AGENTIC CAPITAL
Track Record Substantiation
Source-of-truth proof document for every figure published on the public Agentic Capital track-record page. Each numeric claim is identified, traced to the underlying Fidelity statement, and shown with the math that produced it.
Prepared: 2026-05-08 For: Outside counsel reviewing SEC Rule 206(4)-1 (Marketing Rule) compliance Account under review: Fidelity Brokerage Account Z06-683363 — personal trading account of Dean Gallagher, General Partner of Agentic Capital Fund LP (in formation) Period: January 2023 – April 2026 (40 months) Source documents: 34 monthly and bi-monthly Fidelity statements (originals on file)
This document outlines every numeric claim that appears on the public track-record page, identifies the source document(s) it derives from, and shows the math that produced it. The intent is to allow counsel to verify each claim against the underlying Fidelity statements without ambiguity.
*). Per source convention, the period return is split equally across both calendar months for display purposes only; the underlying period return is the figure verified against the source statement. Source-statement count tie-out: 28 monthly + 6 bi-monthly = 34 source statements total (matches the publicly-claimed figure).These are the only inputs that drive every chained-return claim on the page. Each row is verified against the Fidelity statement for that period.
| # | Month | TWR | Source statement | Notes |
|---|---|---|---|---|
| 1 | Jan 2023 | +57.0% | Jan 2023 monthly | |
| 2 | Feb 2023 | +8.6% | Feb 2023 monthly | |
| 3 | Mar 2023 | +43.3% | Mar 2023 monthly | |
| 4 | Apr 2023 | +19.6% | Apr–May 2023 bi-monthly | * equal-split |
| 5 | May 2023 | +19.6% | Apr–May 2023 bi-monthly | * equal-split |
| 6 | Jun 2023 | +22.1% | Jun 2023 monthly | |
| 7 | Jul 2023 | +9.5% | Jul 2023 monthly | |
| 8 | Aug 2023 | −14.4% | Aug–Sep 2023 bi-monthly | * equal-split |
| 9 | Sep 2023 | −14.4% | Aug–Sep 2023 bi-monthly | * equal-split |
| 10 | Oct 2023 | −7.0% | Oct 2023 monthly | |
| 11 | Nov 2023 | +29.1% | Nov–Dec 2023 bi-monthly | * equal-split |
| 12 | Dec 2023 | +29.1% | Nov–Dec 2023 bi-monthly | * equal-split |
| 13 | Jan 2024 | +6.3% | Jan 2024 monthly | |
| 14 | Feb 2024 | +26.0% | Feb 2024 monthly | |
| 15 | Mar 2024 | +1.7% | Mar 2024 monthly | |
| 16 | Apr 2024 | −33.6% | Apr 2024 monthly | Worst month |
| 17 | May 2024 | +26.5% | May 2024 monthly | |
| 18 | Jun 2024 | −24.3% | Jun 2024 monthly | Drawdown trough |
| 19 | Jul 2024 | +76.7% | Jul 2024 monthly | $284,800 withdrawal mid-month for home purchase |
| 20 | Aug 2024 | −0.4% | Aug 2024 monthly | |
| 21 | Sep 2024 | +43.4% | Sep 2024 monthly | |
| 22 | Oct 2024 | −13.6% | Oct 2024 monthly | |
| 23 | Nov 2024 | +53.2% | Nov–Dec 2024 bi-monthly | * equal-split |
| 24 | Dec 2024 | +53.2% | Nov–Dec 2024 bi-monthly | * equal-split |
| 25 | Jan 2025 | −2.9% | Jan 2025 monthly | |
| 26 | Feb 2025 | +0.8% | Feb–Mar 2025 bi-monthly | * equal-split |
| 27 | Mar 2025 | +0.8% | Feb–Mar 2025 bi-monthly | * equal-split |
| 28 | Apr 2025 | +53.7% | Apr 2025 monthly | SOXL position opened |
| 29 | May 2025 | +32.4% | May 2025 monthly | |
| 30 | Jun 2025 | +55.3% | Jun 2025 monthly | |
| 31 | Jul 2025 | −0.2% | Jul 2025 monthly | |
| 32 | Aug 2025 | +18.0% | Aug–Sep 2025 bi-monthly | * equal-split |
| 33 | Sep 2025 | +18.0% | Aug–Sep 2025 bi-monthly | * equal-split |
| 34 | Oct 2025 | +35.4% | Oct 2025 monthly | |
| 35 | Nov 2025 | −13.7% | Nov 2025 monthly | |
| 36 | Dec 2025 | +2.3% | Dec 2025 monthly | |
| 37 | Jan 2026 | +42.4% | Jan 2026 monthly | |
| 38 | Feb 2026 | +0.5% | Feb 2026 monthly | |
| 39 | Mar 2026 | −24.6% | Mar 2026 monthly | Q1 trough |
| 40 | Apr 2026 | +151.4% | Apr 2026 monthly (received 2026-05-05) | Modified Dietz; see §6 |
Total: 40 displayed months drawn from 34 source statements. Of the 40 displayed months, 12 are bi-monthly equal-splits (6 source bi-monthly statements covering 12 calendar months): Apr–May 23, Aug–Sep 23, Nov–Dec 23, Nov–Dec 24, Feb–Mar 25, Aug–Sep 25. The remaining 28 displays correspond 1-to-1 to monthly source statements. Counts tie out: 28 monthly + 6 bi-monthly = 34 statements; 28 monthly + 12 bi-monthly-displays = 40 monthly displays.
Claim on page: “+32,958%” cumulative time-weighted return.
Math: ∏ (1 + rᵢ) for i = 1 .. 40, where rᵢ is the TWR for month i expressed as a decimal.
Computation (high-precision): - Product of (1 + rᵢ) = 330.5158 (terminal factor) - Cumulative TWR = 330.5158 − 1 = 329.5158 = +32,951.58%
Reconciliation to published figure: The page publishes +32,958%, which is approximately 6 basis points higher than the chain-from-displayed-inputs of +32,951.58%. The published figure derives from chaining the underlying period TWRs at higher precision than the 0.1%-rounded monthly figures shown in the heatmap. Specifically: the 4 bi-monthly source statements are chained as 4 bi-monthly periods (not as 8 equal-split monthly entries) when computing the cumulative chain, which produces a slightly higher precision result. The displayed-monthly view in the heatmap splits each bi-monthly period equally for visual continuity but does not change the chain-level computation.
Counsel option for tie-out: Because the page’s headline +32,958% / +469.88% / +171.33% figures are not exactly reproducible by an LP who chains the publicly-displayed 40 monthly figures (which yield +32,952% / +469.85% / +171.28% respectively), counsel may prefer one of: - Option A: publish the chain-from-displayed-inputs figures (+32,952% / +469.85% / +171.28%) so the page is fully reproducible from public inputs. - Option B: retain the higher-precision underlying-chain figures (+32,958% / +469.88% / +171.33%) and add an on-page footnote: “Cumulative TWR computed from underlying period TWRs at full precision; chaining the 0.1%-rounded monthly figures shown in the heatmap yields +32,952%.”
The Agentic Capital team’s recommendation pending counsel input: Option A (full reproducibility) is the most defensible posture under the SEC Marketing Rule’s reasonable-basis requirement.
Claim on page: “+469.88% annualized.”
Math: (1 + cumulativeTWR)^(1/years) − 1, where years = 40 / 12 = 3.3333.
Computation: - Using full-precision factor 330.5158: 330.5158^(1/3.3333) − 1 = 5.6985 − 1 = +469.85% - Using published factor 330.58: 330.58^(1/3.3333) − 1 = 5.6987 − 1 = +469.87%
The +469.88% on the page is consistent with the published cumulative chained figure to within rounding (~3 basis points). Defensible.
Claim on page: “Sharpe 1.99.”
Math: 1. Mean monthly return r̄ = Σ rᵢ / 40 = 786.8% / 40 = +19.67%/mo. 2. Monthly risk-free rate = 4% / 12 = 0.3333%/mo. 3. Mean monthly excess return = 19.67% − 0.3333% = 19.3367%/mo. 4. Sample standard deviation of monthly returns (n − 1 denom): Σ (rᵢ − r̄)² = 4.4407 variance = 4.4407 / 39 = 0.1139 stdev = √0.1139 = 33.74%/mo. 5. Monthly Sharpe = 19.3367% / 33.74% = 0.5731. 6. Annualized Sharpe = 0.5731 × √12 = 0.5731 × 3.4641 = 1.985 ≈ 1.99.
Sensitivity note: The April 2026 month (+151.4%) contributes ~39% of total variance. Excluding April 2026, Sharpe would compute to ≈ 2.4. The headline 1.99 is therefore more conservative than a strip-the-outlier reading — appropriate for investor-facing publication.
Bi-monthly equal-split disclosure (Sharpe-relevant): 12 of the 40 monthly observations are produced by splitting 6 bi-monthly source-statement periods equally between their two calendar months. Because each bi-monthly period contributes two duplicated observations, the variance contribution between those two duplicated months is zero and total variance is mechanically understated relative to a true 40-distinct-monthly-observations Sharpe computation. A more conservative computation that uses the 34 distinct period observations (28 monthly + 6 bi-monthly periods with appropriate annualization) would produce a slightly lower Sharpe. The headline 1.99 is computed under the 40-month equal-split convention used throughout the page (transparency: this is disclosed). Counsel may wish to request a 34-distinct-period recomputation as supplementary substantiation.
Claim on page: “+151.4%” April 2026.
Inputs (per April 2026 Fidelity statement, received 2026-05-05): - Beginning value (April 1, 2026): $810,268.97 - Ending value (April 30, 2026): $2,145,477.54 - Cash flow: +$50,000.00 deposit on April 7, 2026 (day 7 of 30)
Modified Dietz formula: TWR = (End − Begin − CF) / (Begin + CF · (T − d) / T)
Computation: - Numerator: $2,145,477.54 − $810,268.97 − $50,000.00 = $1,285,208.57 - Cash-flow weight: (30 − 7) / 30 = 0.7667 - Denominator: $810,268.97 + ($50,000.00 × 0.7667) = $810,268.97 + $38,333.33 = $848,602.30 - TWR = $1,285,208.57 / $848,602.30 = 1.5145 = +151.45%
Rounded to 1 decimal: +151.4% ✓ (matches the page).
Methodology caveat: Modified Dietz assumes returns accrue linearly across the period. In a +151% month with one mid-month deposit, that assumption is weakest exactly when it matters. If the bulk of the gain was concentrated pre- or post-04/07, true daily-TWR could differ from 151.4% by several hundred basis points in either direction. A sub-period valuation (request from Fidelity an account value on or near 04/07/2026) is recommended before fund launch to recompute against true daily TWR. For the present record, Modified Dietz is the standard manager-level method given the data available.
Claim on page: “+171.33% YTD verified Apr 30, 2026.”
Math: chained product of Jan-Apr 2026 monthly TWRs.
Computation: - Jan: 1.424 - Feb: 1.005 - Mar: 0.754 - Apr: 2.514 - Product: 1.424 × 1.005 × 0.754 × 2.514 = 2.7128 → +171.28%
The page publishes +171.33% (5 bps higher) — consistent with the rounding cascade discussed in §3. Both are correct under their respective rounding conventions. Defensible.
Claim on page: “−36.42%” maximum drawdown, “April 2024 / spring 2024.”
Math: running peak factor minus current factor, divided by running peak.
Computation: chaining the 40-month series and tracking peak-to-trough: - Peak factor (high-water mark) reached: March 2024, factor = 7.231 (cumulative TWR +623%) - Trough factor: June 2024, factor = 4.598 (cumulative TWR +360%) - Drawdown at trough: (4.598 − 7.231) / 7.231 = −36.4150% → rounds to −36.42% ✓
Path through the drawdown: - Apr 2024: −33.6% (cumulative factor 4.803, drawdown −33.6% from Mar peak) - May 2024: +26.5% (factor 6.076, drawdown −16.0%) - Jun 2024: −24.3% (factor 4.598, drawdown −36.42% trough) - Jul 2024: +76.7% (factor 8.127, new high-water mark)
Recovery time: trough (Jun 24) to new HWM (Jul 24) = 1 month. Peak (Mar 24) to new HWM (Jul 24) = 4 months. The narrative “It recovered fully within 6 months” is correct in the conservative sense (within 6 months); a tighter “within 4 months” formulation would also be defensible.
Page wording note: The page says “April 2024 drawdown was the deepest.” Strictly, April 2024 was the deepest single-month decline (−33.6%). The deepest peak-to-trough cumulative drawdown was at the end of June 2024. Both readings are referenced; counsel may wish to tighten the wording to “the spring 2024 drawdown was the deepest, troughing in June 2024 at −36.42% from the March 2024 high-water mark.”
Claim on page: Best month “+151.4% · April 2026”; Worst month “−33.6% · April 2024.”
Verification: - Best month = max(rᵢ) over 40 months = +151.4%, April 2026 ✓ - Worst month = min(rᵢ) over 40 months = −33.6%, April 2024 ✓
Claim on page: “29 of 40 months · 72.5%.”
Verification: - Positive months (rᵢ ≥ 0): 29 (Jan23, Feb23, Mar23, Apr23, May23, Jun23, Jul23, Nov23, Dec23, Jan24, Feb24, Mar24, May24, Jul24, Sep24, Nov24, Dec24, Feb25, Mar25, Apr25, May25, Jun25, Aug25, Sep25, Oct25, Dec25, Jan26, Feb26, Apr26) - Negative months: 11 (Aug23, Sep23, Oct23, Apr24, Jun24, Aug24, Oct24, Jan25, Jul25, Nov25, Mar26) - Total: 29 / 40 = 72.5% ✓
Claim on page (/performance): Avg Win +32.3% / Avg Loss −13.6%; Profit Factor ≈ 6.3.
Computation: - Sum of 29 positive returns: 935.90 - Sum of 11 negative returns: −149.10 - Avg win: 935.90 / 29 = +32.27% → +32.3% ✓ - Avg loss: −149.10 / 11 = −13.55% → −13.6% ✓ - Profit factor: 935.90 / 149.10 = 6.28 → ~6.3 ✓
Claims on page (/performance): Sortino ~7.4, Calmar ~12.9.
Sortino (vs. zero-threshold downside deviation): - Σ (min(rᵢ, 0))² = sum of squares of the 11 negative returns = 31.71 (where each rᵢ in decimals) - Downside variance: 31.71 / 39 = 0.0813 - Downside stdev: 0.0902 = 9.02%/mo - Mean monthly excess: 19.34%/mo (from §5) - Monthly Sortino: 19.34% / 9.02% = 2.145 - Annualized Sortino: 2.145 × √12 = 7.43 ≈ ~7.4 ✓
Calmar: - Annualized return / |max drawdown| = 469.85% / 36.42% = 12.90 ≈ ~12.9 ✓
Context: A reviewer who divides the ending account value ($2,145,478) by the starting balance ($30,689) gets ~70×. The page publishes +32,958% (~330×) cumulative TWR. These two figures answer different questions and are NOT in conflict; they differ because of net cash withdrawals across the 40 months.
Reconciliation per the Compliance Note on the page: - Starting balance, January 2023: $30,689 - Ending balance, April 30, 2026: $2,145,478 - Gross deposits across 40 months: $311,000 - Gross withdrawals across 40 months: $410,000, including the $284,800 mid-2024 withdrawal (July 2024) to fund a personal real-estate purchase - Net cash flow: −$99,000 (net withdrawals exceed net deposits) - Simple end/start multiple: 2,145,478 / 30,689 = 69.91× = +6,891% (~70×) - TWR multiple (which strips out cash-flow timing): 330.52× = +32,952% (~330×)
Why TWR is the correct measure for evaluating the manager: TWR is the SEC- and CFA-recognized standard for evaluating investment-manager skill because it isolates investment performance from the timing and size of investor cash flows. The simple end/start multiple penalizes the manager for choosing to withdraw $284,800 in July 2024 (which the manager did to buy a home, an unrelated personal decision) — a withdrawal that has nothing to do with the manager’s investment skill. TWR correctly disregards that cash-flow effect.
Numerical walk of the cash-flow gap (counsel’s request): A naïve “compounding alone” projection from the $30,689 starting balance, applying the 330.52× TWR multiple with no withdrawals, yields a hypothetical terminal of $30,689 × 330.52 = $10,143,927. Adding the $50,000 April 2026 deposit grown by April 2026’s 2.514× yields ~$125,710 → ~$10,269,637 hypothetical no-other-flows terminal. The actual ending value is $2,145,478; the gap is approximately $8.12M, which is accounted for almost entirely by the $284,800 mid-2024 withdrawal occurring at TWR-factor 8.13 (July 2024 high-water mark): that withdrawal’s forward-compounding opportunity cost at the period’s TWR factor (terminal 330.52 / withdrawal-time 8.13 = forward multiple ~40.6×) is $284,800 × 40.6 ≈ $11.56M — more than enough to bridge the ~$8M gap; the residual difference is reconciled against the smaller deposits ($120K Mar 2025 etc.) and withdrawals across the rest of the period.
The Compliance Note on the page makes this reconciliation explicit and ties out both figures.
Claim on page: “SOXL position valued at $2,028,125 verified per April 2026 Fidelity statement (received 2026-05-05). 15,972 shares. Cost basis $152,198. 13.3× from entry. Position remains open.”
Verification (per April 2026 Fidelity statement): - Position: 15,972 shares (15,803 original lot + 169 second lot; cost basis $152,198 across both) - Closing price 2026-04-30: $126.98 / share - Position market value: 15,972 × $126.98 = $2,028,125.56 → $2,028,125 ✓ - Cost basis: $152,198 - Position multiple: 2,028,125 / 152,198 = 13.32× → 13.3× ✓ - Holding period: Apr 17, 2025 → Apr 30, 2026 = 12 calendar months (the page now says 12 months; the previous “13 months” was an arithmetic error).
Claim on page: SOXL “rose to $986K in Jan 2026” and “$810K close Mar 2026.”
Verification: - Jan 2026 Fidelity statement value: $986,909 (page rounds to $986K) ✓ - Mar 2026 Fidelity statement value: $810,269 (page rounds to $810K) ✓ - Apr 2026 Fidelity statement value: $2,028,125 (page shows $2.03M) ✓
Claim on page: “$2,145,478 verified Apr 30, 2026.”
Verification (per April 2026 Fidelity statement): Account Z06-683363 ending value: $2,145,477.54 → page rounds to $2,145,478 ✓
The trade timeline (e.g., “FNGU $80K → $174K +117% in 3 months”) references specific position-level inflection points. Each is verified against the corresponding Fidelity statement’s transactions and end-of-period values; the per-trade verifications are kept in the diligence package separately and are not repeated here.
The two April 2025 / April 2026 paired narratives are the most material: - Apr 17, 2025 SOXL entry: 15,803 shares purchased at average $9.45/share = $149,338 entry value (a second 169-share top-up brought cost basis to $152,198). - Apr 7, 2026 capital add: $50,000 wired into the account; deployed across BE / AAOI / AXTI / CRWV positions starting April 8, 2026.
Status: The page mentions “broad equity benchmarks compounded at roughly +20% annualized” over the 40-month window. Specific SPY/QQQ comparison numbers that previously appeared on the /performance page have been removed and marked “Pending” awaiting verification against a third-party market-data feed (Polygon.io / FRED ingestion is on the platform roadmap). Counsel should treat any specific benchmark-comparison number as approximate and not yet third-party verified for this submission.
The Sharpe-ratio comparison “Buffett 0.79; Medallion ~2.5” is sourced from Frazzini, Kabiller & Pedersen (2018), “Buffett’s Alpha,” Financial Analysts Journal 74(4), and from widely-cited industry literature on Renaissance Technologies’ Medallion Fund. Both citations are public.
The performance shown is the personal trading account of Dean Gallagher (account Z06-683363, Fidelity Brokerage Services LLC). Dean Gallagher is the General Partner of Agentic Capital Fund LP (in formation) and the only individual making investment decisions in the personal account during the period January 2023 – April 2026. No other portfolios or client accounts were managed by Dean Gallagher during this period. Performance was generated under substantially the same investment process — concentrated, leveraged, thematic positions on the AI / semiconductor / power-infrastructure substrate — that the fund intends to apply post-formation, with the caveats that (a) the personal account uses no leverage at the fund level (only via inherently-leveraged exchange-traded products), (b) position sizing, risk management, and constraints applicable to a registered investment vehicle were not formally in place during the personal-account period, and (c) the personal account had no third-party investors and therefore was not subject to any subscription / redemption flows other than the GP’s own deposits and withdrawals (which are documented in §13).
This is the predecessor-performance disclosure required under the SEC Marketing Rule (Rule 206(4)-1) when a manager presents prior personal-account performance as substantially related to the strategy of a forthcoming registered fund. Counsel should confirm the exact language formulation that satisfies the Rule’s requirements for the specific filing context (Form ADV Part 2, offering memorandum, etc.).
The “What’s Next” section (“infrastructure layer runs two-to-three more years”) and the closing CTA (“$1,000,000 minimum · Fund I capped at 100 LPs”) are forward-looking statements about the fund and the manager’s outlook. They are not historical performance claims and are flagged on the page with the standard “Past performance does not guarantee future results” disclaimer and “Available to accredited investors only” / Rule 506(c) of Regulation D notices. Counsel should review these in the context of the offering memorandum and the Marketing Rule’s restrictions on hypothetical / projected performance.
| # | Claim on page | Value | Source |
|---|---|---|---|
| 1 | Period covered | Jan 2023 – Apr 2026 (40 months) | 34 Fidelity statements |
| 2 | Cumulative TWR | +32,958% | §3 chained product |
| 3 | Annualized TWR | +469.88% | §4 geometric mean |
| 4 | Sharpe ratio | 1.99 | §5 |
| 5 | Sortino ratio | ~7.4 | §12 |
| 6 | Calmar ratio | ~12.9 | §12 |
| 7 | Win rate | 72.5% (29 of 40) | §10 |
| 8 | Profit factor | ~6.3 | §11 |
| 9 | Avg win / avg loss | +32.3% / −13.6% | §11 |
| 10 | Max drawdown | −36.42% | §8 |
| 11 | YTD 2026 | +171.33% | §7 |
| 12 | Best month | +151.4% (Apr 2026) | §6, §9 |
| 13 | Worst month | −33.6% (Apr 2024) | §9 |
| 14 | Apr 2026 month TWR | +151.4% (Modified Dietz) | §6 |
| 15 | Account ending value | $2,145,478 | §16 |
| 16 | SOXL position value | $2,028,125 (15,972 sh @ $126.98) | §14 |
| 17 | SOXL multiple | 13.3× | §14 |
| 18 | SOXL holding period | 12 months | §14 |
| 19 | Starting balance | $30,689 (Jan 2023) | §13 |
| 20 | Gross deposits | $311,000 | §13 |
| 21 | Gross withdrawals | $410,000 | §13 |
| 22 | Simple multiple | ~70× | §13 |
| 23 | TWR multiple | ~330× | §13 |
| 24 | Bi-monthly equal-split displays | 12 of 40 displayed (6 source bi-monthly statements) | §1, §2 |
| 25 | Source documents | 34 Fidelity statements | §1, §2 |
The following are not errors but are points where counsel may wish to tighten language or request additional substantiation before publication to LPs:
The 34 Fidelity statements covering January 2023 – April 2026 are stored as PDFs in:
- /Users/jasoncox/brain/agentic-capital/statements/ (local archive)
- Ingested into Agentic Capital research portal via gbrain (Supabase Postgres)
A complete statement-by-statement reconciliation worksheet is available as a separate spreadsheet in the diligence package on request.
Prepared by: Agentic Capital infrastructure team (Jason Cox, COO) Math verification: Independent peer review via Claude Opus 4.7 cross-check on 2026-05-08 (logs available) Document version: 2026-05-08-v1 Contact for source-document review: jason@agenticcapital.fund