Apr. 20, 2026
Exhibit A — Comparative Retention Matrix: https://dropbox.com/scl/fi/4k3zsdzziur0lvfpj5l0y/Exhibit_A_Comparative_Retention_Matrix.pdf?rlkey=60axopm5fx300h0fn0fm97m6n&dl=0 Exhibit B — Enforcement Timeline Matrix: https://dropbox.com/scl/fi/sbuj5x9d3cksg5gqjw653/Exhibit_B_Enforcement_Timeline_Matrix.pdf?rlkey=aebswl7gbsc73yta37nyavyq2&dl=0 Exhibit C — AWS Cost-Tier Analysis: https://dropbox.com/scl/fi/99j2k0nv31jk04oyv7k8q/Exhibit_C_AWS_Cost_Tier_Analysis.pdf?rlkey=74z1t3pn7mmaquyysdb1vcq3q&dl=0 Exhibit D — Three-Scenario Cost Model (Excel workbook, 138 formulas): https://dropbox.com/scl/fi/5n9ljq1nhq6pxffr5kdhu/Exhibit_D_Three_Scenario_Cost_Model.xlsxrlkey=9hpz54y5n162v273od8i15usg&dl=0 Exhibit E — Redline of Proposed Amendments to CAT NMS Plan: Appendix D, § 6.4 :https://www.dropbox.com/scl/fi/dbxzs88b6dy9c61htxqhc/Exhibit_E_Redline_Appendix_D_Section_6.4.pdf?rlkey=5of13h0l6snj8abk77jipx46m&e=1&dl=0 Figure 1 "Enforcement Timeline vs. the Three-Year Retention Cliff" below is the entire argument. (each of the following 6 charts are included in the 26 page Comment Letter) Six of the last seven headline market-manipulation cases: Sarao, Goldman 1MDB, JPMorgan spoofing, Archegos, Martoma, Musk/Tesla, resolved past the three-year deletion cliff the Commission just approved for the Consolidated Audit Trail. JPMorgan: 12.8 years. Sarao: 11.1 years. 1MDB: 8.8 years. The Commission just voted to delete the evidence before cases like these close. The one case that fits inside three years (Nuveen / Billimek) is the only one publicly credited to CAT detection. That is not a coincidence. That is the system working, and now being dismantled. A tape that looks 3 years into the past cannot prosecute a 10-year statute of limitations, IMO.Exhibit Figure 2 —"23× Storage Cost Gap" The SEC proposes to shorten CAT retention from 5 years (originally it was 6) down to 3 to save money. Here is the cost argument, at AWS April 2026 actual publicly available pricing: - S3 Standard: $276,000 / PB-year - S3 Glacier Deep Archive: $11,880 / PB-year With compression and an expected federal discount: ~$990 / PB-year That is a 96% reduction... available today, no deletion required. The Commission chose deletion over architecture. It paid for cost savings with evidence. Every SEC enforcement lawyer who has ever needed a five-year-old order book will lose the ability to pull it. Figure 3 — "The U.S. as Retention Outlier" The regulator's own consolidated tape [proposed plan], will then have the SHORTEST retention horizon in U.S. and international securities regulation. Everyone the SEC regulates has to keep records longer than the SEC keeps its own ?!? Currently: - SOX § 802: 7 years - MiFID II transactions: up to 7 years - IRS bad-debt records: 7 years - 17a-4(a) broker-dealer core records: 6 years - CFTC Rule 1.31: 5 years - Bank Secrecy Act SAR docs: 5 years - CAT post-Amendment: 3 years Did you know? A European hedge fund trading a U.S.-listed equity through a CAT-reporting broker has to keep its records for 7 years. The SEC's own tape of the same trade is now proposed to be gone after 3. Can IOSCO peer benchmarking defend the SEC's 3yr approach and logic? Can ESMA cross-border coordination defend the SEC's 3yr approach and logic? In essence, the SEC has assigned itself the shortest memory in the architecture. Make it make sense… Figure 4 — "Statutory Remedies Outlive the CAT" The SEC has the statutory authority to recover scienter-based disgorgement for 10 years. Congress codified that in NDAA § 6501 in 2021. The SEC just proposed to delete the evidence after 3? Example - for conduct occurring in April 2026: - CAT deleted: year 3 (April 2029) - Rule 10b-5 private repose: year 5 - Criminal securities fraud SOL: year 6 - State UFTA reach-back via § 544: year 6 - NDAA § 6501 scienter disgorgement SOL: year 10 Every statutory remedy the Commission is authorized to pursue outlives the evidence it just proposed to destroy. You cannot litigate a 10-year disgorgement claim on a 3-year tape IMO. You are asking the Department of Justice to prosecute with the lights off. Figure 5 — "Longer Retention, Lower Cost" This is what the Commission's Amendment looks like when you actually do the math. When you model the annual storage cost using existing AWS pricing (as of April 2026) for a 30 PB initial storage footprint, with an applied 50% federal enterprise volume discount (full spreadsheet model is provided links above - Exhibit D) : - Scenario 1 (pre-Amendment, 5-year): $1,635,329 / yr - Scenario 2 (approved 3-year Amendment): $1,765,462 / yr - Scenario 3 (the newly proposed 7-year tiered architecture): $389,545 / yr Read that again. The approved 3-year Amendment could be MORE expensive than the 5-year status quo it replaced. And the 7-year alternative proposed in our submitted File No. 4-698 comment costs $1.38 million per year less than the 3-year Amendment, while preserving four extra years of evidence. Feel free to review the provided pricing model spreadsheet and comment... but it looks like our SEC proposal, even with no discount, would lead to: 1. Longest CAT data retention. 2. Lowest CAT data cost. 3. Enhanced enforcement. The Commission can have ALL three, however, it just made a proposal for NONE of them. \\\ Figure 6 — "Proposed Tiered Storage Lifecycle" Here is the architecture I filed with the SEC on File No. 4-698. Ours is NOT deletion, it proposes Tiering: 1. Days 0–90: AWS S3 Standard — $276K / PB-year, millisecond query 2. Months 3–12: AWS Intelligent-Tiering — workload-adaptive 3. Years 1–3: AWS Glacier Instant Retrieval — $48K / PB-year, millisecond query 4. Years 3–7: AWS Glacier Deep Archive + Object Lock (Compliance Mode) — $11,880 / PB-year, 12-hour bulk SLA, WORM-immutable, SHA-256 hashed, chain-of-custody preserved. 1. The three-year "hot" direct-query horizon the Commission said it wanted? Preserved. 2. The cost savings the Commission said it needed? Delivered...bigger + without any Data deletion. This was never a budget problem. It was an architecture problem in my humble opinion. Good news: Commission still has the time and opportunity to avoid selecting the wrong lever. Full submission, exhibits, and redline to CAT NMS Plan § 6.5(b)(i) on the File No. 4-698 docket soon. (In the meantime, check our my dropbox links for full doc files submitted) Hope this helps. THANK YOU @SECGov for the opportunity to comment. Looking forward to additions, corrections, suggestions on this important topic.