SecProbe.io

CORRESP
1
filename1.htm

February
06, 2024

Christopher
Bellacicco

Division
of Investment Management

U.S.
Securities and Exchange Commission

100
F Street, N.E.

Washington,
D.C. 20549

    Re:
    GraniteShares;
    File Nos. 333-214796, 811-23214

Dear
Mr. Bellacicco:

On
November 22, 2023, GraniteShares ETF Trust (the “Trust” or “Registrant”) filed a Registration Statement under
the Securities Act of 1933 and the Investment Company Act of 1940 (the “1940 Act”) on Form N-1A on behalf of GraniteShares
1.5x Long NVDA Daily ETF, GraniteShares 1.5x Long COIN Daily ETF, GraniteShares 1.75x Long BABA Daily ETF, GraniteShares 1.5x Long META
Daily ETF, GraniteShares 1.75x Long GOOGL Daily ETF, GraniteShares 1.5x Long AMZN Daily ETF, GraniteShares 1.75x Long AAPL Daily ETF,
GraniteShares 1.25x Long AMD Daily ETF, GraniteShares 1.25x Long PLTR Daily ETF, GraniteShares 1.5x Long UBER Daily ETF, GraniteShares
1.5x Long DIS Daily ETF, GraniteShares 1.25x Long F Daily ETF, GraniteShares 1.5x Long NVDA Daily ETF, GraniteShares 1.5x Long COIN Daily
ETF, GraniteShares 1.75x Long BABA Daily ETF, GraniteShares 1.5x Long META Daily ETF, GraniteShares 1.75x Long GOOGL Daily ETF, GraniteShares
1.5x Long AAL Daily ETF, GraniteShares 1.5x Short AAL Daily ETF, GraniteShares 1.75x Short AAPL Daily ETF, GraniteShares 1.25x Short
AMD Daily ETF, GraniteShares 1.5x Short COIN Daily ETF, GraniteShares 1.5x Long JPM Daily ETF, GraniteShares 1.5x Short JPM Daily ETF,
GraniteShares 1.5x Long LCID Daily ETF, GraniteShares 1.5x Short LCID Daily ETF, GraniteShares 1.5x Short META Daily ETF, GraniteShares
1.5x Short NVDA Daily ETF, GraniteShares 1.5x Long RIVN Daily ETF, GraniteShares 1.5x Short RIVN Daily ETF, GraniteShares 1.75x Long
TSLA Daily ETF, GraniteShares 1.5x Short TSLA Daily ETF, GraniteShares 1.5x Long XOM Daily ETF, GraniteShares 1.5x Short XOM Daily ETF,
GraniteShares 1.5x Long AAL Daily ETF, GraniteShares 1.5x Short AAL Daily ETF, GraniteShares 1.75x Short AAPL Daily ETF, GraniteShares
1.25x Short AMD Daily ETF, GraniteShares 1.5x Short COIN Daily ETF, GraniteShares 1.5x Long JPM Daily ETF, GraniteShares 1.5x Short JPM
Daily ETF, GraniteShares 1.5x Long LCID Daily ETF, GraniteShares 1.5x Short LCID Daily ETF, GraniteShares 1.5x Short META Daily ETF,
GraniteShares 1.5x Short NVDA Daily ETF, GraniteShares 1.5x Long RIVN Daily ETF, GraniteShares 1.5x Short RIVN Daily ETF, GraniteShares
1.75x Long TSLA Daily ETF, GraniteShares 1.5x Short TSLA Daily ETF, GraniteShares 1.5x Long XOM Daily ETF, and GraniteShares 1.5x Short
XOM Daily ETF (the “Funds”).

The
Trust is providing the below information in response to the comments given by you via telephone to Matthew Tobin on January 30, 2024.
The comments are summarized below, with corresponding responses, which the Registrant has authorized Thompson Hine LLP to make on its
behalf. Capitalized terms used but not otherwise defined herein have the meanings ascribed to them in the document to which the applicable
comment relates.

    1

Comment
1. According to their methodologies, the indexes that GraniteShares plans to use are designed to capture the performance of the top
5 most volatile stocks in the relevant industry. Please explain how the most volatile subset of an industry, and a subset which appears
inconsistent with the funds strategies and into which the underlying stock may never fall, is an appropriate reference portfolio for
the markets or asset classes in which the GraniteShares ETF may invest. In providing your response, please address how the use of these
volatility-based indexes is consistent with the guidance provided on Page 96 of the Rule 18 (f) 4 adopting release which states that
“A funds designated reference portfolio is designed to create a baseline VAR that functions as the VAR of a funds unleveraged portfolio.”

Response:

The
fund’s derivative risk manager (FDRM) believes the statement made by the SEC Staff (the Staff) that “the indexes that GraniteShares
plans to use are designed to capture the performance of the top 5 most volatile stocks in the relevant industry” (emphasis
added) is inaccurate. The index methodology incorporates screens (size, liquidity, trading history and trading location) that are designed
to remove stocks that might skew the overall index volatility higher. In addition, because benchmarks may only contain publicly listed
stocks above a certain size, the indexes are unable to include smaller, younger and/or privately held companies that are arguably riskier
and, as a result, more volatile.

Regarding
the Staff’s statement “(…) a subset (…) into which the underlying stock may never fall, is an appropriate reference
portfolio for the markets or asset classes in which the GraniteShares ETF may invest.”, the FDRM notes that each Underlying Stock
is an eligible constituent in their respective selected benchmark as they share the same FactSet industry classification. For instance,
to be eligible in the Indxx High Volatility Internet Retail Industry Index, a constituent must be classified in the Internet Retail industry
as defined by FactSet. Since Amazon.com is classified by Factset as a stock in the Internet Retail Industry then it is eligible in the
Indxx High Volatility Internet Retail Industry Index. The FDRM acknowledges that although an Underlying Stock is eligible for inclusion
it might never be a constituent of the selected benchmark (what the comment seems to refer to as “fall” in the benchmark).
But the FDRM also notes that just because an Underlying Stock is currently a constituent of a benchmark does not mean it will remain
a constituent going forward. Such modification can occur for various reasons including a change in the benchmark industry taxonomy or
a change in the Underlying Stock’s core business (see Appendix 2). The statement also assumes that the Underlying Stocks are truly
representative of their core industry. The FDRM believes that the assumption is not necessarily accurate and, more importantly, may lead
to biased results for the purpose of Rule 18(f)4 (the Rule). The FDRM observes that most Underlying Stocks operate businesses
diversified across industries instead of being pure industry players. Most Underlying Stocks are also highly successful and established
businesses. Hence it can be argued that for the purpose of the Rule a benchmark including an Underlying Stock, especially if it is granted
a material weight, might materially misrepresent the volatility of the targeted industry.

In
addition, by questioning “(…) how (...) a subset which appears inconsistent with the funds strategies (…), is an
appropriate reference portfolio for the markets or asset classes in which the GraniteShares ETF may invest”, the Staff seems to
imply that while the selected benchmarks might indeed focus on the industries in which the Underlying Stocks operate, a volatility weighted
methodology with 5 components leads to an inaccurate representation of an industry. This, in turn, seems to imply that the Staff
believes that one index methodology over another leads to a better industry representation

The
FDRM observes that the Rule does not specify whether the designated benchmark should use a specific methodology (as to its weighting
or its number of constituents1) or a specific industry taxonomy. There exists today a potentially infinite number of methodologies
that can be used to represent a given industry (see Appendix 1). Designating an industry itself is challenging especially when the impact
of time and evolution of the industries themselves are considered (see Appendix 2). The FDRM is not aware of a market consensus advocating
for the superiority of any benchmark methodology or industry taxonomy but instead observes that decisions are entirely contextual. Given
each decision has its pros and cons, it appears unlikely that there can be a definitive a priori determination as to what should constitute
the most appropriate benchmark to represent an industry.

1 The FDRM acknowledges that the Staff opposed the choice
made by the FDRM in its initial filing dated February 2022 to use single stock indices as selected benchmarks although to the FDRM’s
knowledge the Staff did not make s specific reference to the Rule to support its opposition. The FDRM notes that sponsors have used indices
with less than 5 constituents as selected benchmarks.

    2

Beyond
the theoretical concept regarding which methodology or taxonomy should be used, the Rule imposes practical procedural restrictions on
the FDRM. In particular, the Rule prevents the FDRM from contacting index providers to construct an index for the purpose of complying
with the Rule. Although the FDRM does not deny the rationale of such restriction, its corollary is that the FDRM must select a benchmark
from a set of available solutions where each solution may be questionable. Further, because the set of available solutions may be very
large2, the FDRM believes it is impractical to consider every benchmark that provides sector exposure to the Underlying Stock,
analyze each index methodology and asses the availability of the benchmark data.

Hence,
both from a theoretical and a practical perspective, the selection of a benchmark for the purpose of the Rule can only be assessed on
whether the choice is reasonable or superior to other accepted solutions. As shown in Appendix 3 the choice made by the FDRM meets such
requirements and the FDRM believes that the selected benchmarks address the Staff concerns and for that reason should be considered as
appropriate reference portfolios for the markets or asset classes in which the funds may invest.

Comment
2. The Staff notes that back tested VAR data was provided for a three-month period. Please provide at least six-month of back tested
VAR data.

Response:

Back
tested data is available in Appendix 4.

*     *    *    *    *

If
you have any questions or additional comments, please call the undersigned at 216-566-5706.

    Very
    truly yours,

    /s/
    Andrew Davalla

    Andrew
    Davalla

    cc:
    JoAnn
    M. Strasser

2 On November 16, 2018, the Index Industry Association estimated
that as of June 30, 2018, there were 3.4 million benchmarks and indices in use globally, an increase of 438,000 over the previous 12-month
period.

    3

Appendix
1 – Benchmark Scope and Weighting Methodologies

Once
the benchmark’s industry has been selected the benchmark scope and its weighting methodology must be chosen.

The
scope consists in determining eligibility criteria to be included in a given sector or industry. It is generally unlikely that a benchmark
includes all companies of that sector. Instead, eligibility screens are implemented such as minimum thresholds (turnover, market capitalization,
percentage of revenues derived from the industry) or availability of the data. Some sector benchmarks determine their scope within another
benchmark. For instance, and as detailed in Appendix 3, the S&P 500 Technology Hardware, Storage and Peripherals Index selects stocks
that are members of the S&P 500 and that belong to the targeted GICS. As result the index has 6 components.

There
are three main weighting methodologies:

    (A)
    Market
    capitalization weighted indices

    (B)
    Market
    performance weighted indices (for instance volatility, momentum, liquidity…)

    (C)
    Fundamental
    weighted indices (for instance revenue, dividend rates, earnings, book value… or purely research based)

The
FDRM observes that option (A) naturally tilts its exposure towards the largest companies that operate within the selected industry. Hence
even if the scope of the benchmark was to include all companies operating within the selected industry, there is no mechanism within
a market capitalization weighted index to prevent a company from dominating the benchmark (see for instance Apple Inc in the S&P
500 Technology Hardware, Storage and Peripherals Index as shown in Appendix 3).

Alternative
methodologies have emerged over the years within option (A) to limit this known issue, such as indices using adjusted market capitalization
approaches, for instance, capping the weights of the largest companies or forcing a minimum weight on the smallest companies in the benchmark.

Options
(B) and (C) are more recent categories and are a natural extension of the solutions that aim to reduce the impact of the industry’s
dominant players on a benchmark’s performance.

Overall,
the FDRM notes that any weighting methodology is the result of a choice that entails tradeoffs. Selecting market capitalization weighted
indices related to the Underlying Stocks can lead to questions as to their ability to represent the targeted industry. Volatility weighted
indices have existed since the 1990s, are widely used in the asset management industry3 and should be considered as a reasonable
alternative to address some of the limitations of using market capitalization indices for the purpose of rule 18(f)(4).

Similarly,
the scope of a benchmark cannot be considered individually but instead within the broader scope of the index. Appendix 3, for instance,
shows that an index with 5 components can arguably provide better diversification and sector exposure than a larger index.

3 For instance: S&P 500 Volatility-Highest Quintile Index,
S&P 500 High Beta Index series, iSTOXX high beta index series, Scientific Beta Volatility Index Series

    4

Appendix
2 – Taxonomy and Industry Representativeness

Defining
the industry to which a company belongs at a given point of time can be challenging. This task is even more complicated as both companies
and industries continuously evolve.

Taxonomies
are not homogenous.

Taxonomies
use different terminologies which can lead to confusion when comparing them.

For
instance, the following terminologies are used (from the largest group to the smallest):

    -
    Global
    Industry Classification Standard (GICS): Sector, Industry Group, Industry and Sub-industry

    -
    Industry
    Classification Benchmark (ICB): Industry, SuperSector, Sector, Sub-Sector

The
FDRM also observes that even taken at the same group level the allocation is not homogenous across providers. For instance,

    -
    GICS
    classifies Alphabet Inc in the Communication Sector.

    -
    ICB
    classifies Alphabet Inc in the Technology Industry.

Taxonomies
evolve continuously.

Taxonomies
themselves and allocation across sectors and sub-sectors are also not static as they aim to reflect the continuous changes in the overall
economy. For instance, in March 2023, GICS created 2 new sectors and moved 14 of the S&P 500 constituents into those new sectors.

A
company does not necessarily reflect the industry in which it is allocated.

To
the FDRM’s knowledge, a company can only be allocated to a single industry or sub-sector within a given taxonomy. For instance,
according to Fortune as of 2021 JP Morgan is one of the largest broker dealers in the United States. However, as of January 29, 2024,
JP Morgan does not belong to the S&P Capital Markets Select Industry, but instead belongs to the S&P Banks Select Industry
Index. According to Bloomberg, JP Morgan derives only 53% of its revenues from its banking business4.

As
a result, one company can potentially dominate a sector benchmark even though it only derives a marginal majority of its revenues from
that sector. Based on that observation it could be reasonable to question whether such a company is representative of the sector under
which it is classified. And by extension it could be questioned whether a benchmark allocating a significant portion of its weight to
such a company should be considered as providing an accurate representation of the industry in which the company operates.

Overall,
the FDRM notices that the determination of the “relevant industry” in which the Underlying Stock operates is solely dependent
on the taxonomy used by the index providers when designing its benchmark methodolog