Correspondence 0001592900-24-001319 from EA Series Trust (CIK 0001592900)
EA Series Trust (CIK 0001592900)
Date: Aug. 22, 2024 · CIK: 0001592900 · Accession: 0001592900-24-001319
AI Filing Summary & Sentiment
File numbers found in text: 333-195493, 811-22961
Letter
100 F Street, NE Washington, DC 20549 Post-Effective Amendment No. 352 to the Registration Statement on Form N-1A (the “Amendment”) File Nos.: 333-195493 and 811-22961
RE: EA Series Trust (the “Trust”)
Dear Ms. White:
This correspondence responds to legal comments received by the undersigned from the staff of the U.S. Securities and Exchange Commission (the “Staff” of the “Commission”) on July 19, 2024, with respect to the Post-Effective Amendment No. 352 to the Trust’s Registration Statement on Form N-1A under the Securities Act of 1933 relating to Intelligent Livermore ETF, Intelligent Omaha ETF, Intelligent Equal Select ETF, Intelligent Tech Focus ETF, and the Intelligent Small Cap Select ETF (each a “Fund”, and collectively the “Funds”), each a proposed new series of the Trust. Capitalized terms not otherwise defined have the same meaning as in the Registration Statement.
Global Comments
1.Comment: The human initiation step involves a number of choices that appear relevant to the overall strategy such as intended concentration level, specific factors/themes etc. For each Fund revise the strategy to address the “intended strategy” for the underlying portfolio. Later explain how the analyst defines and collects data and assesses its accuracy, completeness and usefulness.
Response: The Registrant has revised each investment strategy to provide the specific strategy parameters at the beginning of each Principal Investment Strategy section.
The analyst uses a global template for relevant data that includes historical results and forward estimates of major financial items like revenue, earnings, free cash flow, etc. Beyond the global data template, the analyst may add certain additional data that is relevant to the intended strategy (e.g., data on revenue or earnings growth for a growth-oriented strategy). The analyst relies on a long-standing third-party data vendor (e.g., Factset) for accuracy and usefulness combined with testing and review of the outputs of the AI Models.
2.Comment: Please clarify what you mean by “leading large language models.” Is this based on a ranking or objective metrics?
Response: The Registrant has deleted the reference to the term “leading.”
3.Comment: Please clarify what is meant by the statement in Step 2 that the “AI Models represent the Fund’s Investment Committee.”
Response: In many investment processes, investment decisions are made by a committee of several people who are tasked with carrying out the objectives of the investment manager or the specific investment strategy. The aggregate of the committee votes provides the ultimate decision of the committee. Intelligent Alpha, the Sub-Adviser to the Funds (“Intelligent Alpha”), uses multiple large-language AI models to form a committee where AI replaces humans. Intelligent Alpha’s experience suggests that the models answer questions differently than one another, forming a valuable difference of opinion on the committee that we believe can help drive better outcomes. Step 2 of each Principal Investment Strategy section has been updated to include a sentence at the end of the first paragraph as follows:
11300 Tomahawk Creek Pkwy, Suite 310 ● Leawood, KS 66211
Practus, LLP ● Practus.com
The AI Models represent the Fund’s AI investment committee, whereby the AI Models provide their independently created portfolios based on each AI Model’s independent evaluation of the instructions provided to it.
4.Comment: Please include disclosure of securities lending as a principal investment strategy.
Response: Securities lending risk has been deleted from each Fund’s Principal Risks section.
5.Comment: For all Funds except the Small Cap Select ETF, please provide more disclosure regarding how the analyst will determine that 50% data sought about a company is available.
Response: The Analyst determines whether 50% of the data is available by assessing what percentage of data entries in the overall dataset are a null or otherwise a void value that would be useless in quantitative analysis. The Registrant has revised the disclosure for each Fund to include the following disclosure:
Sufficient data is generally considered to 50% of the data entries in the overall dataset that is sought by the Model AI.
6.Comment: Please include disclosure in the strategies regarding why each Fund is subject to high portfolio turnover risk.
Response: The following language has been added to each Fund’s Principal Investment Strategies section:
The Fund is expected to have high portfolio turnover based on historical testing of the Fund’s investment strategy. During the portfolio review process the AI may remove or add stocks to the portfolio or change the weight of stocks already in the portfolio.
7.Comment: Please provide completed fee tables for each Fund.
Response: The Registrant has provided the fee tables under separate cover.
8.Comment: Please clarify what is meant by intended concentration under Step 1 in each Fund’s Principal Investment Strategies section.
Response: The following amendment has been made to each Fund’s disclosures (addition in bold):
A human analyst (the “Analyst”) establishes the intended strategy for the underlying portfolio including the target investment universe, the portfolio size, the intended concentration level (e.g., maximum security weighting), and any specific factors or themes to be highlighted in the portfolio (e.g. dividends, quality).
Intelligent Livermore ETF
9.Comment: Please clarify what the term “Livermore” signifies in the Fund’s name and why it should not be included in an 80% investment policy.
Response: Jesse Livermore was one of the first Great Wall Street traders in the early 1900s. His namesake inspires the strategy’s aims to use the philosophies of the world’s best traders and investors in Intelligent Alpha’s AI-powered selection process. The term does not suggest the fund focuses on a particular type of investment, industry, geographic focus or connote an investment that has particular characteristics under Rule 35d-1 under the Investment Company Act of 1940, as amended (the “1940 Act”).
10.Comment: In the first sentence of the Principal Investment Strategies section, it says that the AI will create an intelligent equal weight portfolio of US large cap stocks. Please reconcile this language with the language on page 2 that says that the portfolio will be investment in global equity securities.
Response: The first sentence has been updated to change “US” to “global.”
11.Comment: In Step 1, please explain what is meant by “concentration level” to be highlighted in the portfolio.
Response: The concentration level refers to the maximum weighting the AI investment committee is allowed to ascribe to a single holding, which is 10% for the Livermore Fund.
12.Comment: For the Livermore and Omaha ETFs, given each Fund’s large cap focus, please delete small- and mid-capitalization risk.
Response: The Registrant has updated the disclosure accordingly but has retained mid-cap risk for the Livermore Fund given the investment universe begins at a market capitalization of $1 billion.
13.Comment: In the second paragraph of step 1, how do you expect that the 4-6 major trading trends will be profitable given the lag time between the identification of the major trading trends as compared to when the portfolio may execute on any such trends?
Response: Intelligent Alpha instructs the AI to optimize for trading trends that have a duration of 6-24 months or more given that the portfolio is generally reviewed on a quarterly basis. Shorter-term trading trends would likely not be profitable unless the strategy were optimized to trade more frequently. Intelligent Alpha believes that identifying trading trends that have been in place for longer durations as described is the best fit for this investment strategy. The disclosure in the second full paragraph of Step 1 has been revised accordingly.
14.Comment: For the Livermore and Omaha ETFs, please provide more information in the filing about how the analyst will measure and define the greatest traders in the world.
Response: The Analyst defines great investors and traders based on an investor’s long-term track record (5+ years) as compared to broad market benchmarks. For the Livermore ETF, the second sentence of the second paragraph has been amended as follows:
The Analyst will define the list of famous traders and investors for the AI by reviewing long-term (5 years or greater) track records of famous investors as compared to broad market benchmarks, which will vary based on the nature of the trading undertaken by the famous investor. A major trading trend is a clearly defined and articulated trading view expressed by the famous trader. The major trading trends identified by the AI define the thematic, sector, and other broad-based exposures that the AI Models believe the portfolio should represent as inspired by the trading trends identified for the list of greatest traders.
For the Omaha ETF, the following has been added as the third to last sentence of the second paragraph in Step 1:
The Analyst focuses on investors with a stronger view toward quality and value investments with the potential for multi-year investment duration.
15.Comment: For the Livermore and Omaha ETFs, please explain in the filing what 13F filings are, including what they show, who files them and when they are filed. In the risk disclosure explain the impact of any lag in reporting on implementing the strategy.
Response: The Registrant has added the following disclosure in Step 1 of the Livermore and Omaha Funds:
Form 13F is a quarterly report filed by institutional investment managers to disclose their U.S. equity holdings to the SEC.
The Registrant has added the following risk disclosure to each of the Livermore and Omaha ETFs:
Risks Related to the Use of Form 13F Data. The Form 13F filings used to analyze trading trends are filed up to 45 days after the end of each calendar quarter. Therefore, a given investor may have already sold its position by the time the AI Models evaluate the filing. Furthermore, the Form 13F filing may only disclose a subset of a particular investor’s holdings, as not all securities are required to be reported on the Form 13F. As a result, the Form 13F may not provide a complete picture of the holdings of a given investor. An investor may hold long positions for a number of reasons, and the AI Models may not appreciate the reasons or the strategies followed by an investor who makes the filings. The analysis of the AI Models may not be representative of the investor’s universe or the strategies that give rise to the reported holdings. Because the Form 13F filing is publicly available information, it is possible that other investors are also monitoring these filings and investing accordingly. This may result in inflation of the share price of securities in which the Fund invests.
16.Comment: On page 2, it provides that the Analyst may in its discretion include leveraged or inverse leveraged index ETFs, please add inverse ETFs to the types of ETFs the Fund may select consistent with risk disclosure. In addition, please clarify why the AI or the Analyst would select a leveraged or inverse leveraged ETF versus a global equity security.
Response: Inverse ETFs have been removed from the risk disclosure. If the AI identifies a particular trend that may be best represented through a long or short ETF as opposed to individual securities or non-levered ETFs (e.g. a trade involving a broad basket of equities), the Analyst will select an ETF.
17.Comment: On page 2, the second sentence of the first full paragraph says that a proprietary curated universe of global equity securities … is created. Please clarify if this universe is created by the Analyst or the AI.
Response: The curated universe is created in tandem between the analyst and AI where the AI identifies relevant themes and potentially sectors while the analyst is responsible for gathering securities relevant to the theme, sector, etc. The second sentence has been amended as follows (addition in bold):
Next, given the trading trends identified, the Analyst curates a proprietary universe of global equity securities with a market capitalization of over $1 billion at the time of the AI’s review is created.
18.Comment: Please provide additional details as to (1) how the AI Models were developed for application to investment selection, including a discussion of signal to noise and interpretability or explainability factors; (2) the types of data inputs entered into the AI Models and the types of signals or outputs achieved; (3) how the AI Models data is validated for accuracy and reliability and by whom and what specific outputs are considered accurate and reliable.
Response: Intelligent Alpha creates specific, detailed instruction sets for investment purposes that are applied to large-language AI models developed by third parties. The instruction sets were developed through an iterative testing process where Intelligent Alpha benchmarks historical and live results compared to relevant comparable indexes. As with all good AI-driven processes, Intelligent Alpha is testing and upgrade our processes on an on-going basis. As part of the testing process, Intelligent Alpha attempts to identify and eliminate signals that appear to weaken outputs based on historical benchmarking.
Explainability is a byproduct of the consistent process we use with the AI models. Intelligent Alpha’s process controls for the types of securities being analyzed, the maximum concentration levels in portfolios, and the philosophy on how the AI should think as an investor. The AI models can be asked why certain securities were selected as part of a portfolio when necessary.
The data input into the AI models consist of large sets of quantitative financial data including historical performance and forward estimates of data points such as revenue, earnings, free cash flow, free cash flow yield, and other financial metrics. We may also utilize proprietary estimates generated by AI as part of portfolio creation. The output of the combination of the data we curate, and the instruction set we provide is a portfolio of specific stocks and weightings (if weightings are requested).
The large-language models that are utilized are built on large proprietary datasets to which Intelligent Alpha does not have access. For the data we utilize in stock analysis and portfolio creation, we rely on third party data vendors with long standing histories in the financial services industry (e.g., Factset). Outputs from our process applied to the AI models are reviewed to determine reliability by the human analyst who has final oversight of the portfolio.
19.Comment: Please clarify the relationship between the 4-6 major trends identified at the start of step 1 and the philosophy developed for use in Step 2. Consider providing examples.
Response: The major trends define the thematic, sector, and other broad exposures the AI Models believe the portfolio should represent as inspired by the great traders identified. The Philosophy defines how the AI Models select securities to represent those themes, sectors, etc. For example, the AI Models may identify a trend to invest in European stocks. The Philosophy may instruct the AI Models to optimize for European stocks that have certain characteristics like strong business moats, stock momentum, high free cash flow yields, etc. The Registrant has added this information before the last sentence of the third paragraph in Step 1.
20.Comment: Please disclose the Livermore ETFs total number of fund holdings.
Response: The portfolio will generally consist of 60-90 total holdings. The following language has been added as
Show Raw Text
CORRESP 1 filename1.htm Document August 22, 2024 Ms. Alison White U.S. Securities and Exchange Commission 100 F Street, NE Washington, DC 20549 RE: EA Series Trust (the “Trust”) Post-Effective Amendment No. 352 to the Registration Statement on Form N-1A (the “Amendment”) File Nos.: 333-195493 and 811-22961 Dear Ms. White: This correspondence responds to legal comments received by the undersigned from the staff of the U.S. Securities and Exchange Commission (the “Staff” of the “Commission”) on July 19, 2024, with respect to the Post-Effective Amendment No. 352 to the Trust’s Registration Statement on Form N-1A under the Securities Act of 1933 relating to Intelligent Livermore ETF, Intelligent Omaha ETF, Intelligent Equal Select ETF, Intelligent Tech Focus ETF, and the Intelligent Small Cap Select ETF (each a “Fund”, and collectively the “Funds”), each a proposed new series of the Trust. Capitalized terms not otherwise defined have the same meaning as in the Registration Statement. Global Comments 1.Comment: The human initiation step involves a number of choices that appear relevant to the overall strategy such as intended concentration level, specific factors/themes etc. For each Fund revise the strategy to address the “intended strategy” for the underlying portfolio. Later explain how the analyst defines and collects data and assesses its accuracy, completeness and usefulness. Response: The Registrant has revised each investment strategy to provide the specific strategy parameters at the beginning of each Principal Investment Strategy section. The analyst uses a global template for relevant data that includes historical results and forward estimates of major financial items like revenue, earnings, free cash flow, etc. Beyond the global data template, the analyst may add certain additional data that is relevant to the intended strategy (e.g., data on revenue or earnings growth for a growth-oriented strategy). The analyst relies on a long-standing third-party data vendor (e.g., Factset) for accuracy and usefulness combined with testing and review of the outputs of the AI Models. 2.Comment: Please clarify what you mean by “leading large language models.” Is this based on a ranking or objective metrics? Response: The Registrant has deleted the reference to the term “leading.” 3.Comment: Please clarify what is meant by the statement in Step 2 that the “AI Models represent the Fund’s Investment Committee.” Response: In many investment processes, investment decisions are made by a committee of several people who are tasked with carrying out the objectives of the investment manager or the specific investment strategy. The aggregate of the committee votes provides the ultimate decision of the committee. Intelligent Alpha, the Sub-Adviser to the Funds (“Intelligent Alpha”), uses multiple large-language AI models to form a committee where AI replaces humans. Intelligent Alpha’s experience suggests that the models answer questions differently than one another, forming a valuable difference of opinion on the committee that we believe can help drive better outcomes. Step 2 of each Principal Investment Strategy section has been updated to include a sentence at the end of the first paragraph as follows: 11300 Tomahawk Creek Pkwy, Suite 310 ● Leawood, KS 66211 Practus, LLP ● Practus.com The AI Models represent the Fund’s AI investment committee, whereby the AI Models provide their independently created portfolios based on each AI Model’s independent evaluation of the instructions provided to it. 4.Comment: Please include disclosure of securities lending as a principal investment strategy. Response: Securities lending risk has been deleted from each Fund’s Principal Risks section. 5.Comment: For all Funds except the Small Cap Select ETF, please provide more disclosure regarding how the analyst will determine that 50% data sought about a company is available. Response: The Analyst determines whether 50% of the data is available by assessing what percentage of data entries in the overall dataset are a null or otherwise a void value that would be useless in quantitative analysis. The Registrant has revised the disclosure for each Fund to include the following disclosure: Sufficient data is generally considered to 50% of the data entries in the overall dataset that is sought by the Model AI. 6.Comment: Please include disclosure in the strategies regarding why each Fund is subject to high portfolio turnover risk. Response: The following language has been added to each Fund’s Principal Investment Strategies section: The Fund is expected to have high portfolio turnover based on historical testing of the Fund’s investment strategy. During the portfolio review process the AI may remove or add stocks to the portfolio or change the weight of stocks already in the portfolio. 7.Comment: Please provide completed fee tables for each Fund. Response: The Registrant has provided the fee tables under separate cover. 8.Comment: Please clarify what is meant by intended concentration under Step 1 in each Fund’s Principal Investment Strategies section. Response: The following amendment has been made to each Fund’s disclosures (addition in bold): A human analyst (the “Analyst”) establishes the intended strategy for the underlying portfolio including the target investment universe, the portfolio size, the intended concentration level (e.g., maximum security weighting), and any specific factors or themes to be highlighted in the portfolio (e.g. dividends, quality). Intelligent Livermore ETF 9.Comment: Please clarify what the term “Livermore” signifies in the Fund’s name and why it should not be included in an 80% investment policy. Response: Jesse Livermore was one of the first Great Wall Street traders in the early 1900s. His namesake inspires the strategy’s aims to use the philosophies of the world’s best traders and investors in Intelligent Alpha’s AI-powered selection process. The term does not suggest the fund focuses on a particular type of investment, industry, geographic focus or connote an investment that has particular characteristics under Rule 35d-1 under the Investment Company Act of 1940, as amended (the “1940 Act”). 10.Comment: In the first sentence of the Principal Investment Strategies section, it says that the AI will create an intelligent equal weight portfolio of US large cap stocks. Please reconcile this language with the language on page 2 that says that the portfolio will be investment in global equity securities. 2 Response: The first sentence has been updated to change “US” to “global.” 11.Comment: In Step 1, please explain what is meant by “concentration level” to be highlighted in the portfolio. Response: The concentration level refers to the maximum weighting the AI investment committee is allowed to ascribe to a single holding, which is 10% for the Livermore Fund. 12.Comment: For the Livermore and Omaha ETFs, given each Fund’s large cap focus, please delete small- and mid-capitalization risk. Response: The Registrant has updated the disclosure accordingly but has retained mid-cap risk for the Livermore Fund given the investment universe begins at a market capitalization of $1 billion. 13.Comment: In the second paragraph of step 1, how do you expect that the 4-6 major trading trends will be profitable given the lag time between the identification of the major trading trends as compared to when the portfolio may execute on any such trends? Response: Intelligent Alpha instructs the AI to optimize for trading trends that have a duration of 6-24 months or more given that the portfolio is generally reviewed on a quarterly basis. Shorter-term trading trends would likely not be profitable unless the strategy were optimized to trade more frequently. Intelligent Alpha believes that identifying trading trends that have been in place for longer durations as described is the best fit for this investment strategy. The disclosure in the second full paragraph of Step 1 has been revised accordingly. 14.Comment: For the Livermore and Omaha ETFs, please provide more information in the filing about how the analyst will measure and define the greatest traders in the world. Response: The Analyst defines great investors and traders based on an investor’s long-term track record (5+ years) as compared to broad market benchmarks. For the Livermore ETF, the second sentence of the second paragraph has been amended as follows: The Analyst will define the list of famous traders and investors for the AI by reviewing long-term (5 years or greater) track records of famous investors as compared to broad market benchmarks, which will vary based on the nature of the trading undertaken by the famous investor. A major trading trend is a clearly defined and articulated trading view expressed by the famous trader. The major trading trends identified by the AI define the thematic, sector, and other broad-based exposures that the AI Models believe the portfolio should represent as inspired by the trading trends identified for the list of greatest traders. For the Omaha ETF, the following has been added as the third to last sentence of the second paragraph in Step 1: The Analyst focuses on investors with a stronger view toward quality and value investments with the potential for multi-year investment duration. 15.Comment: For the Livermore and Omaha ETFs, please explain in the filing what 13F filings are, including what they show, who files them and when they are filed. In the risk disclosure explain the impact of any lag in reporting on implementing the strategy. Response: The Registrant has added the following disclosure in Step 1 of the Livermore and Omaha Funds: Form 13F is a quarterly report filed by institutional investment managers to disclose their U.S. equity holdings to the SEC. The Registrant has added the following risk disclosure to each of the Livermore and Omaha ETFs: 3 Risks Related to the Use of Form 13F Data. The Form 13F filings used to analyze trading trends are filed up to 45 days after the end of each calendar quarter. Therefore, a given investor may have already sold its position by the time the AI Models evaluate the filing. Furthermore, the Form 13F filing may only disclose a subset of a particular investor’s holdings, as not all securities are required to be reported on the Form 13F. As a result, the Form 13F may not provide a complete picture of the holdings of a given investor. An investor may hold long positions for a number of reasons, and the AI Models may not appreciate the reasons or the strategies followed by an investor who makes the filings. The analysis of the AI Models may not be representative of the investor’s universe or the strategies that give rise to the reported holdings. Because the Form 13F filing is publicly available information, it is possible that other investors are also monitoring these filings and investing accordingly. This may result in inflation of the share price of securities in which the Fund invests. 16.Comment: On page 2, it provides that the Analyst may in its discretion include leveraged or inverse leveraged index ETFs, please add inverse ETFs to the types of ETFs the Fund may select consistent with risk disclosure. In addition, please clarify why the AI or the Analyst would select a leveraged or inverse leveraged ETF versus a global equity security. Response: Inverse ETFs have been removed from the risk disclosure. If the AI identifies a particular trend that may be best represented through a long or short ETF as opposed to individual securities or non-levered ETFs (e.g. a trade involving a broad basket of equities), the Analyst will select an ETF. 17.Comment: On page 2, the second sentence of the first full paragraph says that a proprietary curated universe of global equity securities … is created. Please clarify if this universe is created by the Analyst or the AI. Response: The curated universe is created in tandem between the analyst and AI where the AI identifies relevant themes and potentially sectors while the analyst is responsible for gathering securities relevant to the theme, sector, etc. The second sentence has been amended as follows (addition in bold): Next, given the trading trends identified, the Analyst curates a proprietary universe of global equity securities with a market capitalization of over $1 billion at the time of the AI’s review is created. 18.Comment: Please provide additional details as to (1) how the AI Models were developed for application to investment selection, including a discussion of signal to noise and interpretability or explainability factors; (2) the types of data inputs entered into the AI Models and the types of signals or outputs achieved; (3) how the AI Models data is validated for accuracy and reliability and by whom and what specific outputs are considered accurate and reliable. Response: Intelligent Alpha creates specific, detailed instruction sets for investment purposes that are applied to large-language AI models developed by third parties. The instruction sets were developed through an iterative testing process where Intelligent Alpha benchmarks historical and live results compared to relevant comparable indexes. As with all good AI-driven processes, Intelligent Alpha is testing and upgrade our processes on an on-going basis. As part of the testing process, Intelligent Alpha attempts to identify and eliminate signals that appear to weaken outputs based on historical benchmarking. Explainability is a byproduct of the consistent process we use with the AI models. Intelligent Alpha’s process controls for the types of securities being analyzed, the maximum concentration levels in portfolios, and the philosophy on how the AI should think as an investor. The AI models can be asked why certain securities were selected as part of a portfolio when necessary. 4 The data input into the AI models consist of large sets of quantitative financial data including historical performance and forward estimates of data points such as revenue, earnings, free cash flow, free cash flow yield, and other financial metrics. We may also utilize proprietary estimates generated by AI as part of portfolio creation. The output of the combination of the data we curate, and the instruction set we provide is a portfolio of specific stocks and weightings (if weightings are requested). The large-language models that are utilized are built on large proprietary datasets to which Intelligent Alpha does not have access. For the data we utilize in stock analysis and portfolio creation, we rely on third party data vendors with long standing histories in the financial services industry (e.g., Factset). Outputs from our process applied to the AI models are reviewed to determine reliability by the human analyst who has final oversight of the portfolio. 19.Comment: Please clarify the relationship between the 4-6 major trends identified at the start of step 1 and the philosophy developed for use in Step 2. Consider providing examples. Response: The major trends define the thematic, sector, and other broad exposures the AI Models believe the portfolio should represent as inspired by the great traders identified. The Philosophy defines how the AI Models select securities to represent those themes, sectors, etc. For example, the AI Models may identify a trend to invest in European stocks. The Philosophy may instruct the AI Models to optimize for European stocks that have certain characteristics like strong business moats, stock momentum, high free cash flow yields, etc. The Registrant has added this information before the last sentence of the third paragraph in Step 1. 20.Comment: Please disclose the Livermore ETFs total number of fund holdings. Response: The portfolio will generally consist of 60-90 total holdings. The following language has been added as