OCC's Vanilla Options model, which prices American 
and European options and generic futures.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on the Vanilla Options model. The current 
rule text includes other information related to the model. 
Specifically, the current rule text includes the following:
     Rationale and testing to support the number of steps used 
in the Leisen-Reimer binomial tree: OCC does not believe the rationale 
to support this model choice is needed to understand how the model 
currently functions.
     Equations describing the calculation of various ``Greeks'' 
(i.e., Gamma, Vega, Theta, and Rho), restatements of standard Black's 
formulas, and a restatement of the standard Leisen-Reimer binomial 
tree: This information, while relevant to understanding how the model 
functions, is foundational information commonly understood in 
quantitative finance and readily found in public academic literature. 
To the extent the text does not describe OCC's application of the 
``Greeks,'' Black's formulas, and the Leisen-Reimer binomial tree, OCC 
does not believe this information needs to be maintained in OCC's 
rules.
     A list of control parameters of the Newton-Raphson method 
used to calculate implied volatilities for vanilla options: These 
control parameters are configuration choices made by OCC that are not 
inherent to the model's selection or design and that do not materially 
impact the model's results, which OCC may from time to time determine 
it should modify based on current market conditions and business 
practices.
    OCC believes that this information is more appropriately covered in 
the Vanilla Options Model Whitepaper and other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
Asian FLEX Options
    Like European options, Asian FLEX options are priced based on a 
Black-Scholes formula.\44\ Asian FLEX options are modeled with 
assumptions that volatility, interest rates and cost of carry remain 
constant across an option's tenor. Furthermore, implied volatility is 
determined from ``terminal'' option (i.e., the last option in a series) 
volatilities, which are obtained from prices of available regular 
options expiring at the same tenor or, in their absence, by 
interpolating terminal volatilities of existing tenor regular options 
using an internal calculator developed by OCC.
---------------------------------------------------------------------------

    \44\ See Securities Exchange Act Release No. 74966 for further 
information on how STANS models Asian-style options.
---------------------------------------------------------------------------

Cliquet Options
    STANS also prices Cliquet options using a Black-Scholes model. Like 
Asian FLEX options, Cliquet options are modeled with assumptions that 
volatility, interest rates, and cost of carry remain constant across an 
option's tenor. STANS calculates options premiums based on the premiums 
of the individual forward starting options that comprise the Cliquet 
option. This valuation is then repeated for each ``reset period'' of 
the Cliquet option.
Forward Start Options
    STANS can also be used to price forward start options. Forward 
start options are options for which the strike price in dollars is 
unknown prior to the determination date of the strike shortly before 
expiration.\45\ Forward start option values depend on the same input 
model parameters as vanilla options and on the determination date of 
the strike. Using the Black-Scholes framework, the pricing problem of a 
forward-start option prior to strike determination can be transformed 
into the valuation of a plain vanilla option at determination time, 
after which the option can be priced using a standard application of 
Black's formula.
---------------------------------------------------------------------------

    \45\ Instead, forward start options trade with strikes defined 
as a fraction [alpha], known prior to expiration, of the underlying 
closing price on the determination date.
---------------------------------------------------------------------------

iv. Aggregation
    The STANS Methodology Description would next describe how STANS 
applies the theoretical derivatives prices to actual positions in 
Clearing Members' accounts to calculate margin requirements. This is 
accomplished by aggregating (i) a base margin charge, which consists of 
an ES calculation with the addition of Extreme Value Theory (``EVT'') 
loss modeling and a stress test component; (ii) an error compensation 
charge; (iii) a liquidation cost charge; (iv) a positive risk reversal 
charge; and (v) various add-on charges that are applied based on 
accounting principles.
Base Margin Charge
    STANS first calculates the base margin charge. This is accomplished 
by identifying the positions present in a Clearing Member's 
account,\46\ multiplying the values of those positions to each of the 
10,000 theoretical values calculated in the above step, then adding the 
products' values together to obtain possible 10,000 net asset values 
(``NAVs'') for the account. The account's actual NAV is then subtracted 
from each of these 10,000 possible NAVs to obtain 10,000 possible 
Profit and Loss (``P&L'') statements. STANS then constructs a VaR line 
separating the 100 most extreme negative projected P&L

[[Page 85797]]

statements over a two-day horizon from the remaining 9,900 simulated 
outcomes, representing the worst 1% of the projected scenarios, and 
calculates the average of these 100 values to obtain a single ES value 
for the account. This is called the empirical ES because STANS uses 
actual historical prices in calibrating the simulation, which 
represents the historical dependence among the various risk factors.
---------------------------------------------------------------------------

    \46\ The netting/offsetting of a Clearing Member's positions 
within an account pursuant to OCC's rules occurs outside STANS 
before the position data is brought into STANS for this step.
---------------------------------------------------------------------------

    In addition to calculating the empirical ES, STANS applies EVT to 
parametrically fit the largest losses and parametrically calculate ES. 
EVT is based on a tenet of probability theory that the distribution of 
extremes of a univariate random variable converge to a Generalized 
Pareto distribution.\47\ The parametric EVT estimator can use a larger 
tail sample than the empirical estimator, which, for ES at the 99th 
percentile, is limited to 100 (i.e., 1% of 10,000) points. Empirical ES 
is used when there is indication that the tail is not well fit by EVT.
---------------------------------------------------------------------------

    \47\ A Generalized Pareto distribution is a type of continuous 
probability distribution that can be used to model the distribution 
of the tail of another underlying distribution.
---------------------------------------------------------------------------

    STANS next applies a stress test component to its base charge. This 
component includes additional calculations related to (i) 
concentration, which is intended to consider extreme idiosyncratic 
moves in concentrated positions and to counteract ``survivor bias'' in 
historical equity returns data (i.e., that historical data typically 
does not incorporate certain extreme movements in a firm's stock 
prices, such as when a firm declares bankruptcy or is subject to a rich 
takeover); and (ii) dependence, in which the ES calculations described 
above are performed twice again, once assuming perfect correlation 
among the various risk factors and once assuming no correlation among 
the various risk factors. After performing these concentration and 
dependence calculations, STANS takes the higher of the two factors and 
combines it with the empirical ES to create a more conservative margin 
requirement for the account.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's chapter on the base charge, stress-test add-on 
charge, and total margin charge. The current rule text also includes a 
summary section summarizing historical changes OCC has made to the 
manner in which STANS calculates a total margin charge. OCC does not 
believe this information is needed to understand how STANS currently 
functions. OCC further believes that this information is more 
appropriately covered in the Portfolio Risk Measures Model Whitepaper 
or other internal OCC documentation rather than in OCC's rules for this 
reason. Therefore, OCC proposes to delete this rule text in its 
entirety without adding new, corresponding rule text in the STANS 
Methodology Description.
Error Compensation
    An inherent property of ES calculations is the existence of 
estimation error. To compensate for the potential risk that a STANS ES 
calculation includes such an error on the positive (lower loss) side, 
the ES value based on the simulated results is shifted through a 
compensation term to a more conservative level. Mathematically, the 
error compensator shifts ES to the left by 1.2 standard deviations of 
the loss tail, covering the 70% quantile of estimation error. The 
extent to which this alters the calculated ES in absolute varies based 
on the distribution's kurtosis (i.e., the shift is more significant for 
distributions with fatter tails).
Liquidation Cost Charge
    The default of a Clearing Member requires OCC to close-out that 
Clearing Member's positions, which results in OCC incurring costs. 
Closing out positions in a defaulted portfolio may also entail selling 
long positions at the current bid price and covering short positions at 
the current ask price, which could create additional costs based on the 
bid-ask spread. To account for these costs, STANS calculates a daily 
liquidation cost charge based on a liquidation cost grid, calibrated 
with data from historical stressed periods, and applies this calculated 
cost as an add-on charge. In general, the Liquidation Charge model 
calculates two risk-based liquidation costs for a portfolio, Vega \48\ 
liquidation cost (``Vega LC'') and Delta liquidation cost (``Delta 
LC''), using ``Liquidation Grids.'' More specifically, the model 
consists of: (1) The decomposition of the defaulter's portfolio into 
sub-portfolios by underlying security; (2) the creation and calibration 
of Liquidation Grids used to determine liquidation costs; (3) the 
calculation of the Vega LC (including a minimum Vega LC charge) for 
options products; (4) the calculation of Delta LCs for both options and 
Delta-one products; (5) the calculation of Vega and Delta concentration 
factors; and (6) the calculation of volatility correlations for Vega 
LCs.\49\ STANS applies both Vega and Delta LCs to options products, but 
only applies a Delta charge to Delta-one \50\ products such as futures 
contracts, Treasury securities, and equity securities.
---------------------------------------------------------------------------

    \48\ The Delta and Vega of an option represent the sensitivity 
of the option price with respect to the price and volatility of the 
underlying security, respectively.
    \49\ See Securities Exchange Act Release No. 85755 for more 
detail on the liquidation cost model used by STANS.
    \50\ ``Delta one products'' refer to products for which a change 
in the value of the underlying asset results in a change of the 
same, or nearly the same, proportion in the value of the product.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on the Liquidation Charge model. The 
current rule text also includes other information related to the model. 
Specifically, the current rule text includes the following:
     Background historical information on adoption of the 
model: OCC does not believe this historical information is needed to 
understand how the model currently functions.
     Classifications OCC applies to an underlying equity 
security based on the security's liquidity level to determine which 
liquidation grid is most appropriate: These details represent 
configuration choices made by OCC that are not inherent to the model's 
selection or design and that do not materially impact the model's 
results, which OCC may from time to time determine it should modify 
based on current market conditions and business practices.
     Intermediate equations used to define variables for 
calculating Vega LC: OCC does not believe these intermediate, 
explanatory equations are needed to understand how the model currently 
functions.
     Descriptions of the parameters used to calibrate 
liquidation grids: These calibration parameters represent configuration 
choices made by OCC that are not inherent to the model's selection or 
design and that do not materially impact the model's results, which OCC 
may from time to time determine it should modify based on current 
market conditions and business practices.
    OCC believes that this information is more appropriately covered in 
the underlying Liquidation Charge Model Whitepaper and other internal 
OCC documentation rather than in OCC's rules for the reasons listed 
above. Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
Positive Risk Reversal
    As an additional conservative measure, STANS applies a ``positive

[[Page 85798]]

risk reversal'' charge to ensure that the total calculated margin 
requirement is at least equal to the estimated liquidation cost, even 
in the event a position is liquidated at the current market price (or a 
more favorable price). STANS incorporates the positive risk reversal 
charge by simply applying a minimum margin requirement to a position 
that is equal to the estimated liquidation cost charge.
Various Add-On Charges
    In addition to the charges described above, OCC may, pursuant to 
its rules, elect to apply additional charges to a Clearing Member's 
margin requirements for various reasons; e.g., based on the Clearing 
Member's Watch Level status or to account for rebates, adjustments and 
add-ons related to stock loan positions.\51\ These additional charges 
occur outside of STANS and are outside the scope of the STANS 
Methodology Description.
---------------------------------------------------------------------------

    \51\ See Securities Exchange Act Release No. 82355, which states 
that OCC's Margin Policy establishes the application of add-on 
charges.
---------------------------------------------------------------------------

    The proposed text would replace current OCC rule text from a 
section in the Margins Methodology's base charge, stress-test add-on 
charge, and total margin charge chapter covering add-on charges. The 
current rule text notes that OCC may apply various add-on charges to 
its margin requirements outside the STANS methodology, which could 
include additional margin charges related to (i) cross-margin accounts, 
established by OCC Rule 704; (ii) placement on a heightened Watch Level 
based on OCC's credit risk surveillance, established by OCC's 
Counterparty Credit Risk Management Policy; \52\ (iii) interest 
payments and adjustments to stock loan positions, established by OCC 
Rule 601, Interpretation & Policy .05; (iv) customer positions subject 
to certain margin requirements promulgated by the U.S. Commodity 
Futures Trading Commission, established by OCC Rule 601, Interpretation 
& Policy .07; (v) concentration risk for equity securities exceeding an 
average daily trading volume threshold, established by OCC's Collateral 
Risk Management Policy; \53\ and (vi) OCC's extended trading hours 
program, established generally by OCC's Margin Policy and specified in 
OCC's Extended Trading Hours Set-Up and Monitoring Procedure.\54\
---------------------------------------------------------------------------

    \52\ See Securities Exchange Act Release No. 81949 (October 26, 
2017), 82 FR 50719 (November 1, 2017) (SR-OCC-2017-009) for more 
information on OCC's Watch Level framework. OCC has filed a proposed 
rule change with the Commission to adopt a new TPRMF, which would 
replace the Counterparty Credit Risk Management Policy and provide 
an overview of OCC's overall approach to third-party risk 
management. See supra note 3.
    \53\ See Securities Exchange Act Release No. 82009, which 
describes OCC's Collateral Risk Management Policy.
    \54\ The specific margin add-on charges OCC may apply are 
subject to change in accordance with internal governance established 
by OCC's Margin Policy and supporting procedures.
---------------------------------------------------------------------------

    As outlined above, these add-on charges are applied pursuant to 
other OCC rules, policies, and/or procedures, and are established 
outside of the STANS methodology. Therefore, OCC believes that they are 
more appropriately covered in the underlying OCC rules, policies, and 
procedures that establish them, and, accordingly, proposes to delete 
this text in its entirety without adding new, corresponding rule text 
in the STANS Methodology Description.
Model Utilities
    The STANS Methodology Description would next describe several 
``model utilities'' that are applied at various points in the STANS 
methodology, to incorporate various market and operational factors that 
affect options pricing and thereby produce model results which more 
accurately reflect current and potential market conditions.
i. Dividends
    The STANS Methodology Description would describe how STANS 
incorporates expected cash dividends on a stock into options 
pricing.\55\ STANS obtains daily information on general dividend yields 
and discrete dividends from pricing vendors, then applies a dividend 
growth rate to this information to forecast dividends (typically) 16 
quarters into the future.
---------------------------------------------------------------------------

    \55\ OCC considers the potential effects of stock dividends 
outside of STANS.
---------------------------------------------------------------------------

    STANS accounts for the possibility that cash dividends may be paid 
on stocks, which would affect their pricing, through a dividend utility 
that interacts with the pricing models in STANS. Daily, STANS retrieves 
from an external vendor data on forecasted cash dividends and yield 
curves associated with the issuance of those dividends. STANS uses this 
data to forecast when a security may go ex-dividend, and accordingly 
incorporates this into pricing the associated equity security. STANS 
also accounts for the possibility that an option may be exercised early 
to obtain a cash dividend on the underlying security. Using the same 
external dividend data, STANS calculates when an option would likely be 
exercised early to receive the dividend and prices it accordingly.
ii. Interest Rate Curve
    This model utility calculates the yield curve using (i) overnight, 
one-week, one-month, two-month, and three-month cash deposit interest 
rates; (ii) Eurodollar interest rate futures with three-month to two-
year tenors; and (iii) interest rate swaps with three-year to 30-year 
tenors. The model utility calculates a discount factor from a given 
date to any future date along the curve. This discount factor is used 
as an input to pricing models to generate theoretical prices for 
instruments based on these rates.
iii. Overnight and Daily Returns
    STANS calculates margin requirements on a daily basis, using prices 
from that day's market close. However, some positions may expire or be 
exercised during a business day and before the following day's margin 
settlement. Since OCC clears derivatives that are settled on both 
opening and closing prices, both types of events affect derivatives 
prices and their corresponding margin requirements. Therefore, the 
STANS Methodology Description would describe how STANS obtains relevant 
risk factors for both the most recent opening price and the most recent 
closing price. STANS includes within the copula it constructs, 
described previously, a joint distribution of both overnight and daily 
returns on relevant risk factors.
    The proposed text would replace current OCC rule text from the 
Margins Methodology's section on overnight and daily innovations. The 
current rule text also includes other information on the overnight and 
daily returns model utility. Specifically, the current rule text 
includes the following:
     Details on how OCC implemented the model utility in its 
technology systems: These implementation details relate to OCC's 
internal administration of its technology systems and are not needed to 
understand how the model currently functions. Because these details are 
not inherent to the model's selection or design, OCC could also change 
them from time to time without affecting the model's results.
     Redundant detail related to the copula constructed by 
STANS: These details, described above, would be described in the STANS 
Methodology Description's section on the Student-t Copula model, and 
OCC does not believe repeating it here is needed to understand how the 
model utility currently functions.
    OCC believes that this information is more appropriately covered in 
the Daily and Overnight Theoretical Price

[[Page 85799]]

Scenario Simulation Model Whitepaper or other internal OCC 
documentation rather than in OCC's rules for the reasons listed above. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
iv. One-Day and Two-Day Scenarios
    As noted previously, OCC has established margin levels to cover the 
costs of liquidating positions over a two-day margin period of risk. 
Furthermore, and as described above, during this interval expiring OCC-
cleared or cross-margined positions may experience final settlement 
based on either the opening or closing price of the underlying 
security. Therefore, the STANS Methodology Description would describe 
how STANS calculates for each underlying price scenario simulated 
prices at four different points in time: (i) Opening price on day one; 
(ii) closing price on day one; (iii) opening price on day two; and (iv) 
closing price on day two. STANS must account for these additional 
prices to avoid under-margining portfolios with both expiring and non-
expiring positions in a risk group, and to reflect the prices of 
underlying securities and associated derivatives that are forecasted to 
go ex-dividend or ex-coupon on T+1 or T+2 (where T represents the 
activity date). To calculate the two additional prices that may be 
observed over the two-day margin period of risk, STANS applies a 
randomly generated permutation to the return scenarios. The second-day 
return scenarios and securities that go ex-dividend on T+2 are then 
applied scenario-by-scenario to the first-day results in the same 
fashion.
v. Portfolio Specific Haircuts
    Some Clearing Members have deposited securities as margin 
collateral that are also used in STANS as risk factors, and therefore 
potential price movements in these securities are factored into margin 
requirement calculations. However, a Clearing Member may want--or be 
required--to withdraw or deposit such margin collateral intraday. This 
would change the concentration of the Clearing Member's collateral 
types and would also change the sensitivity of how the Clearing 
Member's portfolio responds to such changes. To account for these 
changes in concentration and sensitivity, the STANS Methodology 
description would describe how STANS utilizes a Portfolio Specific 
Haircuts model. This model provides haircut values for withdrawals or 
deposits of collateral, which are then applied to the previous day's 
collateral values to arrive at the impact of the collateral movements 
on the margin requirement. These haircuts represent the sensitivity of 
that Clearing Member account's risk profile to its position in the 
collateral security being withdrawn or deposited. These haircuts are 
designed to provide an estimate of the resulting change in margin 
requirements if the entire margin calculation were re-run following the 
withdrawal or deposit. A different haircut is associated with each 
combination of Clearing Member account and security posted as margin 
collateral.
Margins Methodology Chapters Not Found in STANS Methodology Description
    The current rule text from the Margins Methodology describes that 
STANS uses historical and current prices for listed tenors of energy 
and other commodity futures to simulate prices of energy and other 
commodity futures using two variants of a two-factor Schwartz and 
Smith's model: \56\ One variant to incorporate the effects of 
seasonality \57\ for pricing futures related to nonseasonal commodities 
such as crude oil and the other variant to incorporate the effects of 
seasonality and is used to price futures related to seasonal 
commodities such as natural gas, heating oil, gasoline, electricity, 
and petrochemicals. The products for which OCC previously used this 
model to calculate margin requirements are no longer listed, and 
therefore OCC has decommissioned this associated pricing model. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
---------------------------------------------------------------------------

    \56\ The Schwartz and Smith's model is a two-factor model of 
commodity prices that allows for mean reversion in short-term prices 
and uncertainty in the long-term equilibrium level to which prices 
revert. See Schwartz, E. and Smith, E., ``Short-Term Variations and 
Long-Term Dynamics in Commodity Prices,'' 46 Mgmt. Sci. 7, 893-911 
(2000) (describing the Schwartz and Smith's model).
    \57\ Seasonality is a characteristic of futures products that 
exhibit regular and predictable price changes that recur every 
calendar year.
---------------------------------------------------------------------------

    The current text from the Margins Methodology also includes 
information on a model used to price European-style binary options. The 
products for which OCC used this model to calculate margin requirements 
are no longer listed, and OCC decommissioned the model. The current 
text also includes information on OCC's use of the Vanilla Options 
model to calculate margin requirements for Currency Options and Foreign 
Index Futures, both of which are products OCC no longer clears. 
Therefore, OCC proposes to delete this rule text in its entirety 
without adding new, corresponding rule text in the STANS Methodology 
Description.
    The current rule text from the Margins Methodology also includes 
information on certain processes OCC uses to operationalize the STANS 
methodology in its systems. Specifically, these processes are (i) daily 
calibration and transfer, which describes implementation of the 
processes to daily obtain pricing data and calibrate pricing models; 
(ii) Monte Carlo marginals, which describes implementation of the 
processes that create price scenarios for underlying risk factors from 
either copula draws or (in the absence of a copula) conditional or 
default simulations; (iii) Monte Carlo theoreticals, which describes 
implementation of the processes that calculate theoretical values for 
futures and options; and (iv) monthly copula estimation and simulation, 
which describes implementation of the processes that generate copula 
scenarios for underlying risk factors based on calibrated parameters.
    These chapters describe implementation details related to OCC's 
internal administration of its technology systems and are not needed to 
understand how the STANS models currently function. Because these 
details are not inherent to model selection or design, OCC could also 
change them from time to time without affecting model results. OCC 
believes that this information is more appropriately covered in the 
underlying Model Whitepapers and other internal OCC documentation 
rather than in OCC's rules for this reason. Therefore, OCC proposes to 
delete this rule text in its entirety without adding new, corresponding 
rule text in the STANS Methodology Description.
Margin Policy
    Lastly, OCC would make conforming changes to its Margin Policy to 
reflect the adoption of the STANS Methodology Description and the 
retirement of the Margins Methodology. OCC would also make other non-
substantive changes to the Margin Policy to correct typographical 
errors, update references to other related internal OCC policies and 
procedures, and conform the policy to OCC's current internal policy 
template. The proposed changes are intended to promote the accuracy and 
clarity of OCC's Margin Policy and would not impact OCC's margin 
setting practices or processes.

[[Page 85800]]

(2) Statutory Basis
    OCC believes that the proposed rule change is consistent with 
Section 17A of the Act \58\ and the rules thereunder applicable to OCC. 
Section 17A(b)(3)(F) of Act \59\ requires, among other things, that the 
rules of a clearing agency be designed to promote the prompt and 
accurate clearance and settlement of securities transactions and 
derivative agreements, contracts, and transactions. The purpose of the 
proposed rule change is to adopt a STANS Methodology Description 
document to clearly and concisely describe the material aspects of 
OCC's quantitative methodology for calculating Clearing Member margin 
requirements. OCC uses the margin it collects to limit its credit 
exposures to participants and to protect other Clearing Members from 
losses that may arise as a result of a default and ensure that OCC is 
able to continue the prompt and accurate clearance and settlement of 
its cleared products. As a result, OCC believes the proposed STANS 
Methodology Description is designed to promote the prompt and accurate 
clearance and settlement of securities transactions and derivative 
agreements, contracts, and transactions in accordance with Section 
17A(b)(3)(F) of the Act.\60\
---------------------------------------------------------------------------

    \58\ 15 U.S.C. 78q-1.
    \59\ 15 U.S.C. 78q-1(b)(3)(F).
    \60\ Id.
---------------------------------------------------------------------------

    Rule 17Ad-22(b)(1) \61\ requires that a registered clearing agency 
that performs central counterparty services establish, implement, 
maintain and enforce written policies and procedures reasonably 
designed to measure its credit exposures to its participants at least 
once a day and limit its exposures to potential losses from defaults by 
its participants under normal market conditions so that the operations 
of the clearing agency would not be disrupted and non-defaulting 
participants would not be exposed to losses that they cannot anticipate 
or control. As described above, the proposed STANS Methodology 
Description described herein details the risk-based margin methodology 
by which OCC measures its credit exposures to its participants on a 
daily basis and determines margin requirements based on such 
calculations. OCC believes that the proposed STANS Methodology 
Description would result in a more transparent and clearly 
understandable description of the methodology used to measure and 
mitigate credit exposures to OCC's Clearing Members, and that the 
proposed rule change is therefore designed to ensure that OCC sets 
margin requirements that would serve to limit OCC's exposures to 
potential losses from defaults by its participants under normal market 
conditions so that the operations of OCC would not be disrupted, and 
non-defaulting participants would not be exposed to losses that they 
cannot anticipate or control. Accordingly, OCC believes the proposed 
rule change is consistent with Rule 17Ad-22(b)(1).\62\
---------------------------------------------------------------------------

    \61\ 17 CFR 240.17Ad-22(b)(1).
    \62\ Id.
---------------------------------------------------------------------------

    Rule 17Ad-22(b)(2) \63\ further requires, in part, that a 
registered clearing agency that performs central counterparty services 
establish, implement, maintain and enforce written policies and 
procedures reasonably designed to use margin requirements to limit its 
credit exposures to participants under normal market conditions and use 
risk-based models and parameters to set margin requirements. The STANS 
Methodology Description is intended to better describe how the STANS 
methodology is designed to limit OCC's credit exposures to participants 
under normal market conditions in a manner consistent with Rule 17Ad-
22(b)(2).\64\
---------------------------------------------------------------------------

    \63\ 17 CFR 240.17Ad-22(b)(2).
    \64\ Id.
---------------------------------------------------------------------------

    Rules 17Ad-22(e)(6)(i), (iii), and (v) \65\ further require that a 
covered clearing agency establish, implement, maintain and enforce 
written policies and procedures reasonably designed to cover its credit 
exposures to its participants by establishing a risk-based margin 
system that, among other things: (1) Considers, and produces margin 
levels commensurate with, the risks and particular attributes of each 
relevant product, portfolio, and market; (2) calculates margin 
sufficient to cover its potential future exposure to participants in 
the interval between the last margin collection and the close out of 
positions following a participant default; and (3) uses an appropriate 
method for measuring credit exposure that accounts for relevant product 
risk factors and portfolio effects across products. As described in 
detail above, OCC believes that the proposed STANS Methodology 
Description would result in a clearer, more transparent document 
describing OCC's risk-based margin system that, among other things: (1) 
Considers, and produces margin levels commensurate with, the risks and 
particular attributes of each relevant product, portfolio, and market; 
(2) calculates margin sufficient to cover its potential future exposure 
to participants in the interval between the last margin collection and 
the close out of positions following a participant default; and (3) 
uses an appropriate method for measuring credit exposure that accounts 
for relevant product risk factors and portfolio effects across 
products. OCC therefore believes the proposed STANS Methodology 
Description is reasonably designed to consider and produce margin 
levels commensurate with the risks and particular attributes of 
products cleared by OCC, calculate margin sufficient to cover its 
potential future exposure to participants in the interval between the 
last margin collection and the close out of positions following a 
participant default, and apply an appropriate method for measuring 
credit exposure that accounts for risk factors and portfolio effects of 
products cleared by OCC in a manner consistent with Rules 17Ad-
22(e)(6)(i), (iii), and (v).\66\
---------------------------------------------------------------------------

    \65\ 17 CFR 240.17Ad-22(e)(6)(i), (iii), and (v).
    \66\ Id.
---------------------------------------------------------------------------

    Rule 17Ad-22(e)(23) \67\ further requires, in part, that a covered 
clearing agency establish, implement, maintain, and enforce written 
policies and procedures reasonably designed to provide sufficient 
information to enable participants to identify and evaluate the risks, 
fees, and other material costs they incur by participating in the 
covered clearing agency. The STANS Methodology Description is designed 
to provide Clearing Members with greater transparency into the STANS 
Methodology than the current rule text of the Margins Methodology, 
which OCC does not make generally available to participants and 
includes various details that, as described herein, OCC does not 
believe constitute material aspects of the STANS methodology. In 
addition, OCC has organized and written the STANS Methodology 
Description in a way that would more clearly identify for Clearing 
Members the material aspects of the STANS methodology. Specifically, 
OCC has organized the STANS Methodology Description in a way that 
enables a reader to better understand how the various quantitative 
model components of STANS function in concert to produce OCC margin 
requirements, rather than organizing the document in a way that would 
serve OCC's internal purposes but not facilitate comprehension of the 
STANS methodology by a third party. Furthermore, by including in the 
STANS Methodology Description only the OCC rule text covering the 
material, quantitative aspects of the STANS methodology, and either not 
describing extraneous or immaterial aspects of the STANS methodology in 
the STANS Methodology Description or referring

[[Page 85801]]

the reader to other OCC or external sources where appropriate,\68\ the 
proposed STANS Methodology Description would more clearly identify for 
an informed reader how the STANS methodology's quantitative model 
components form OCC's basis for calculating margin requirements, and 
what aspects of the STANS methodology OCC may adjust in the course of 
its business pursuant to its other rules and internal policies and 
procedures. OCC believes that this additional clarity, transparency, 
and enhanced readability regarding the material quantitative model 
components of the STANS methodology promote the requirements of Rule 
17Ad-22(e)(23).
---------------------------------------------------------------------------

    \67\ 17 CFR 240.17Ad-22(e)(23).
    \68\ For example, the STANS Methodology Description would refer 
to other OCC rules to establish manual, non-modeled margin 
components or adjustments made by OCC, and would refer to public 
academic sources for descriptions of common mathematical theories 
and methods that do not represent OCC-specific applications or 
modifications of those theories and methods.
---------------------------------------------------------------------------

    Finally, Section 19(b)(1) of the Act and Rule 19b-4 thereunder set 
forth the requirements for SRO proposed rule changes, including the 
regulatory filing requirements for ``stated policies, practices and 
interpretations.'' \69\ OCC proposes to retire its existing Margins 
Methodology, which was, in part, previously filed as an OCC ``rule'' 
with the Commission, as the STANS Methodology Description would 
supersede the Margins Methodology in its entirety. Under the proposal, 
the material aspects of the STANS methodology would be contained in the 
proposed STANS Methodology Description described herein.
---------------------------------------------------------------------------

    \69\ See supra note 13.
---------------------------------------------------------------------------

    As described in detail herein, various details in the current 
Margins Methodology would no longer be OCC rule text following adoption 
of the STANS Methodology Description. These internal procedural and 
administrative details used by OCC's model developers and model 
validators would: (1) Be reasonably and fairly implied by the proposed 
STANS Methodology Description, OCC's Margin Policy,\70\ OCC's Model 
Risk Management Policy,\71\ and other OCC rules; and/or (2) otherwise 
not be deemed to be material aspects of OCC's risk-based margin system. 
Specifically, OCC believes the details it proposes to remove from OCC's 
rule text are consistent with Section 19(b)(1) of the Act and Rule 19b-
4 thereunder for the following reasons:
---------------------------------------------------------------------------

    \70\ See Securities Exchange Act Release No. 82355 (December 19, 
2017), 82 FR 61058 (December 26, 2017) (SR-OCC-2017-007).
    \71\ See Securities Exchange Act Release No. 82473 (January 9, 
2018), 83 FR 2271 (January 16, 2018) (SR-OCC-2017-011).
---------------------------------------------------------------------------

     To the extent the current rule text includes details on 
OCC's historical modeling practices and potential future enhancements, 
OCC does not believe such text constitutes an SPPI of OCC because it 
does not describe OCC's current practices;
     To the extent the current rule text includes details on 
the exact set of current products applied to each STANS component, 
which will change from time to time as OCC-cleared products are listed 
and delisted, OCC believes such text is reasonably and fairly implied 
by the proposed rule text establishing the scope of instruments for 
which the STANS methodology calculates margin requirements;
     To the extent the current rule text includes details on 
various configuration choices made by OCC, such as data sources, model 
parameters, and model performance monitoring, that are not inherent to 
model selection or design and that do not materially impact a model's 
results, which OCC may from time to time determine it should modify 
based on current market conditions and business practices, OCC does not 
believe such text constitutes an SPPI because it does not describe a 
material aspect of the operation of the facilities of OCC;
     To the extent the current rule text includes details on 
testing results and explanatory rationale supporting a model, OCC does 
not believe such text constitutes an SPPI because it does not describe 
an OCC policy, practice, or interpretation;
     To the extent the current rule text includes recitations 
of standard mathematical and economic theories/techniques that are 
well-known in quantitative finance, readily found in public sources, 
and do not include OCC-specific modifications or applications, OCC 
believes such text is reasonably and fairly implied by the rule text 
establishing the theories/techniques selected by OCC if OCC has not 
applied such theories/techniques in a modified or idiosyncratic manner;
     To the extent the current rule text includes redundant 
descriptions of a model component appearing in multiple chapters, the 
rule text has been consolidated to describe the model component in the 
single location;
     To the extent the current rule text includes details on 
OCC's implementation of a model in its internal technology systems and 
application of model results in operational procedures that are not 
inherent to a model and that OCC could change them from time to time 
without affecting a model's results, OCC does not believe such text 
constitutes an SPPI because (1) it does not describe a material aspect 
of the operation of the facilities of OCC and (2) it is reasonably and 
fairly implied that the calculations described in the STANS Methodology 
Description must be implemented in some manner through internal OCC's 
systems and processes. For example, current chapters of the Margins 
Methodology describe the processes run by internal OCC systems to 
execute the calculations described in the proposed STANS Methodology 
Description. These chapters do not describe material aspects of OCC's 
models or methodology. Rather, these chapters describe, for example, 
the timing and sequencing of various processes and the code libraries 
maintained by OCC to support the STANS methodology. Changes in such 
processes would not be considered changes to OCC's models/methodology 
and would not materially impact OCC's margin requirements. Moreover, 
Clearing Members and market participants can reasonably and fairly 
infer that OCC maintains such systems and processes to effectuate the 
daily calculation of margin requirements using the models and 
methodology described herein; and
     To the extent the current rule text includes manual margin 
adjustments and add-ons OCC employs pursuant to OCC rules, policies, 
and/or procedures outside the STANS methodology, OCC does not believe 
such text constitutes an SPPI because it is reasonably and fairly 
implied by other existing rules of OCC.
    Accordingly, OCC believes the proposed changes would be consistent 
with the requirements of Section 19(b)(1) of the Act and Rule 19b-4 
thereunder.\72\
---------------------------------------------------------------------------

    \72\ See 15 U.S.C. 78s(b)(1) and 17 CFR 240.19b-4.
---------------------------------------------------------------------------

(B) Clearing Agency's Statement on Burden on Competition

    Section 17A(b)(3)(I) of the Act requires that the rules of a 
clearing agency do not impose any burden on competition not necessary 
or appropriate in furtherance of the purposes of Act.\73\ OCC does not 
believe that the proposed rule change would impact or impose any burden 
on competition. The proposed STANS Methodology Description describes 
OCC's STANS margin setting methodology that currently applies to all 
Clearing Members. Therefore, the proposal has no impact on Clearing 
Members, and OCC does not believe that the proposed rule change would

[[Page 85802]]

unfairly inhibit access to OCC's services or disadvantage or favor any 
particular user in relationship to another user. In addition, the 
proposal currently applies uniformly to all Clearing Members in 
establishing their margin requirements.
---------------------------------------------------------------------------

    \73\ 15 U.S.C. 78q-1(b)(3)(I).
---------------------------------------------------------------------------

    For the foregoing reasons, OCC believes that the proposed rule 
change is in the public interest, would be consistent with the 
requirements of the Act applicable to clearing agencies, and would not 
impact or impose a burden on competition.

(C) Clearing Agency's Statement on Comments on the Proposed Rule Change 
Received From Members, Participants or Others

    Written comments were not and are not intended to be solicited with 
respect to the proposed rule change and none have been received.

III. Date of Effectiveness of the Proposed Rule Change and Timing for 
Commission Action

    Within 45 days of the date of publication of this notice in the 
Federal Register or within such longer period up to 90 days (i) as the 
Commission may designate if it finds such longer period to be 
appropriate and publishes its reasons for so finding or (ii) as to 
which the self- regulatory organization consents, the Commission will:
    (A) By order approve or disapprove the proposed rule change, or
    (B) institute proceedings to determine whether the proposed rule 
change should be disapproved.

IV. Solicitation of Comments

    Interested persons are invited to submit written data, views and 
arguments concerning the foregoing, including whether the proposed rule 
change is consistent with the Exchange Act. Comments may be submitted 
by any of the following methods:

Electronic Comments

     Use the Commission's internet comment form (http://www.sec.gov/rules/sro.shtml); or
     Send an email to rule-comments@sec.gov. Please include 
File Number SR-OCC-2020-016 on the subject line.

Paper Comments

     Send paper comments in triplicate to Secretary, Securities 
and Exchange Commission, 100 F Street NE, Washington, DC 20549-1090.

All submissions should refer to File Number SR-OCC-2020-016. This file 
number should be included on the subject line if email is used. To help 
the Commission process and review your comments more efficiently, 
please use only one method. The Commission will post all comments on 
the Commission's internet website (http://www.sec.gov/rules/sro.shtml). 
Copies of the submission, all subsequent amendments, all written 
statements with respect to the proposed rule change that are filed with 
the Commission, and all written communications relating to the proposed 
rule change between the Commission and any person, other than those 
that may be withheld from the public in accordance with the provisions 
of 5 U.S.C. 552, will be available for website viewing and printing in 
the Commission's Public Reference Room, 100 F Street NE, Washington, DC 
20549, on official business days between the hours of 10:00 a.m. and 
3:00 p.m. Copies of such filing also will be available for inspection 
and copying at the principal office of OCC and on OCC's website at 
https://www.theocc.com/Company-Information/Documents-and-Archives/By-Laws-and-Rules#rule-filings.
    All comments received will be posted without change. Persons 
submitting comments are cautioned that we do not redact or edit 
personal identifying information from comment submissions. You should 
submit only information that you wish to make available publicly.
    All submissions should refer to File Number SR-OCC-2020-016 and 
should be submitted on or before January 19, 2021.
---------------------------------------------------------------------------

    \74\ 17 CFR 200.30-3(a)(12).

    For the Commission, by the Division of Trading and Markets, 
pursuant to delegated authority.\74\
J. Matthew DeLesDernier,
Assistant Secretary.
[FR Doc. 2020-28662 Filed 12-28-20; 8:45 am]
BILLING CODE 8011-01-P


