Jpmam

George C.W. Gatch
Managing Director
CEO – Global Funds Management
& Institutional

January 13, 2016
Mr. Brent Fields
Secretary
Securities and Exchange Commission
100 F Street, NE
Washington, DC 20549-1090
Re: Open-End Fund Liquidity Risk Management Programs; Swing Pricing; Re-Opening of
Comment Period for Investment Company Reporting Modernization Release (File No. S716-15)
Dear Mr. Fields:
J.P. Morgan Asset Management (“JPMAM”)1 is pleased to respond to the Securities and Exchange
Commission’s (the “SEC” or the “Commission”) request for comment on its proposal on liquidity
risk management programs for open-end funds (the “proposal”). JPMAM offers 155 mutual funds
and ETFs in the US (excluding money market funds), with a total of approximately $268 billion in
assets under management at the end of 2015.
JPMAM supports the SEC’s goals of promoting effective liquidity risk management throughout the
open-end fund industry, reducing the risk that funds will be unable to meet redemption obligations,
and mitigating dilution of the interests of fund shareholders. We believe the SEC, as the primary
regulator of the asset management industry, is well placed to promulgate regulations regarding
liquidity risk management, and to supervise funds and monitor potential liquidity risks on an
ongoing basis.
More specifically, JPMAM supports the SEC’s general approach to liquidity risk management for
funds. We believe enhanced liquidity monitoring and reporting obligations, along with an enhanced
oversight and governance framework, will improve funds’ ability to manage their liquidity risks. At
the same time, we appreciate the Commission’s recognition that, for a range of reasons, different
funds have different liquidity profiles, and therefore a one-size-fits-all approach to liquidity
management (such as by requiring specified cash buffers or other liquidity allocations) would not be
helpful.

1

J.P. Morgan Asset Management is a marketing name for the investment management subsidiaries of JPMorgan.
270 Park Avenue, New York, New York 10017-2014
Telephone: 212 648 2357 Facsimile: 212 648 2587 [email protected]
J.P. Morgan Investment Management Inc.
“J.P.Morgan Asset Management is the marketing name for the asset management businesses of JPMorgan Chase & Co. those businesses include, but are not limited to,
J.P. Morgan Investments Management Inc., JPMorgan Investment Advisors, Inc., Security Capital Research & Management Incorporated and J.P. Morgan Alternative Asset Management, Inc.”

We write to provide comments on certain critical elements of the proposal. Our comments
underscore our view that markets and liquidity are dynamic. We support the SEC’s proposed
approach of assessing a the liquidity of a fund’s portfolio securities on a continuum from most to
least liquid, rather than the current bilateral construct under SEC guidance, in which assets are
considered to be liquid if they are not in the 15% illiquid bucket. We believe that, even when market
conditions change dramatically, relative assessments of liquidity should typically hold, i.e., assets
classified as more liquid are likely to remain more liquid than those classified as less liquid. At the
same time, during highly stressed markets, even the most liquid assets could experience more price
volatility and lower market depth than during ordinary markets. For these reasons, we have
concerns about how the proposal articulates its liquidity categories, as well as the concept of the
“three-day liquid asset minimum.” We also offer our views on the proposal’s reporting and
disclosure requirements, and swing pricing.
Our comments can be summarized as follows:


Classification of Liquidity of Portfolio Positions: JPMAM supports a requirement that funds
classify their portfolio holdings across a spectrum of liquidity. However, we have several
concerns with the proposed approach, including the granularity of and precision implied by
the proposed six buckets based on the estimated days to liquidate a position, as well as the
requirement that a sale must not materially impact the price of the asset. We offer an
alternative approach that we believe would meet the Commission’s objectives of enhancing a
fund’s ability to monitor and manage its liquidity and providing sufficiently granular data for
the Commission to monitor liquidity risk in funds and identify outliers.



Management of Liquidity Risk – Three-Day Liquid Asset Minimum: JPMAM supports
improving funds’ consideration of, and placing a governance structure around, the
maintenance of highly liquid assets. We are concerned, however, that the proposed threeday liquid asset minimum, which requires board approval, is not sufficiently dynamic to
respond to changing market conditions, and does not offer sufficient flexibility to manage
funds efficiently. We also have concerns about the proposed liquidity classifications based
on “days to liquidate.” We recommend instead that the board oversee a management
committee that supervises a highly liquid asset threshold.



Reporting and Disclosure Requirements: While we support monthly reporting to the SEC
of the information described in the rule (subject to our comments), as well as additional
disclosure to investors on Form N-1A, we question the value of quarterly disclosure of
liquidity classifications and the three-day liquid asset minimum to the public. We are
concerned that the classifications could be used to provide outdated metrics about a fund’s
portfolio. Disclosure of the liquid asset minimum or threshold could also have unintended
consequences.

2



Swing Pricing: We generally support the proposal to permit funds to use swing pricing. We
recommend that the Commission permit, but not require, the consideration of market
impact in setting a swing factor, and we suggest more flexibility in the factors for establishing
a swing threshold. We also describe some operational challenges that we hope the industry
will resolve should the Commission permit swing pricing.

Classification of Liquidity of Portfolio Positions
JPMAM supports the concept of requiring funds to analyze the liquidity of their portfolios, and to
monitor these classifications on an ongoing basis. We believe liquidity exists on a spectrum from
highly liquid to illiquid. Analysis of how a portfolio’s assets are spread across this range is an
important input into the assessment, monitoring, and management of a fund’s liquidity risks.
While we support the concept of a position-level assessment of liquidity, we have several comments
and concerns with the proposed approach to this assessment – specifically, the proposed
classification of funds’ positions into six categories based on the time period in which the position
“would be convertible to cash at a price that does not materially affect the value of that asset
immediately prior to sale.” Our concerns are set out below, after which we offer an alternative
approach.
As a preliminary matter, the SEC’s proposed factors to be considered in classifying the liquidity of
positions are generally consistent with our view of relevant inputs to such an analysis.2 However,
because we agree that the enumerated factors may not apply to all funds or instruments, and are also
not an exhaustive list, we recommend that these factors be provided as guidance in an adopting
release, rather than enumerated in the rule text.
With respect to the proposed liquidity classifications, we note that outside of US equities and a
limited range of government bonds, an assessment of days to convert a position to cash can be
highly subjective and imprecise, even without the added stipulation that the transaction must not
materially affect the value of the asset. Importantly, such an estimate is likely to be most inaccurate
when it is most scrutinized – that is, during unpredictable market events or volatility. Even in
normal market conditions, these assessments could change from week to week or even daily. We
are thus concerned that the proposed “days to liquidate” designations imply a false sense of

These factors are: existence of an active market for the asset, including whether the asset is listed on an exchange, as
well as the number, diversity, and quality of market participants; frequency of trades or quotes for the asset and average
daily trading volume of the asset (regardless of whether the asset is a security traded on an exchange); volatility of trading
prices for the asset; bid-ask spreads for the asset; whether the asset has a relatively standardized and simple structure; for
fixed income securities, maturity and date of issue; restrictions on trading of the asset and limitations on transfer of the
asset; the size of the fund’s position in the asset relative to the asset’s average daily trading volume and, as applicable, the
number of units of the asset outstanding; and relationship of the asset to another portfolio asset.
2

3

precision. Our concern is compounded by the designation of six categories – this level of
granularity further implies a level of precision which is unrealistic in most asset classes.
The proposed liquidity assessment is further complicated by the requirement that conversion to cash
must not materially affect the value of the asset. We recognize the concern underlying this
requirement, namely that if funds sell assets at “fire sale” prices there can be negative price pressure
on those assets as well as correlated assets, which could transmit stress to other funds or portions of
the market.3 In theory, almost any security can be sold quickly with a high enough discount; we
therefore acknowledge the Commission’s desire to impose some boundary on the price at which an
asset can be converted to cash in order to qualify as liquid.
However, the requirement that a conversion must not affect the value (or not “materially”)
discounts two very important facts. First, buying or selling virtually any position size by definition
affects pricing, with the possible exception of relatively small blocks of actively traded securities.
Moreover, even for the most liquid instruments, in dynamic markets prices change constantly. From
a practical perspective, it would be impossible to determine the potential effect of an individual trade
on an asset’s price in order to comply with such a requirement.
As important, while mutual funds promise daily redemptions, there has never been an expectation
that such redemptions must take place without impacting market prices. Investors understand that
sales of fund portfolio assets in a declining market could result in the fund receiving less than the
asset’s carrying value. Such investment risk is fundamental to the success of the capital markets and
to mutual funds themselves. Any suggestion that mutual funds offer some level of price certainty
should be treated with extreme caution.
To address these concerns while remaining consistent with the proposal’s approach of classifying
liquidity on a continuum, we recommend that funds be required to conduct position-level
assessments and classify assets into five buckets with descriptors that connote relative liquidity, while
removing the “days to liquidate” indicator. These buckets are described below. To provide
additional color, we have included examples of security types that we would generally consider for
each bucket if we were to conduct such an analysis today. We would not recommend that these
examples be codified in the final rule, however, because as we have explained, markets are dynamic
and conditions change frequently. As a result of the required ongoing monitoring, the illustrative
security types we provide could require different classification in response to changing market
conditions.

See, e.g., Open-End Fund Liquidity Risk Management Programs; Swing Pricing; re-Opening of Comment Period for
Investment Company Reporting Modernization Release, Release Nos. 33-9922, IC-31835 (Sept. 22, 205), 90 Fed. Reg.
62274 (Oct. 15, 2015) (“Proposing Release”) at 24.
3

4

1) Cash and cash equivalents: “daily liquid assets” as defined in Rule 2a-7, i.e., i) cash, ii) direct
obligations of the U.S. Government, or iii) securities that will mature or are subject to a
demand feature that is exercisable and payable within one business day;
2) Highly liquid instruments: instruments with daily trading/widely available market quotes
even during periods of market stress (e.g., large cap equities, G4 government and agency
bonds (US, UK, Germany and Japan), new issue US Agency MBS passthroughs, investment
grade corporate bonds that have large amounts outstanding and are constituents of major
benchmarks);
3) Moderately liquid instruments: instruments that are considered liquid in ordinary markets,
but which may become less liquid in stressed conditions (e.g., most high yield bonds, major
emerging market equities, investment grade corporate bonds other than those described
above, “Tier 1” EMD sovereign bonds, senior tranche RMBS and CMBS);
4) Less liquid instruments: instruments for which there are few active participants, execution is
sporadic and/or in small sizes (e.g., non-rated municipal securities, high yield bonds that are
particularly vulnerable to a negative credit event, frontier country EMD sovereign bonds,
junior tranches of structured products);4
5) Illiquid instruments: instruments that are private or restricted securities (other than 144A or
similarly liquid securities), internally fair valued, and/or are highly distressed.
We believe assigning portfolio positions to these five buckets, using the SEC’s proposed factors5 as
appropriate along with any other relevant information, is preferable to requiring an estimate of “days
to liquidate,” for several reasons. First, it addresses the Commission’s desire to see liquidity analyzed
on a spectrum, and its interest in receiving data it can use to monitor liquidity risk in the industry.
Although there may be some differences around the margins in how funds assign specific positions
to these buckets, this approach also is likely to produce more comparable results across funds
compared to the SEC’s proposed buckets, which require an element of guesswork (e.g., to determine
whether a position will take 7 or 8 days to liquidate).
In addition, our proposed approach negates the need for the requirement that a potential transaction
not materially impact the price. As noted above, using a metric of “days to liquidate” requires the
imposition of a backstop against selling quickly for pennies on the dollar (i.e., in theory almost any
security can be sold quickly with a high enough discount, which the proposal seeks to avoid).
Because our approach looks at the market characteristics of each position, there is no need for such
While there may not be standing bids or offers for some instruments in this category, traders execute the desired
transactions by contacting a number of potential counterparties.
4

5

See supra note 2.

5

a backstop on price; put starkly, a fund could not put a high yield bond in the “highly liquid” bucket
on the theory that it could be converted to cash quickly at a 50 percent discount, because the market
characteristics of such a bond do not warrant this classification.
Lastly, because our recommended approach is less subjective than an estimate of “days to liquidity,”
we believe it is more amenable to inputs from third-party service providers. This is particularly
important given that the proposal is based on a position-level liquidity analysis. We expect that over
time, funds’ approach to liquidity classification will look similar to valuation procedures, whereby we
receive a data feed with a variety of quantitative information from third-party service providers,
which we review and test (daily and retrospectively), and scrutinize any position that triggers a flag.
Without such an automated approach to narrow down the securities that require individual
attention, it would be challenging to assess liquidity at a position level on an ongoing basis.6 To
ensure that such an approach retains appropriate rigor and supervision, we support the
recommendation in the Proposing Release that funds and advisers should review the data quality,
methodologies and metrics provided by third parties, and consider the need for modifications to
accurately reflect the liquidity characteristics of a given fund’s holdings.7
In sum, we believe our proposed approach would, consistent with the Commission’s intent, enhance
a fund’s ability to monitor and manage its liquidity and plan how to meet redemptions, and would
provide sufficiently granular data to allow the Commission to monitor liquidity risk in funds and
identify outliers.
Management of Liquidity Risk – Three-Day Liquid Asset Minimum
JPMAM supports improving funds’ consideration of, and putting a governance structure around, the
maintenance of highly liquid assets in their portfolios. While we carefully monitor liquidity in our
funds and expect that many of our competitors do the same, we recognize that, with no current
regulatory requirements in this area, practices vary widely. Given the global concerns about market
liquidity8 and the growing participation of funds in the securities markets, we believe it is sensible to
examine how funds determine the appropriate level of highly liquid assets, and to place more

6

We estimate that our US funds hold a combined total of over 30,000 individual CUSIPs.

7

Proposing Release at 82.

See, e.g., Financial Stability Oversight Council 2015 Annual Report, available at
https://www.treasury.gov/initiatives/fsoc/studies-reports/Documents/2015%20FSOC%20Annual%20Report.pdf at 4
(“As this evolution of market structure plays out across a broader collection of asset classes and markets, market
participants and regulators should continue to monitor how it affects the provision of liquidity and market functioning,
including operational risks.”); Statement by Mark Carney, Chairman, Financial Stability Board, to the International
Monetary and Financial Committee, April 18, 2015, available at
https://www.imf.org/External/spring/2015/imfc/statement/eng/FSB.pdf (“Market participants need to be mindful of
risks of diminished market liquidity, asset price discontinuities, and contagion across markets.”).
8

6

rigorous governance and oversight around such considerations. However, we have several concerns
with the proposed approach to requiring a three-day liquid asset minimum.
As a preliminary matter, as discussed above, we have concerns about classifying the liquidity of
assets by the number of days in which they can be converted to cash without materially impacting
the price. We therefore recommend that the “three-day liquid assets” concept be replaced with
“highly liquid assets” or “HLA,” consisting of cash and highly liquid instruments, i.e., the first two
categories of our proposed liquidity classifications.
Perhaps more importantly, and consistent with our view that markets and liquidity are dynamic, we
believe that liquidity management likewise ought to be a dynamic process. Markets and outlooks
change daily, and portfolio managers have always taken current events into consideration when
managing liquidity. By contrast, the proposed requirement that an HLA minimum, and any changes
to it, be approved by the fund board will lead to the minimum being relatively static. JPMAM’s fund
board holds five scheduled meetings per year; even assuming that an HLA minimum was presented
at each meeting, we do not believe that this allows for sufficient flexibility.
There are two possible outcomes from such a static approach, neither of which benefits fund
investors. Some funds are likely to take a conservative approach and continually maintain an HLA
minimum that is appropriate for high-stress or worst-case market environments, even when not
demanded by current or foreseeable market conditions. This could have a negative impact on fund
returns, and could have the perverse effect of driving investors toward funds with a less conservative
approach, and correspondingly higher returns. Alternatively, some boards may agree to approve a
low HLA minimum, based on circumstances at the time of the approval; in such cases the fund
would presumably increase its liquid holdings as market conditions changed, but subject to limited
board oversight.
We are also concerned that an HLA minimum could unnecessarily hamper beneficial fund
investments. The proposal is clear that a fund that fell below its three-day liquid asset minimum
would not be forced to sell less liquid assets to return to its minimum, but would be prohibited from
buying additional less liquid assets. Nonetheless, there may be missed opportunities as a result of an
absolute minimum, for example if a fund has an impending inflow (e.g., a new mandate) and would
like to begin to invest in advance of the flows, or when the level of HLA has been reduced due to
appreciation in securities in an upward trending market and additional buying opportunities are
identified in less liquid assets. As noted above, these possibilities could lead fund boards to approve
artificially low minimums.
To allow for a more dynamic and flexible approach, while ensuring enhanced governance and
oversight of the establishment and maintenance of an appropriate level of HLA, we recommend
several changes to the Commission’s proposal. First, the board should be permitted to delegate its
authority to a liquidity management committee comprised of employees of the fund adviser

7

appointed by the board. Much like the current valuation process, the liquidity management
committee could, subject to board-approved policies and procedures, have responsibility for
establishing, monitoring and, as necessary, approving changes to the maintenance of HLA, based on
the factors proposed by the Commission and other relevant information.9 The committee could
report to the board on a regular basis.
Second, rather than an HLA minimum, we recommend an HLA threshold. The key distinction here is
that a fund would not be prohibited from transacting in less liquid assets when the threshold is
breached; however, such a breach would necessitate a timely review by the liquidity management
committee and a report to the board.
Finally, because a liquidity management committee would be better positioned than the board to
oversee HLA on an ongoing basis, we recommend that the liquidity management committee could
be charged with maintaining an HLA threshold that is designed to address current and reasonably
foreseeable near-term market conditions. This standard should improve results for investors by allowing
funds to be more fully invested during ordinary markets, while requiring them to be cognizant of
and prepared for changing conditions.
Separately, we request that the Commission clarify its intent regarding the use of HLA to meet
redemptions. We agree with the SEC that better oversight and governance around HLA may
improve liquidity management and reduce the likelihood that funds would need to dispose of less
liquid assets at “fire sale” prices to meet redemptions. However, we do not believe that the existence
of such assets should create an expectation or requirement that redemptions will be funded from
those assets. Indeed, such a requirement could actually create an incentive for investors to redeem
early, to be assured of getting out before the fund depleted its HLA; even absent such “runs,” a
requirement to meet redemptions from HLA would by definition increase liquidity risk for
remaining shareholders.
We are concerned that, based on the language in the proposal, a portfolio manager’s election to sell a
less liquid asset (and accept a lower price or “liquidity haircut”) on or about the same time that the
fund experiences redemptions could be viewed as violating the rule. We request that the
Commission clarify that, while the HLA requirement is designed to impose specific requirements on
and governance around the liquidity management process, the ultimate determination of how to
manage the portfolio and meet individual redemptions remains with the portfolio management
team.
The SEC’s proposed factors to consider in establishing a three-day liquid asset minimum are: the size, frequency, and
volatility of historical purchases and redemptions of fund shares during normal and stressed periods; the fund’s
redemption policies; the fund’s shareholder ownership concentration; the fund’s distribution channels; and the degree
of certainty associated with the fund’s short-term and long-term cash flow projections. As with the factors to be
considered for purposes of liquidity classifications, we believe the proposed list of factors is not exhaustive and that the
appropriate factors will vary by asset class and fund. We therefore recommend that these factors be provided as
guidance in an adopting release, rather than enumerated in the rule text.
9

8

Reporting and Disclosure Requirements
JPMAM supports the proposed requirements for increased reporting to the Commission on
position-level liquidity classifications and other elements of funds’ liquidity risk management
programs on Forms N-Port and N-CEN. As the primary regulator of the fund industry in the US,
the SEC should have the data it needs to monitor fund holdings and liquidity determinations,
examine potential outliers and, if an unexpected market event occurs (e.g., the default of a significant
institution), quickly assess the potential impact on mutual funds it supervises.
Indeed, we would support the sharing of such information across other financial regulators, subject
to thorough consideration of confidentiality and data security issues. In light of the concerns
regarding potential systemic risk in the asset management industry both in the US and abroad,10 we
believe the dialogue can only be improved with better data about fund holdings.
We are also supportive of the proposed additional disclosures to investors on Form N-1A regarding
the number of days in which a fund will pay redemption proceeds, and a description of the methods
they use, such as in-kind redemptions, and available funding sources, such as lines of credit, to fulfill
their redemption obligations. We agree with the SEC that this information will improve shareholder
and market participant knowledge regarding fund redemption procedures and liquidity risk
management. Similarly, for funds that elect to use swing pricing, the proposed disclosure will be
important to enhance public understanding regarding the benefits and risks of swing pricing.11 We
would further support the Commission requiring funds to include a discussion of their liquidity risk
management policies and procedures, similar to what is currently required on Form N-1A for
policies and procedures regarding proxy voting (items 17 and 27) and valuation procedures (item
23), among others.
On the other hand, we question the value of providing quarterly public disclosure of liquidity
classifications. If, as the Proposing Release suggests, third-party data analyzers used this information
to produce metrics for investors, which would presumably be updated quarterly in conjunction with
each public disclosure, such metrics could be nearly five months old when they are replaced (every
three months plus a 60-day lag). As such, they may no longer represent the fund’s current portfolio
composition and liquidity profile. While index funds and others that intentionally manage to their
See, e.g., Financial Stability Oversight Council, Notice Seeking Comment on Asset Management Products and Activities
(Dec. 24, 2014), available at
https://www.treasury.gov/initiatives/fsoc/rulemaking/Documents/Notice%20Seeking%20Comment%20on%20Asset
%20Management%20Products%20and%20Activities.pdf; Financial Stability Board, Assessment Methodologies for
Identifying Non-Bank Non-Insurer Global Systemically Important Financial Institutions (March 4, 2015), available at
http://www.fsb.org/wp-content/uploads/2nd-Con-Doc-on-NBNI-G-SIFI-methodologies.pdf.
10

As discussed below, JPMAM currently employs swing pricing on the vast majority of our Luxembourg-domiciled
UCITS funds, as well as our UK-domiciled Open End Investment Companies. Information on swing pricing can be
found in those funds’ prospectus at:
http://www.jpmorganassetmanagement.lu/EN/dms/JPMorgan%20Funds%20[PRO]%20[GB_EN].pdf (pp. 17-18).
11

9

benchmarks are likely to maintain relatively steady allocations to each asset class or sector in their
portfolio, a large and growing12 number of non-traditional and multi-asset funds employ flexible
investment approaches across sectors and even asset classes. Such a fund could have a substantially
different portfolio than is indicated by a published liquidity metric that is up to five months old.
Additionally, we are concerned that the proposed assessment of days to convert a position to cash is
subjective, which could lead very similar funds to disclose different liquidity profiles. Our proposed
approach to liquidity classifications will reduce the variability across funds but, particularly for the
first year or two, there are likely to be discrepancies in how funds report their liquidity. While the
SEC will have the underlying data to identify and calibrate these differences, we are concerned that
third party ratings would simply take the funds’ own assessments. This could lead to comparisons
that disadvantage funds with a more conservative approach to classification, and provide false
assurances to investors regarding funds that are more aggressive in their approach to classifications.13
With respect to public disclosure of a fund’s HLA threshold, we have two concerns. First, as
discussed above, we believe liquidity management should be a dynamic and flexible process, and that
a fund’s HLA threshold may change based on reasonably foreseeable near term market conditions.
Thus, a publicly disclosed HLA threshold may be a stale snapshot in time, and potentially not
reflective of the fund’s actual HLA or portfolio management approach at the time it is being relied
upon by an investor. Perhaps more importantly, the public disclosure of such a number could
actually incentivize investors to redeem quickly at the first indication of potential market stress.
Although the HLA threshold is not necessarily a reflection of the actual portfolio, its disclosure may
create the perception that a fund has only the enumerated amount of liquidity available to meet
redemptions, which could drive investors to redeem before this supposed liquidity runs out.
In sum, we support monthly reporting of liquidity classifications and HLA thresholds to the SEC,
which could potentially be shared with other regulators. We also support the proposed disclosure to
investors on Form N-1A, and would further support requiring a discussion of liquidity management
policies and procedures. However, because we think position-level liquidity classifications are not
likely to be useful to investors and could in fact be an inaccurate representation of a fund’s current

Since 2010, AUM in non-traditional bond funds have increased approximately 40% (from $55B in Dec. 2010 to $139B
in Nov. 2015); assets in multi-sector bond funds have increased approximately 50% (from $90B in Dec. 2010 to $166B
in Nov. 2015); and AUM in multi-asset funds have increased over 60% (from $644B in Dec. 2010 to $997B as of Nov.
2015. These funds typically have flexible investment mandates, and their holdings and liquidity profiles could change
substantially over the course of several months.
12

It should be noted that, even without disclosing funds’ liquidity classifications on Form N-Port, third-party data
analyzers can and do develop liquidity ratings by imposing their own liquidity metrics on funds’ disclosed portfolio
holdings. See, e.g., “Press Release: Interactive Data Set to Launch Liquidity Indicators Service,” May 18, 2015, available at
https://www.interactivedata.com/Assets/DevIDSite/PR-2015/Interactive-Data-Set-to-Launch-Liquidity-IndicatorsService.pdf. These metrics are likely to be more comparable, since the service providers are applying consistent liquidity
classifications.
13

10

portfolio holdings, and because disclosure of an HLA threshold could potentially incentivize first
movers, we recommend that the Commission not require such disclosure to the public.
Swing Pricing
JPMAM applauds the Commission’s willingness to permit swing pricing for open-end funds. We
employ swing pricing on the vast majority of our Luxembourg-domiciled UCITS funds, as well as
our UK-domiciled Open End Investment Companies.14 We believe that, by reducing dilution of the
fund as a result of transaction costs, swing pricing is likely to benefit our long-term shareholders.15
As a general matter, we support the SEC’s proposed approach to swing pricing, including that its use
would be voluntary, governed by board-approved policies and procedures, and subject to disclosure
and reporting requirements on Forms N-1A and N-CEN. We recommend that the Commission
permit, but not require, the consideration of market impact in setting a swing factor, and we suggest
more flexibility in the factors for establishing a swing threshold. We also describe several
operational obstacles to implementation of swing pricing at this time. Nonetheless, we believe that
if the SEC permits the use of swing pricing, the industry will work with its partners in the fund
distribution chain to find solutions to the operational concerns. We therefore support the proposal,
subject to the following comments.
First, we do not support the proposed requirement that a fund’s swing factor take into account
“market impact costs.” We agree that including market impact costs in a fund’s swing factor would
make swing pricing an even better tool for addressing liquidity risks, because it would charge
transacting shareholders for changes to the market value of securities bought or sold as a result of
their transactions. However, based on our experience with swing pricing in Luxembourg and the
UK, we do not believe there currently exists a conventional or commonly used method for
incorporating market impact costs into a swing factor. We believe this would take substantial
additional consideration, data, modeling and testing.16

We would be pleased to meet with the Commission or its staff to discuss our experience with swing pricing in more
detail.
14

As part of our annual swing threshold review for these funds, we quantify the dilution saved, in dollars and as a
percentage of AUM.
15

The Proposing Release concedes that “[m]arket impact cost cannot be calculated directly. It can be roughly estimated
by comparing the actual price at which a trade was executed to prices that were present in the market at or near the time
of the trade.” Proposing Release at 187-88, n. 415. Absent some ability to model and predict such costs, we are
uncertain as to whether they could be included in a predetermined and periodically reviewed swing factor, or
alternatively whether a fund’s swing factor would have to be determined on a daily basis, which may not be feasible.
16

11

Meanwhile, swing pricing as currently employed, which takes into account such costs as transaction
costs, spreads, and taxes such as stamp duties,17 can still benefit long-term shareholders by reducing
dilution, and can act as a disincentive for investors to redeem in times of stress by imposing a charge
when the swing threshold is breached. We therefore recommend that the SEC permit swing pricing
as currently used, while making the consideration of market impact costs optional for purposes of
establishing a fund’s swing factor.18 Over time, we believe the industry would explore ways to
incorporate market impact. Funds that employ swing pricing could also be required to describe in
their prospectus the costs that are considered in establishing a swing factor, so that investors are
aware of their funds’ practices.
Secondly, we do not believe a fund’s investment strategy, the liquidity of portfolio assets, or the
costs associated with transactions should be required considerations for determining a swing
threshold. These criteria are important in considering a swing factor, which may be set very low for a
fund that typically has high liquidity and low transaction costs. To set a threshold, on the other
hand, a fund may wish to look primarily at its flows, and determine the amount of flows it wishes to
capture with swing pricing to reduce dilution. Even minimal dilution can be impactful over time,
and a fund may elect to implement swing pricing, with a low swing factor, to address such dilution.
Additionally, fund complexes may prefer to implement swing pricing at the same threshold on the
full range of funds, even in cases where the swing factor is low (because transaction costs are
minimal), rather than establish different thresholds for each fund. Thus, we recommend that these
factors be optional considerations for the determination of a swing threshold.
Finally, as an operational matter with respect to implementation of swing pricing in the US, we are
not confident in our ability to determine a fund’s net flows in a timely fashion and with sufficient
certainty to warrant adjusting the NAV. In the US, we are required to accept shareholder orders
until the fund is closed for valuation, i.e., 4 pm. NAVs must be submitted for publication by 6 pm.
Meanwhile, trade flows for funds that are sold via third party distributors are generally not available
until the early morning of T+1.19 By contrast, in Luxembourg, we stop accepting orders at 2:30 pm,

See “Swing pricing: The J.P. Morgan Asset Management approach in the Luxemburg domiciled SICAVs JPMorgan
Funds and JPMorgan Investment Funds,” April 2015, available at
http://www.jpmorganassetmanagement.de/DE/dms/Swing%20Pricing%20[MKR]%20[IP_EN].pdf.
17

This would be consistent with the recently released Swing Pricing Guidelines from the Association of the Luxembourg
Fund Industry, which includes market impact as an item that could be considered when deriving the swing factor. See
Association of the Luxembourg Fund Industry, Swing Pricing Guidelines, December 2015, available at
http://www.alfi.lu/sites/alfi.lu/files/Swing-Pricing-guidelines-final.pdf, at 12.
18

Indeed, some of those flows, particularly those for 401(k) plans, cannot be calculated until after the NAV is submitted.
Since 401(k) flows tend to be steady, however, it is possible that they could be estimated with reasonable certainty.
19

12

Central European Time, and receive cash projections at 5:30 pm. NAVs are released at 7:15 pm.
This timeline is sufficient to analyze cash flows and determine whether to swing the NAV.20
We have begun conversations with our distribution partners about the feasibility of obtaining cash
flow protections on T, but these discussions are in the early stages. Moreover, we believe such talks
are more likely to bear fruit if they are part of an industry call to improve cash flow projections,
rather than bespoke arrangements between various fund complexes and distributors. Another
potential improvement would be to delay the NAV publication time from 6 pm to 8 pm. We
support the industry efforts to resolve these operational issues,21 and we encourage the SEC to
consider how it could engage in such efforts, such as by participating in industry roundtables.
*

*

*

JPMAM appreciates the opportunity to comment on the Commission’s proposed rule. We would
be pleased to provide any further information or respond to any questions that the Commission or
the staff may have.

Very truly yours,
/s/ George C.W. Gatch
George C.W. Gatch
Cc:

The Honorable Mary Jo White, Chair
The Honorable Kara M. Stein, Commissioner
The Honorable Michael S. Piwowar, Commissioner
David W. Grim, Director, Division of Investment Management

We back-tested our cash flow projections versus revised cash flows over three years, and found very high accuracy in
the swing pricing determinations (i.e., sufficient to satisfy the SEC’s proposed “determination on the basis of
information obtained after reasonable inquiry”).
20

We concur with the operational challenges described by the Global Association of Risk Professionals in its comment
letter on this proposal. See Letter from Richard Apostolik, President and CEO, Global Association of Risk Professionals,
to Brent Fields, Secretary, Securities and Exchange Commission, dated Jan. 12, 2016, available at
https://www.sec.gov/comments/s7-16-15/s71615-33.pdf.
21

13