Ambrus Capital Management / Ambrus Volatility Fund: Hedge Fund Diligence (First Pass)
Ambrus Capital Management / Ambrus Volatility Fund: Hedge Fund Diligence (First Pass)
Prepared 2026-10-20 from the eight files in sources/. The newest evidence is the January 2026 DDQ and the December 2025 investor deck, so everything here is roughly ten months old. No 2026 performance, AUM or team data was available. This is a first-pass note, not full investment diligence, not ODD, and not a recommendation.
Source key: [DDQ26] DDQ, Jan 2026 · [Deck25] Investor Presentation, Dec 2025 · [PPM25] Portfolio Supplement, Liberty Investment Park SPC / Ambrus Volatility Fund SP, Feb 2025 · [Sub25] SP Subscription Agreement (KKWC markup, 28 Feb 2025) · [TS21] Tear Sheet, Jun 2021 · [Memo21] Tear Sheet Memo, Jun 2021 · [Pitch21] Pitch deck, Jun 2021 · [Shock21] Shock & Slide Levels, 8 Sep 2021
1. First Impression
What it is. Ambrus runs a "carry-neutral" tail-risk fund. It buys deep out-of-the-money VIX calls, S&P 500 puts and sector-ETF puts with 1 to 18 month tenors, and it pays for that bleed with a short-horizon "flow trading" sleeve. The flow sleeve trades US equities and options around price-insensitive flows such as ETF rebalances, vol-target funds, dealer gamma hedging and retail thematics [DDQ26 §2.1.2; Deck25 pp.12, 24–28]. The fund is built to pay off when VIX is above about 40 and accelerating. It does not try to protect against slow, orderly declines [DDQ26 §2.1.6–2.1.7; Deck25 p.18]. AUM was about $437.6M firm-wide, including $395.6M in the master fund, as of 1 Jan 2026 (unaudited) [DDQ26 §1.2.13].
Why it's worth a meeting. The live record from June 2022 to November 2025 matches the mandate. Net returns were roughly flat for 26 months, from June 2022 to July 2024 (about +1% cumulative). The fund then had two spike months: +11.89% in August 2024 and +25.21% in April 2025 [Deck25 p.2]. Over the live period the worst monthly loss was -2.86%, and my computed peak-to-trough drawdown is about -3.6%. That combination of near-zero bleed after fees and positive convexity is uncommon in long-vol funds. Service providers are institutional: NAV Consulting, EY and Richey May, Goldman Sachs, KKWC, Maples and Carey Olsen.
What is unclear, and the biggest gaps:
- Track-record restatement. The 2025 deck's 2020–May 2022 history is labelled hypothetical. It includes +336.95% for March 2020. It also contradicts the actual SMA figures in the June 2021 tear sheet for the same months (see §4). The verified live record is only about 3.5 years and includes two payoff events.
- Strategy evolution. In 2021 the stated sleeves were dispersion, VIX contango and capped short vega, with tails held only at 1–2 month tenors. In 2025 the fund describes itself as "no short optionality", with 1–18 month tails and a flow book [Pitch21 pp.20, 23, 27; Shock21; Deck25 p.27]. The current process is about three years old.
- Realized vs. modeled convexity. The stress tables, already halved, imply much larger gains at VIX 40–45 than the fund delivered in August 2024 or April 2025 (see §4).
- Capacity vs. growth. The deck calls the flow edge "capacity constrained", yet master AUM is about $396M. Ambrus also runs a friends-and-family fund, Medici, that trades only the bleed-mitigation strategies [DDQ26 §1.2.11; Deck25 pp.7, 8, 12].
- Thin institution. The firm has five employees across four cities. The two co-CIOs own 79% and also run risk. The CCO is also COO and a 9% owner. There is no SOC 1 report, internal audit, board or standalone crime policy [DDQ26 §§1.2, 4.1.1, 6.3–6.7, 10.1, 12].
2. Team
| Person | Role (per sources) | Ownership [DDQ26 §1.2.8] | Notes |
|---|---|---|---|
| Kris Sidial | Co-CIO / PM | 39.6% | Previously at BMO (exotic derivatives) and Chimera/Xanthus (prop) [DDQ26 §1.2.9] |
| William Wise | Co-CIO / PM | 39.6% | Independent trader; built the flow-type signals [DDQ26 §1.2.9] |
| Michael Perry | CCO [DDQ26 §10.1] / COO [Deck25 p.10] | 9.1% | Technology background at private funds |
| Sal Abbasi | "Managing Partner" [Deck25 p.2]; "Partner" [Deck25 p.10]; "advisory capacity" [DDQ26 §1.1.4] | 6.8% (2.8% direct + 4.0% via trust) | Ex-Citadel quant credit tech, Morgan Stanley |
| Matt Sutton | Execution trader (joined Sep 2024) | 0% | Authorized to place orders [DDQ26 §9.2] |
| Aditya Nemani | Quant researcher (joined Feb 2025) | — | Ex-Walleye, Citadel [Deck25 p.10] |
| Oceanfront Consulting LLC | — | 4.9% | Beneficial owner not identified |
Facts. There are 5 employees: 2 PMs, 1 analyst, 1 junior trader and 1 administrator [DDQ26 §1.2.1–1.2.5]. They work in Tampa, Hauppauge NY, San Diego and Boston, and coordinate by video during the trading day [DDQ26 §1.1.4]. Paul Gustafson left in November 2025; his role is not stated [DDQ26 §1.2.6]. Bret Lannert was a named partner in 2021 [Pitch21 p.5; TS21] but is absent from 2025 materials. His departure falls outside the DDQ's two-year leaver window, so it is undocumented in the sources. The firm holds two $2M key-person life policies, one on each co-CIO [DDQ26 §12].
Bio drift to verify, not a finding. Sidial was described in 2021 as having "5 years" of institutional experience, including managing a "personal book over $1.5 billion" and "finishing his masters at UPenn" [Pitch21 p.5; TS21]. By 2025 the claim was "studied at University of Pennsylvania (MS program)" with a Long Island University BS [Deck25 p.10]. Wise went from "independent trader for over six years" and founder of an online trading chat/platform [Pitch21 p.5; TS21] to "highly successful independent trader for more than a decade" [Deck25 p.2]. His independent-trading record is the basis for the flow edge, and no evidence of it appears in the sources.
Read. The edge appears tied to the two co-CIOs: Wise for the flow signals and Sidial for the tail book and monetization. Recent hires add research and execution depth but have no ownership. The "50% quant / 50% discretionary" monetization process [Deck25 p.21] and the roughly 25% discretionary trade share [DDQ26 §2.1.2] put key-person risk squarely on the two founders.
3. Terms, Liquidity and Alignment
| Term | Deck25 (p.2, p.36) | PPM25 (Offshore SP, Class A) | Note |
|---|---|---|---|
| Mgmt fee | 2% | 2% p.a., quarterly (in arrears for SP; in advance for domestic LP) | Can be rebated to insiders; part goes to Kroma (placement agent) [PPM25 pp.12, 15–16] |
| Incentive | 20% | 20%, annual and on redemption; HWM via per-series Loss Recovery Account | No hurdle; no clawback disclosed |
| Minimum | $1,000,000 | $275,000 (floor $100,000) | Conflict; deck likely refers to the domestic LP |
| Liquidity | Monthly, 30 days' notice | Monthly, 30 days' notice; 100% paid within 30 days of NAV | — |
| Lock-up | "None" | 2% redemption fee within 12 months (soft lock) | Conflict; confirm by vehicle |
| Gate / suspension | — | Refers to base Memorandum, which is not in sources/ | Gap |
| Side letters | — | Contemplated [Sub25 §16] | Existence and terms unknown |
| Expenses | — | Broad pass-through: research and data, IT, OMS and risk systems, regulatory filings, Director fees, organizational and offering costs [PPM25 pp.17–21] | Total expense ratio not disclosed |
Alignment, on available information.
- The GPs say they have personal capital in the fund, and Sidial says he has invested "50% of my investable capital" [Deck25 p.34]. No dollar amount or percentage of AUM is disclosed.
- The fee structure fits the strategy unevenly. A fund designed to be flat in normal years still earns 2% a year. The flow sleeve therefore has to cover bleed plus fees, so being "flat net" requires about +2–3% gross.
- With annual crystallization and no clawback, a spike year pays the GP 20% of a gain that is supposed to be monetized and given back as the hedge resets. The HWM limits but does not eliminate this.
- The IMA can be terminated on 75 days' notice [PPM25 p.12]. No key-person redemption right is disclosed.
4. Performance and Drawdowns
Reported net returns (%) [Deck25 p.2]. The deck states these are derived from administrator NAV, net of 2/20.
| Year | Reported YTD | Status | Comment |
|---|---|---|---|
| 2020 | 269.27 | Hypothetical | Mar +336.95. Monthly figures compound to about 336%, not 269%, and do not reconcile |
| 2021 | 2.21 | Hypothetical (actual SMAs were trading) | Monthly figures compound to about 2.8%, not 2.21%. Conflicts with TS21 (below) |
| 2022 | -1.73 | Hypothetical Jan–May; live from launch in May 2022 | Live Jun–Dec: -0.5% |
| 2023 | 1.71 | Live | — |
| 2024 | 13.29 | Live | Aug +11.89 (VIX spike to about 45) |
| 2025 (Jan–Nov) | 28.37 | Live | Apr +25.21 (VIX to about 60) |
Live period, Jun 2022 to Nov 2025 (42 months, my calculation from the deck's monthly figures): about +47% cumulative, about 11.7% annualized, about 15% annualized volatility, about -3.6% maximum drawdown, 28 of 42 months positive. Excluding August 2024 and April 2025, the other 40 months compound to about +5.0% (about 1.5% a year). Those two months account for essentially all of the return. The deck's own figures (rolling 24-month return 45.78%, rolling 24-month max drawdown -2.64%) reconcile with my numbers for Dec 2023 to Nov 2025. December 2025 is missing from the table.
Restatement and conflict.
- 2021. The June 2021 tear sheet reports actual SMA "Convexity" returns for Jan–May 2021 of +3.20, -2.57, -2.21, -1.31 and -1.10 (YTD -4.02%), net of commissions but gross of fees [TS21]. For the same months, the 2025 deck shows hypothetical returns of -3.46, +5.36, -0.53, -1.22 and -1.20 [Deck25 p.2]. The two series differ in sign in February. The deck's footnote calls the Jan 2020 to May 2022 history hypothetical even though the DDQ says SMAs traded from January 2021 [DDQ26 §1.1.1].
- 2020. The tear sheet shows no 2020 track record ("the firm was focused on capital raising"). The pitch deck says the group had one account in Jan–Feb 2020 [Pitch21 p.28]. The +336.95% for March 2020 is therefore a backfilled simulation. It sits close to the 326.4% "‑30% month" scenario in [Shock21], which was a hypothetical repricing of September 2021 positions.
- 2019. A "Yield" composite of +45.11% from team-managed customer accounts appears in 2021 materials [TS21; Pitch21 p.28] and was dropped later.
- The deck's footnote says the figures come from administrator NAV and that the pre-May 2022 figures are hypothetical. Both cannot be true for 2020–2021.
Realized vs. modeled convexity. The deck's stress tables cover January to December 2024 only, so they are stale in a December 2025 deck. They were already "cut in half to be conservative" [Deck25 pp.4, 38]. For August 2024 positioning they show +121.66% at VIX 40 in a COVID-type path and +212%/+269% at VIX 40/45 in a 2015-type path. Actual August 2024 was +11.89%, and actual April 2025, with VIX above 40 for about a week, was +25.21%. The deck's caveat is that the tables assume full crash dynamics, not "a momentary touch" [Deck25 p.4], and both 2024–25 spikes faded quickly. That explanation is plausible but unverified. The tables also have internal oddities: for November 2024 positioning, the Volmageddon payoff at VIX 30 (681.58%) exceeds the payoff at VIX 40 (596.82%).
Drawdowns. The live period has no meaningful drawdown. That fits the design, but the sample includes no sustained high-vol regime and no flow-sleeve failure. 2022 was a slow decline, which the fund says it does not cover, and the year was about flat [Deck25 pp.18, 22]. Which sleeve produced the P&L month by month is not disclosed.
5. Strategy and Edge
| Element | Claim | Evidence in sources | Still to diligence |
|---|---|---|---|
| Tail sleeve | Cheap convexity via vol-surface relative value; tenors spread across 1–18 months to manage sequence risk; execution below the inside bid using algorithms and dealer inventory [Deck25 pp.14–17] | One SPY put example (Oct 2023, $0.13) and a 40,000-contract fill at $0.01 against a 0.00/0.05 NBBO | Tail budget per month; vega and premium by tenor; attribution of the Aug 2024 and Apr 2025 gains |
| Monetization | Model bands plus discretionary overlay, about 50/50 [Deck25 pp.20–21] | Narrative only | Actual monetization logs for Aug 2024 and Apr 2025; how much P&L was given back |
| Flow sleeve | Signals that detect price-insensitive flow; about 10 bps of AUM at risk per trade; holds seconds to days; "uncorrelated" [DDQ26 §2.1.2; Deck25 pp.25–28] | Illustrative trade types only; the JPM collar is a public example, not an Ambrus P&L | Standalone sleeve returns, hit rate, turnover and capacity curve; correlation in vol spikes; signal decay as AUM grew |
| Uniqueness | "The only firm to bring the prop-style, alpha-driven approach of carry-neutral tail risk hedging into the hedge fund arena" [Deck25 p.5] | Marketing claim | Peer comparison with other bleed-offset long-vol managers |
Edge read. The tail sleeve's design (tenor laddering, avoiding "dead vol", discretionary monetization) is coherent and close to standard practice. Whether the structure works depends on the flow sleeve's ability to cover carry plus fees at roughly $400M. The sources assert this but never show the sleeve on its own. The 2021 memo itself said "it is unrealistically optimistic to think that tail risk hedges can make money systematically over time" [Memo21 p.5], while the 2025 deck markets "crash protection with zero bleed" [Deck25 p.12]. The live record supports "low bleed" so far. It does not yet separate skill from a favourable 2023–25 environment for short-horizon equity flow trading.
Mandate breadth. The PPM allows any strategy and any instrument, including written options, private placements and swaps, "without notice to, or the consent of, investors" [PPM25 pp.4–5]. That contrasts with "no short optionality" in the deck and with the DDQ's statement that only listed instruments are used [DDQ26 §3.5.1].
6. Portfolio, Risk and Capacity
- Leverage: "The Fund does not use leverage." Active trades typically use 15% of capital or less, and total usage "rarely" exceeds 30% [DDQ26 §3.1].
- Controls: about 10 bps of AUM maximum loss per flow trade; tiered monthly drawdown scaling; alpha-decay monitoring; stops or offsetting options on short positions, which are closed intraday if unhedged [DDQ26 §3.2.2; PPM25 pp.3–4]. The thresholds themselves are not disclosed.
- Activity: about 20,000 contracts a day, about 90% electronic, OMS is SpiderRock [DDQ26 §3.5].
- Liquidity and valuation: listed instruments only; exchange or third-party prices verified by the administrator; no Level 3 [DDQ26 §3.3]. Deep OTM options are wide in calm markets.
- Risk governance: the co-CIOs own day-to-day risk, and the CCO (Perry) provides "independent oversight" [DDQ26 §4.1.1]. Perry is also COO and an owner, so there is no independent risk function.
- Capacity: the strategy operates "within defined capacity parameters" [DDQ26 §2.1.9], but the number is not given. The deck stresses that the edges are capacity constrained [Deck25 pp.7–8, 12].
- Allocation conflict: the Medici fund (friends and family, 27.2M) and Sundial ($2.2M) are traded separately [DDQ26 §§1.2.11, 3.5.7]. Allocation of scarce flow signals across accounts needs to be checked.
7. Key Counterparties and ODD Snapshot
These are disclosed facts only, not ODD.
| Function | Provider |
|---|---|
| Vehicle | Master: Ambrus Volatility Master Fund LP (Cayman); feeders: Ambrus Volatility Fund LP (Delaware) and Ambrus Volatility Fund SP, a segregated portfolio of Liberty Investment Park SPC, a third-party Cayman platform [PPM25 p.1] |
| Administrator | NAV Fund Administration Group (NAV Consulting); also AML officer [DDQ26 §11.2] |
| Auditors | EY Ltd. (Cayman SP); Richey May (master and domestic) |
| Prime brokers / custody | Goldman Sachs, Clear Street, Velocity Clearing [PPM25 pp.22–23]. The deck also lists Lightspeed, and lists CIBC as "Custodian" (CIBC is the bank per [DDQ26]) |
| Legal | KKWC (US), Maples (master), Carey Olsen (SP) |
| Compliance / finance | Pillar Compliance Group; Covalent (outsourced CFO) |
| IR / distribution | Kroma Capital Partners, paid from the management fee [PPM25 p.12] |
| Regulatory | The Ambrus Group LLC is SEC-registered; the IM is a relying adviser. A Florida OFR exam began 8 May 2024, outcome not stated [DDQ26 §§1.1.3, 1.2.12] |
| Controls / insurance | No SOC 1, internal audit or board. D&O/E&O aggregate limit is $1M (Markel). No standalone crime policy. No in-house IT [DDQ26 §§6.7, 12, 13] |
Gaps: SPC Directors are not identified in the supplement. The base Memorandum is not in the sources. Valuation policy, audited financials and Form ADV were not reviewed.
8. Known / Unknown / Ask Next
| Area | What we know | What we don't know | Ask next |
|---|---|---|---|
| Team | Two co-CIO founders own 79%; 5 staff; hires in 2024–25; one leaver Nov 2025; Lannert gone since 2021 | Decision rights between the co-CIOs; succession; Gustafson's role and why Lannert left; Oceanfront's owner | Org chart, comp and equity plan, leaver history since 2020 |
| Terms / liquidity | 2/20, HWM, monthly with 30 days' notice; 2% fee within 12 months on SP | Gate and suspension terms; side letters; MFN; the domestic LP's own terms | Base Memorandum, LPA, side-letter summary |
| Performance | 42 live months, +47% net, two spike months drive the return | Actual SMA record for 2021–22; why history was restated as hypothetical; Dec 2025 and 2026 YTD | Admin-signed monthly NAV since launch; audited financials for 2022–25; actual SMA composite |
| Strategy / edge | Two-sleeve design; listed instruments; about 10 bps risk per trade | Standalone flow-sleeve P&L; tail budget; monetization record | Monthly attribution by sleeve; Aug 2024 and Apr 2025 trade logs |
| Risk / exposures | Capital usage ≤15–30%; tiered drawdown controls | Thresholds; vega and premium budget by tenor; stress realized vs. modeled | Current risk report; updated 2025 stress tables; backtest of the 2024–25 events |
| Capacity | "Capacity constrained" edges; master about $396M | Hard cap; flow-sleeve capacity curve; soft-close status | Capacity estimate and how it was derived; AUM and flow history |
| ODD / counterparties | Tier-1 and mid-tier providers; outsourced compliance and CFO | Directors; valuation policy; Florida exam result; SPC platform governance | Form ADV 1/2A, exam letter, valuation policy, SPC Director list |
| Alignment | GPs say they are invested; Sidial says 50% of his investable capital | Dollar GP commitment; trend over time | GP and employee capital by date, and its share of AUM |
9. Diligence Questions
Incentive Alignment
- What is the dollar amount of GP and employee capital in the fund today, and how has it changed since May 2022?
- Are there side letters with capacity rights, fee breaks or different liquidity? Is there an MFN?
- Why charge 20% with annual crystallization and no clawback on a strategy whose spike gains are meant to be monetized and recycled?
Edge & Repeatability 4. Please show monthly P&L for the flow sleeve and the tail sleeve separately since launch. Did the flow sleeve cover the tail premium plus fees in every year? 5. What was the flow sleeve's hit rate and P&L during the Aug 2024 and Apr 2025 spikes? Was it correlated with the tail book? 6. The strategy moved from short vega, VIX contango and 1–2 month tails in 2021 to no short optionality, flow trading and 1–18 month tails. What drove that change, and when did the current process go live? 7. Who built the flow signals, and what is Wise's independently verifiable record from before Ambrus?
Performance & Drawdowns 8. Why does the 2025 deck show hypothetical returns for January 2020 to May 2022 when SMAs were trading from January 2021? Why do the deck's 2021 figures differ from the actual figures in the June 2021 tear sheet? 9. How was the 2020 series simulated? Why don't the 2020 and 2021 YTD figures compound from the monthly numbers? 10. Walk us through the August 2024 and April 2025 monetizations: VIX path, amounts monetized, P&L given back, and the gap against the stress tables.
Team Stability 11. How are decisions split between the co-CIOs, and who has final say on monetization? What happens if one leaves? 12. What were Paul Gustafson's and Bret Lannert's roles, and why did they leave? Who is Oceanfront Consulting LLC? 13. How are Sutton and Nemani paid, and do they have a path to equity?
Risk, Financing & Capacity 14. What is the hard capacity for the flow sleeve and for the fund, and is the fund soft-closed? 15. How are flow signals allocated between the master fund, Medici, SMAs and Sundial? 16. What are the tiered drawdown thresholds, and the monthly tail-premium budget as a percentage of NAV? 17. What did the Florida OFR exam (May 2024) find? When did the firm move to SEC registration?
10. Reference Check Prep
- Current investors: How was the April 2025 monetization communicated? Have the monthly letters and attribution been consistent? Any surprises on liquidity or redemption fees?
- Former investors / SMA clients from 2021–22: What did your actual 2021 returns look like? Why did you leave? How was execution handled (the broker transition in 2021)?
- Former partners and employees (Lannert, Gustafson): Who really runs the book? How are disagreements between the co-CIOs resolved? What is the quality of risk discipline? Why did you leave?
- Service providers (NAV Consulting, Richey May/EY, Goldman Sachs): How long have you worked with them? Any valuation disputes, reconciliation breaks or late deliverables? How are SMA and master-fund books separated?
11. Suggested Materials Request List
- Administrator-confirmed monthly net returns since May 2022, plus the actual 2021–22 SMA composite and its methodology
- Audited financial statements for 2022–2025 (master, SP and domestic)
- Monthly attribution by sleeve (tail vs. flow), and premium spent vs. flow P&L
- AUM and investor-count history since 2021, including flows and concentration of the largest investors
- Current risk report: vega and premium by tenor and underlying, capital usage, drawdown thresholds
- Updated stress tables (2025 positioning), plus realized-vs-modeled analysis for Aug 2024 and Apr 2025
- Base Offering Memorandum for Liberty Investment Park SPC, the domestic LP's LPA and PPM, and a side-letter summary
- Form ADV Parts 1 and 2A, the Florida OFR exam correspondence, and the compliance manual table of contents
- Valuation policy, list of SPC Directors, and the trade-allocation policy covering Medici, SMAs and Sundial
- GP and employee capital statement; organization chart; leaver list since 2020
- Sample monthly investor letter; reference list of current and former investors
12. Sources Reviewed
| File | Type | Date |
|---|---|---|
| Ambrus DDQ - January 2026.xlsx | Manager DDQ | Jan 2026 (AUM as of 1 Jan 2026) |
| Ambrus Volatility Fund Investor Presentation (December 2025).pdf | Pitch deck, 39 pp. | Dec 2025 (performance through Nov 2025) |
| PPM Portfolio Supplement - Ambrus Volatility Fund SP - FINAL - February 2025.pdf | Offering document (SP supplement only) | Feb 2025 |
| Ambrus Volatility Fund SP - Subscription Agreement.pdf | Subscription docs (KKWC markup) | 28 Feb 2025 |
| The Ambrus Group Pitchdeck June 2021.pdf | Pitch deck | Jun 2021 (stale; pre-fund) |
| Ambrus Tear Sheet June 2021.pdf | Tear sheet | Jun 2021 (stale) |
| Ambrus Tear Sheet Memo June 2021.pdf | Investor memo | Jun 2021 (stale) |
| Ambrus Shock & Slide Levels September 8th 2021.pdf | Hypothetical stress scenarios | 8 Sep 2021 (stale) |
Not used as evidence. Two internal meeting memos in the fund folder sit outside sources/ and are treated as prior work. An iConnection profile there was not reviewed. The later memo contains unverified claims to test with the manager, including AUM of about $650M, a "soft close", backing from a large multi-strategy fund, and "top-performing volatility fund for 5 years". None of these claims appears in sources/. The DDQ reports $437.6M AUM as of 1 Jan 2026.
