One International Place, 40th Floor 100 Oliver Street Boston, MA 02110-2605 +1 617 728 7100 Main +1 617 426 6567 Fax www.dechert.com
| ||||
STEPHANIE CAPISTRON
stephanie.capistron@dechert.com +1 617 728 7127 Direct +1 617 275 8364 Fax |
May 3, 2018
VIA EDGAR CORRESPONDENCE
Mr. Jay Williamson
U.S. Securities and Exchange Commission
Division of Investment Management
100 F Street, NE
Washington, D.C. 20549
Re: | Goldman Sachs ETF Trust (the “Registrant”), SEC File Nos.333-200933 and811-23013 |
Dear Mr. Williamson:
This letter responds to the comments you provided to (i) Lori Zedeck of Dechert LLP and me in a telephonic discussion on January 16, 2018 with respect to your review of Post-Effective Amendment No. 92 (the “Amendment”) to the Registrant’s registration statement filed with the U.S. Securities and Exchange Commission (“SEC”) on December 11, 2017 and (ii) to Lori Zedeck on March 22, 2018 with respect to your review of the Amendment and a correspondence filing dated March 19, 2018 that responded to certain of the comments provided on January 16, 2018 (the “Correspondence”). The Amendment was filed pursuant to Rule 485(a) under the Securities Act of 1933 for the purpose of registering shares of Goldman Sachs JUST U.S. Large Cap Equity ETF, a new series of the Registrant (the “Fund”). We have reproduced your comments below, followed by the Registrant’s responses. Capitalized terms have the meanings attributed to such terms in the registration statement.
General
Comment 1. Please provide a copy of the Index methodology white paper and a model portfolio as of a recent date. The model portfolio should include: name of security, weight of security in the Index, market capitalization, and (for the ten largest holdings only), a brief explanation of why the security is included in the Index.
Response 1. The Correspondence contained a copy of the Index methodology and an accompanying model portfolio.
Mr. Jay Williamson May 3, 2018 Page 2 |
Prospectus
Comment 2. The reference to “U.S. large cap equity” in the Fund’s name implicates the provisions of Rule35d-1 under the Investment Company Act of 1940, as amended (the “1940 Act”) and creates a separate obligation for the Fund to invest 80% of its net assets plus borrowings for investment purposes in U.S. large cap companies. If the existing 80% policy is designed to meet these requirements, please confirm that the existing policy is sufficient to satisfy the names rule requirement.
Response 2. The Registrant believes that the existing 80% policy is sufficient to satisfy the names rule requirement.
Comment 3. In the“Summary–Principal Investment Strategies”and “Investment Management Approach–Principal Investment Strategies” sections, the disclosure indicates that the Index is “designed to deliver exposure to equity securities of large capitalization U.S. issuers that engage in “just business behavior” based on rankings produced by the [Index Provider] . . . based on their quantitative performance assessment of seven issue areas . . ..” The SEC staff (“Staff”) notes that many of the issue areas, such as “how they treat the communities they interact with,” appear to require extensive subjective judgment. Please supplementally explain the Index Provider’s process for assigning quantitative measures across each of the seven issue areas. The discussion should address both quantitative and qualitative factors, as well as who is responsible for considering them and assigning the weightings. In addition to the foregoing, to the extent that Index constituent determination involves extensive subjective judgment, please explain why the Fund believes that the Index is sufficiently rules-based to be labeled as an index and why the Index Provider’s activities do not constitute a securities purchase recommendation.
Response 3. The Registrant provided a response to this comment in the Correspondence.
Comment 4. In the“Index Provider” section, the disclosure states that the Index furthers the Index Provider’s purposes of “encouraging greater accountability among business corporations with respect to the issue areas addressed in the Reference Index by providing a practical application of the issues on which the Index Provider seeks to educate the public and supporting development of a platform for investors to demand greater social responsibility from businesses. . .” Please include a clear, concise and understandable discussion of the Index Provider’s purpose in the“Summary–Principal Investment Strategies”
Mr. Jay Williamson May 3, 2018 Page 3 |
and“Investment Management Approach–Principal Investment Strategies” sections, and explain how the Index methodology is designed to achieve that purpose. In addition, to the extent that the purpose may come at the expense of shareholder returns, please ensure that the Fund’s risk factors appropriately reflect the trade-offs that investors may be making.
Response 4. The Registrant provided a response to this comment in the Correspondence.
Comment 5. The Staff notes that the Fund provides risk disclosures responsive to Item 4 and Item 9 of FormN-1A in the“Summary–Principal Risks of the Fund” section and“Risks of the Fund” section. Please explain why another set of risk factors is necessary in the“Appendix A–Other Portfolio Risks” section, or consider relocating the disclosure to the SAI.
Response 5. The Registrant respectfully submits that the “Appendix A–Other Portfolio Risks” section is intended to provide investors with more information about the general risks of the Fund, as well as additional broad background regarding portfolio securities techniques.
SAI
Comment 6. In the“Investment Restrictions” section, the Fund states that, in applying its concentration policy to derivative instruments, it “will look to the industry of the reference asset(s) and not to the counterparty or issuer.” As the counterparty is responsible under contract, it is unclear why the application of the fundamental restriction does not involve both the reference asset and counterparty. Please revise or explain.
Response 6. The Registrant utilizes derivative instruments to gain exposure to the reference asset(s) underlying the derivative, rather than to the counterparty or issuer of the derivative. In general, the value of such instruments changes in large part based on changes in the value of the reference asset(s). Accordingly, the Registrant believes that the industry of the reference asset(s), rather than that of the counterparty or issuer, should be considered for concentration purposes. In addition, the Registrant notes that counterparties in derivative transactions are generally securities-related issuers (“SRIs”) under Section 12(d)(3) of the 1940 Act. Section 12(d)(3) and Rule12d3-1 thereunder appropriately and adequately address the risks associated with counterparty exposure to SRIs. Accordingly, counterparty exposure to SRIs need not be addressed by reference to the 1940 Act’s portfolio concentration requirements. The Registrant notes that GSAM and/or its affiliates currently review counterparties to assess their creditworthiness. For these reasons, the Registrant respectfully declines to revise the disclosure.
Mr. Jay Williamson May 3, 2018 Page 4 |
Comment 7. In the “Proxy Voting” section, given the Fund’s focus on social impact investing, please advise whether the Fund will follow the same proxy voting policies as other funds in the Goldman Sachs Fund Complex. To the extent that the policies will be unique to the Fund, please explain the material differences in policies, including a discussion regarding how the Fund’s policies were developed, and what parties (including any third parties) participated in developing them.
Response 7. The Registrant hereby confirms that the Fund will follow the same proxy voting policies as other funds in the Goldman Sachs Fund Complex.
Correspondence
Comment 8. The Staff notes that the Registrant has not responded to several comments on the Amendment. Please provide responses to those comments prior to effectiveness.
Response 8. The Registrant notes that it has provided responses to all remaining comments on the Amendment.
Comment 9. The Staff notes the Registrant’s response to Comment 2 of the Correspondence. Please supplementally provide the credible objective sources used as inputs into the ranking methodology. In addition, please provide enhanced disclosure in response to Item 9, as it relates to the quantitative performance assessments used and overall Index methodology. In this respect, the Staff believes the executive summary of the Index methodology available on JUST Capital’s website can be condensed and should be described to investors. The Staff further believes additional information about the drivers and weights contained in Section 5 of the ranking methodology should be provided. Please revise the disclosure as necessary.
Response 9. The Registrant notes that JUST Capital is a third-party index provider and neither the Fund nor GSAM is involved in the Index construction process. The Registrant understands that JUST Capital collaborated with numerous third parties to generate information for the ranking methodology. Discussion of sources used for development of the “Drivers” of just business behavior is included in the attached Exhibit A.
Mr. Jay Williamson May 3, 2018 Page 5 |
JUST Capital has established “Metrics” for each “Component” (or items determined by the American public to be most important to “justness”) described in Exhibit B. The Metrics have been developed to meet the following criteria: (1) are judged to best reflect the public’s definition of the Component in question; (2) accurately measure company performance or managerial commitment; (3) require as few assumptions and as little subjective interpretation as possible; (4) can be assessed with clear units of measurement, or binary outcomes; (5) reflect measurement best practices; and (6) are broadly applicable to all companies in the universe, regardless of size, industry or business model. Where it is not possible to design Metrics that reasonably encompass performance for companies across all industries, JUST Capital designs additional metrics to ensure a subset of Metrics within each Component is relevant to each company.
JUST Capital employs Metrics that measure actual performance (e.g. megawatt hours of electricity used) and assess a company’s management practices (e.g. quality and comprehensiveness of a company’s environmental management system) on related issues.
JUST Capital analysts have reviewed, screened and consolidated over 5,000 individual data points from reputable third parties across all aspects of just corporate performance. These include: publicly available company reports (e.g., audited company filings, annual reports, corporate social responsibility and sustainability reports, etc.); third-party data vendors (i.e.,for-profit companies that collect and distribute data – both financial andnon-financial – including those focused on environmental, social and governance issues); U.S. government agency data (e.g., Bureau of Labor Statistics); academic andnon-for-profit data; national media and press; and crowd-sourced data. From this review and in response to stakeholder feedback, JUST Capital has selected, developed and classified its Metrics according to the following taxonomy:
Metric-type | Description | |
Performance | Measures and assessments of actual company performance (e.g. Quality of a company’s 401k plan on a0-100 point scale; Percentage change in US workforce; Electricity use in MWh/USD million of revenue; CEO or company director involvement in material related-party transactions, reported as True or False). | |
Management | Companies’ policies, commitments and management practices, which meet an established minimum standard or best practice. Typically measured on a binary Yes/No or True/False basis. |
Mr. Jay Williamson May 3, 2018 Page 6 |
Controversies | Controversies reported by influential media, stakeholder, and third-party sources captured according to a strict, rules-based research process. Measured by the number and severity of cases. | |
Crowd-sourced | Reviews and salary disclosures from current and former employees of ranked companies. | |
Fines | Fines levied by relevant regulatory authorities. Measured in US Dollars. |
The disclosure has been revised to include additional information about the drivers and weights and is included in the attached Exhibit B.
Comment 10. The Staff notes the Registrant’s response to Comment 2 of the Correspondence that JUST Capital selects data for its ranking model taking into account a number of key criteria and following strict protocol. Please explain how the Index Provider determines the data is complete (i.e., the data it chooses not to include) and who is responsible for analyzing and interpreting the data collected. If possible, please provide some examples to the Staff so that we may understand the Registrant’s response.
Response 10. As noted above, JUST Capital is a third-party index provider and neither the Fund nor GSAM is involved in the Index construction process. The Registrant understands that JUST Capital evaluates a wide range of reputable and widely-utilized data providers focusing on environmental, social and governance issues, includingfor-profit andnon-profit data vendors, and government data. Where available, JUST Capital reviews data dictionaries and evaluates them using a framework that considered applicability (e.g.,how directly a data point relates to the component in question), robustness (e.g.,the transparency of methodology and reputation of the organization responsible for its collection), and completeness (e.g.,the extent to which data was available for all companies within JUST Capital’s coverage universe).
The Registrant notes that additional information is included in the attached Exhibit A.
Mr. Jay Williamson May 3, 2018 Page 7 |
Comment 11. The Staff notes the Registrant’s response to Comment 3 of the Correspondence, with respect to considering the degree to which the Index Provider does not believe that its purpose is achieved at the expense of investor returns. While the Staff does not have a view on this, the Index methodology (which is based off of polling) contained in Section 5 of the ranking methodology puts investor returns at a lower weight than other factors, such as the way companies treat their workers. The Staff believes this is relevant to an investor’s understanding of how the Index works. Please consider additional Index disclosure, as necessary.
Response 11. The strategy disclosure and risk disclosure have been revised accordingly.
* * *
We believe that the foregoing has been responsive to the Staff’s comments. Please call the undersigned at (617)728-7127 if you wish to discuss this correspondence further.
Sincerely,
/s/ Stephanie A. Capistron
Stephanie A. Capistron
cc: | Joseph McClain, Goldman Sachs & Co. LLC |
Mr. Jay Williamson May 3, 2018 Page 8 |
Exhibit A
Drivers Development 2015-2018
Natalie Jackson, Managing Director, Polling
The top level of JUST Capital’s ranking model is the set of “Drivers” that define the high-level categories of issues on which the public thinks companies should be ranked. The current set of Drivers—those used in the 2017 model, and which will be used moving forward—were derived from processes spanning 2015 to 2017. This memo describes those processes.1
2015-2016: Deriving the Original 10 Drivers
Over the course of 2015, JUST Capital completed its inaugural survey work: an exhaustive, multiphase, seven-month research study to determine how Americans define JUST corporate behavior. At the commencement of the project, JUST Capital had no preconceived notion or expectation of the results the research would yield. JUST Capital followed no agenda other than asking about people’s views of corporations today, identifying the public’s definition of corporate JUSTness, and assessing the relative importance of components identified as important to JUST behavior. We believe it to be one of the largest sets of surveys of its kind ever done.
The purpose of the first phase of quantitative research, which was completed in Q2 of 2015, was to organize and prioritize the comprehensive list of 188 items which emerged from qualitative work – an extensive set of focus groups, interviews, and online discussions in more than two dozen locations nationally - into a list of the most importantComponents. JUST Capital surveyed the American people to confirm on a large scale what the qualitative respondents had said, based on a demographically representative online sample of 15,000 people. JUST Capital chose this large sample size in order to allow us to make statistically valid assessments of subgroup opinions. It then followed this survey with a confirmatory poll of 5,000 individuals.
Information JUST Capital gathered on the public’s definitions and priorities in each category was used to consolidate and organize the 188 items into a hierarchy containing the 36 most importantComponents of JUSTness, and further grouped into 10 common themes calledDrivers.
These are the specific steps JUST Capital took to build the hierarchy:
1. | Determine item importance: An analysis of the15,000-respondent survey was used to determine the most important of the 188 items. In order to qualify for continued consideration as part of the model, items had to meet the following criteria: |
a. | At least 75% of the American public selected the item as highly relevant to JUST corporate behavior. |
1 | For further details, including technical reports describing JUST Capital’s qualitative and quantitative survey research practices, please refer to the2017-18 JUST Capital Ranking Methodology and itsSurvey Research Appendices. |
Mr. Jay Williamson May 3, 2018 Page 9 |
b. | Each item must capture a separate and distinct behavior from every other item. |
c. | The item must be measurable and broadly applicable to most companies. |
2. | Create the Drivers: Additional analysis from the survey of 15,000 respondents used factor analysis to determine how the items that met the criteria at Step 1 grouped together statistically. The results of multiple factor analyses—with multiple specifications used to test the robustness of the results—demonstrated that the items could be grouped into 10 discrete,non-overlapping categories of behavior, which we named “Drivers.” |
3. | Create the Components:The remaining items were factor-analyzed within the Driver category they loaded onto in order to group items together for the Components. Again, multiple specifications were used to check the robustness of the results. |
These steps created the 10 Drivers and 36 Components that were used in the initial industry-level company ranking that JUST Capital undertook in 2016. Based on lessons learned from this experience and feedback from our research advisory council, a decision was taken to closely examine and reconfirm the driver groupings in 2017 in order to bring greater stability to the model over time.
2017: Public Testing and Refinement – From 10 to 7 Drivers
The Driver weights for 2017 differ from the 2016 weights for two reasons:
1. | In 2017, JUST Capital shifted the model from 2016’s Component-levelbottom-up approach, in which the American public rated Components and those were summed to get the Driver weights, to atop-down Driver-level approach in which Americans rated the Drivers themselves. |
2. | Additionally, as a result of qualitative and quantitative work in early 2017, the number of Drivers was reduced from ten to seven. |
Both changes were made to create a more stable model of JUST corporate behavior over time.We do not anticipate making such fundamental changes to the Drivers or the construction of the weights in the future, unless there are substantial movements in American opinion that demand it.
2016 Drivers and weights | ||||
Worker Pay & Benefits | 25 | % | ||
Worker Treatment | 24 | % | ||
Leadership & Ethics | 17 | % | ||
Customer Treatment | 7 | % | ||
Product Attributes | 6 | % | ||
Domestic Job Creation | 6 | % | ||
Environmental Impact | 5 | % | ||
Supply Chain Impact | 4 | % | ||
Investor Alignment | 4 | % | ||
Community Well-Being | 2 | % |
2017 Drivers and weights | ||||
Workers | 23 | % | ||
Customers | 19 | % | ||
Products | 17 | % | ||
Environment | 13 | % | ||
Communities | 11 | % | ||
Jobs | 10 | % | ||
Shareholders & Management | 6 | % |
Mr. Jay Williamson May 3, 2018 Page 10 |
1) Thetop-down approach
In 2016, the Component weights were summed to become the Driver weights. In thetop-down approach for 2017, Drivers are identified independently, and then Components are associated with them by theme. This architecture calls for a2-level process of weighting.
First, relative importance is derived at the Driver level using a MaxDiff design. The Driver weights derived are then compared to those obtained from at least one additional weighting methodology (point allocation) as well as a rank ordering exercise, to enhance JUST Capital’s confidence in the MaxDiff outcome. In 2017, the rank order and weights obtained from these additional approaches substantially replicated that of the MaxDiff.
Then, within each Driver, the relative importance of the Components is determined. This means that the full set of Components isn’t tested against itself. The Weight of the Components will be applied only within the group associated with the relevant Driver.
This decision resulted in changes to the Driver weights, as well as how individual Components affect the overall model. In 2016, summing the components to create the Driver weights meant that more Components within a given Driver caused higher weights for that Driver. For the 2017 model, survey respondents rated the Drivers separately so that the number of Components no longer has that effect on the model. That methodological shift accounts for the substantial differences in weights between 2016 and 2017.
2) Determining the Drivers – From 10 to 7
While the 10 groups had been statistically determined from the initial 188 items, those 10 groups had not been tested with the public after they were created. In the focus groups conducted by NORC in April of 2017, it became clear that respondents did not divide just corporate behavior into as many high-level categories as comprised the10-Driver setup. The focus groups were not given the actual Drivers from 2016, but were asked to think about the basic stakeholders in corporate justness. The focus groups identified5-8 high-level categories that could be Drivers.
These differed slightly from the 2016 Drivers. One clear conclusion was that most participants put all worker issues together instead of separating pay and benefits from treatment. Other recommendations were not as clear. For example, some of the focus groups thought customers and products should go together, while others did not agree. Some groups put leadership and investors together, and some did not.
Since we did not want to determine Driver composition based solely on the focus groups, JUST Capital conducted a series of quantitative surveys to ascertain those details from the broader American public. This work consisted of a nationally representative survey of 1,800 Americans designed for latent class analysis (LCA) conducted by NORC, and four additional nationally representative surveys of 1,000 Americans asking questions about the Drivers conducted by YouGov.
The LCA survey asked respondents to rate all the Components that had emerged from the Component-oriented focus groups by importance to justness. With this information, a latent class analysis was performed. LCA techniques are statistical methods designed to uncover an underlying construct that is not directly measured. Thus, we hoped to uncover the potential Drivers using these Component-level responses. Both NORC and JUST independently analyzed these data using slightly different approaches –NORC used cluster analysis, and JUST used factor analysis – to look at the number of Drivers and how
Mr. Jay Williamson May 3, 2018 Page 11 |
the Components aligned within them. The cluster and factor analyses independently demonstrated that seven was the most statistically robust number of Drivers, that the two workers Drivers should be combined, and suggested that we should examine how we composed the communities, international, investor, and leadership issues into Drivers.
To investigate those issues further, we conducted a series of four web panel surveys with YouGov. The first targeted the number and composition of the Drivers, asking respondents to select which of a few lists of potential Drivers made most sense to them. (The other three focused on the Components within Drivers and are discussed in the Components section.) We also asked if combinations of the Drivers made sense – for example, should “creating jobs” be its own Driver, or should it be combined with “communities” or “workers”? Should “leadership” and “investors” go together, or be separate Drivers? The results suggested that we should consider combining international and local communities, leadership and investors, and keep jobs independent.
With all this information and input from members of its Research Advisory Council, JUST Capital formulated its recommendation that a7-Driver composition be used in the 2017 model. In the assessment of its analysts, JUST Capital took the view that this7-Driver structure presented the most accurate, unbiased and comprehensive picture of corporate justness as defined by the public. This recommendation was endorsed by the Research Committee of the Board. The changes are as follows:
• | All the focus group and survey work confirmed that there should be one worker category, not two. |
• | There was some ambiguity with treating international communities (“supply chain” in the 2016 Drivers) as separate from domestic communities in the focus groups. Some viewed these as united under the “communities” umbrella, others did not. In thefollow-up survey work, a majority of respondents said that international communities should be treated the same as U.S. communities. The cluster and factor analyses confirmed that this was a reasonable combination, and so “supply chain impact” and “community well-being” became “communities.” |
• | Similarly, the focus groups were mixed on putting leadership and investor issues together. A majority of quantitative survey respondents said these should go together, and again, the cluster and factor analyses supported that decision. “Leadership and ethics” and “investor alignment” became “shareholders and management.” |
2018 and Beyond: Stability
Given the extremely robust nature of the analyses that has gone into determining the Drivers – qualitative analysis followed by quantitative modeling and confirmation in 2015 and 2016, followed by testing with the public in 2017 and more quantitative modeling and confirmation in 2017 – at the end of our 2017 process, our Board, Research Committee, and staff agreed that the Drivers are in their ideal form. We do not anticipate replicating those processes unless we detect a large shift in public opinion concerning the key stakeholders of corporate justness. We are vigilant in searching for signals of such shifts in all our survey research. The magnitude of a shift in public opinion required to disrupt our current Driver composition would likely manifest itself in both our qualitative and quantitative work via measures of respondent comprehension. We test for this comprehension and clarity in all of our survey processes.
Mr. Jay Williamson May 3, 2018 Page 12 |
Exhibit B
The Fund seeks to achieve its investment objective by investing at least 80% of its assets (exclusive of collateral held from securities lending) in securities included in its underlying index, in depositary receipts representing securities included in its underlying index and in underlying stocks in respect of depositary receipts included in its underlying index. Shareholders will be provided with sixty days’ notice in the manner prescribed by the Securities and Exchange Commission (“SEC”) before any change in the Fund’s policy to invest at least 80% of its assets in securities included in its underlying index.
The Index is designed to deliver exposure to equity securities of large capitalization U.S. issuers that engage in “just business behavior” based on rankings produced by the Index Provider. The Index Provider publishes an annual ranking of issuers in the Reference Index based on their quantitative performance assessment of seven issue areas: how they treat their workers, their customers, the communities they interact with, the environment, and their shareholders, their commitment to making quality and beneficial products, and job creation.
The Index Provider seeks to encourage greater accountability in the business community and drive positive change among large publicly-traded U.S. corporations by (a) defining business behaviors that the American public cares most about (through extensive qualitative and quantitative survey research), (b) developing metrics that correspond to these issues in accordance with a robust, transparent methodology, (c) ranking the largest publicly traded U.S. companies on the basis of these metrics, and (d) developing tools and products that allow investors to direct capital towards more “just” companies.
To construct the annual ranking, the Index Provider conducts representative surveys of the American public on a regular basis in order to understand what issues represent “just” corporate behavior, how these issues should be described, and the relative importance of each issue. For the ranking published in 2017, the Index Provider utilized the following “drivers” (weighted as indicated), as an indication of “just” corporate behavior based on its survey research:
• | Workers (23%) –Components include payment of living and fair wages, safety in the workplace, and access to health insurance |
• | Customers (19%) –Components includenon-discrimination policies and customer privacy |
• | Products (17%) –Components include quality of products, benefits to health, environment or society, and fair pricing |
• | Environment (13%) –Components include minimization of pollution and efficient use of resources |
• | Communities (11%) –Components include avoidance of business relationships with oppressive or abusive entities |
• | Jobs (10%) –Components include creation of jobs in the United States and absolute number of jobs in the United States |
• | Shareholders and Management (6%) –Components include compliance with laws, integrity, and investor returns |
The Index Provider develops metrics and collects data on how the companies in the Reference Index perform across these issues. To assess each metric, the Index Provider utilizes data derived from various sources, including publicly available company reports, third-party data vendors, government data, academic andnot-for-profit data, press coverage, and crowd-sourced data. The Index Provider constructs a ranking model that ranks each company based on these metrics.