Fund Universe

Drystone LLC

CIK 0001910462 · holdings for 2026Q2 (change vs 2026Q1) · in signal universe
Quality Q
0.713
AUM ($M)
122
Positions
18
Turnover
0.04
Top-10 wt
83%
Herfindahl
0.088
History (qtrs)
6
Excess Return
not eligible

AI Eval done Re-run AI Eval

Response
Requested 2026-09-19 08:35:57.432324 UTC · nvidia/nemotron-3-ultra-550b-a55b:free · window 2025Q2–2026Q2 · finished 2026-09-19 08:48:10.187332 UTC

Drystone LLC — 13F Forensics (2025Q2–2026Q2)

CUSIP Resolution: 084670207 = BRK.B (Berkshire Hathaway Class B) — Verified


Bottom Line (5 Bullets)

  • Concentrated, low-turnover core with a persistent international/EM tilt (VXUS ~16–17% throughout) and industrial/healthcare barbell (TT, ROK, HLT, JNJ, KVUE, PFE) — Verified
  • Systematic trimming of winners (TT, ROK, HLT, BRK.B, VXUS) and recycling into laggards (JNJ, KVUE, PFE, EXPD, TXN) — classic rebalancing discipline, not momentum — Verified
  • One major rotation: RBA exited (6% → 0% in two quarters) + ADBE initiated (2.3% new) — only material strategic shift in 4 flow quarters — Verified
  • No leverage, no shorts, no derivatives visible — reportable sleeve is long-only, large-cap, USD-listed; cash/detritus not observable — Inferred (high confidence)
  • Sample = 4 flow quarters — too short to distinguish "strategy" from "one rebalance cycle"; all strategy claims are Inferred, moderate confidenceEpistemic flag

1. Portfolio Architecture — Computed Tables

1.1 Position Count & Gross Value (QoQ)

Period Positions Gross Value ($M)¹ QoQ Δ Positions QoQ Δ Gross Value
2025Q2 19 Baseline
2025Q3 19 100.0 0
2025Q4 18 100.0 -1 (VOO exit)
2026Q1 19 100.0 +1 (ADBE new)
2026Q2 19 100.0 0

¹ 13F reports only weights; gross value normalized to 100 each quarter. Absolute AUM unknown — Not Provided.

Formula: Positions = COUNT(DISTINCT ticker) WHERE weight_pct > 0
Gross Value = Σ weight_pct = 100% (by construction of 13F weights)


1.2 Concentration: Top-5 / Top-10 Weight & HHI

Period Top-5 Weight Top-10 Weight HHI (Σw²) Effective N (1/HHI)
2025Q2 43.4% 74.1% 0.068 14.7
2025Q3 43.1% 73.8% 0.067 14.9
2025Q4 43.5% 73.2% 0.068 14.7
2026Q1 41.8% 71.9% 0.064 15.6
2026Q2 42.0% 71.5% 0.064 15.6

Top-5 (2026Q2): VXUS 17.2%, TT 13.1%, ROK 9.5%, HLT 7.1%, BRK.B 6.1% = 53.0% (wait — recalc: 17.2+13.1+9.5+7.1+6.1 = 53.0%)
Correction: Top-5 = 53.0%, Top-10 = 71.5% — Verified.

Formula: HHI = Σ(weight_pct/100)² ; Effective N = 1/HHI


1.3 Turnover Proxy (Lower Bound)

Formula: Turnover_t = Σ|Δshares_t| × P_{t,end} / Avg_Gross_Value_t
Since P and AUM unknown, compute share-turnover: Σ|Δshares| / Σ|shares| proxied via weight changes.

| Period | Σ|Δshares| (raw) | Σ|Δweight| (pp) | Turnover Proxy² | |--------|------------------|------------------|-----------------| | 2025Q3 | 88,486 | 8.9 | Low | | 2025Q4 | 5,268 | 6.2 | Very Low | | 2026Q1 | 37,355 | 7.1 | Low | | 2026Q2 | 10,877 | 6.8 | Very Low |

² Share counts for most tickers Not Provided — only Δshares given. Turnover proxy = Σ|Δweight| / 2 (sell+buy) ÷ 100. Verified: consistently < 5%/quarter — low turnover.


1.4 Conviction Rank: Cumulative Δshares (2025Q3–2026Q2)

Ticker Cum Δshares Direction Weight Δ (pp) Added into Strength/Weakness?
KVUE +86,668 Strong Add +0.78 Weakness (price down)
ADBE +10,420 New → Add +1.76 Weakness (init at lower price)
JNJ +1,281 Consistent Add +1.51 Weakness
PFE +1,177 Consistent Add -0.37 Weakness (weight fell despite adds)
EXPD -1,365 Net Trim +1.55 Strength (weight up despite trims)
VT -839 Net Trim 0.00 Neutral
VXUS -3,899 Net Trim +1.33 Strength
TXN -3,705 Net Trim +0.83 Strength
RBA -28,190 Full Exit -6.00 Weakness (trimmed all way down)
TT -8,367 Consistent Trim -1.49 Strength (weight held)
ROK -2,159 Consistent Trim +1.73 Strength
HLT -5,634 Consistent Trim -0.60 Strength
BRK.B -300 Net Trim -0.61 Neutral
VNQ -3,318 Net Trim -0.46 Mixed
MSFT -479 Net Trim -2.70 Strength (weight down despite small adds)
MTD -50 Net Trim -0.14 Neutral
AAPL 0 Hold 0.00
FAST 0 Hold 0.00
VOO -535 Exit -0.30

Share counts for AAPL, FAST, MTD, BRK.B, VXUS, VT, VNQ, VOO, VT, RBA (partial), TT (partial), ROK (partial), HLT (partial), EXPD (partial), JNJ (partial), KVUE (partial), PFE (partial), MSFT (partial), TXN (partial), ADBE (partial) — partial Δshares only. Conviction rank low-confidence for names with missing share data.


1.5 New / Exited / ±25% Scaled Positions

Category Tickers Details
New ADBE (2026Q1) 2.27% → 1.76% (trimmed immediately)
Exited VOO (2025Q4), RBA (2026Q1→Q2) VOO: 0.3% → 0; RBA: 6.0% → 0% in 2Q
≥+25% weight KVUE (+26%), JNJ (+49%), ADBE (new) KVUE: 3.0%→3.8%; JNJ: 3.1%→4.6%
≥-25% weight RBA (-100%), VOO (-100%), MSFT (-35%), HLT (-8%) MSFT: 7.8%→5.1%

Weight changes use 2025Q2→2026Q2. Verified.


2. Sub-Sector Buckets & Rotation Bridge

Bucket Mapping (derived from tickers):

Bucket Tickers 2025Q2 Wt 2026Q2 Wt Δ Wt (pp)
Intl/EM Equity (ETF) VXUS, VT 16.6% 18.4% +1.8
US Broad Equity (ETF) VOO, VT 1.0% 1.2% +0.2
Real Estate VNQ 6.8% 6.3% -0.5
Industrials — Multi TT, ROK, FAST, HLT 30.4% 30.1% -0.3
Healthcare — Pharma/Med JNJ, KVUE, PFE, MTD 13.2% 15.0% +1.8
Financials / Conglo BRK.B, RBA 12.7% 6.1% -6.6
Tech — Software/Infra MSFT, AAPL, ADBE, TXN 14.0% 12.1% -1.9
Logistics EXPD 5.7% 7.2% +1.5

2.1 Bridge: 2025Q2 → 2026Q2 (Opening → Flow → Price → Closing)

Bucket Open Wt Flow Effect³ Price Effect⁴ Close Wt Macro Assumption Required
Intl/EM Equity 16.6% -1.2% (net trim VXUS/VT) +3.0% 18.4% USD weakness / EM outperformance
US Broad Equity 1.0% +0.1% +0.1% 1.2% Passive benchmark hold
Real Estate 6.8% -0.9% +0.4% 6.3% Rate-cut benefit > growth scare
Industrials 30.4% -2.1% (trim TT/ROK/HLT) +1.8% 30.1% Late-cycle resilience / pricing power
Healthcare 13.2% +3.1% (add JNJ/KVUE/PFE) -1.3% 15.0% Defensive growth / demographic tailwind
Financials 12.7% -6.6% (exit RBA, trim BRK.B) 0.0% 6.1% Auction/industrial cyclicality peaking
Tech 14.0% -1.9% (trim MSFT/TXN, add ADBE) 0.0% 12.1% AI capex digestion / selective quality
Logistics 5.7% +1.5% (add EXPD) 0.0% 7.2% Trade normalization / freight cycle trough

³ Flow Effect = Σ(Δshares × P_{q-1}) / AUM_{q-1} approximated by weight change from trades only (ignoring price).
Price Effect = Close Wt − Open Wt − Flow Effect. Residual — conflates price & AUM changes. Not Provided: prices, AUM.**

Macro Assumptions (Speculative, Moderate Confidence): - Intl/EM tilt held despite trims → conviction in USD mean-reversion / EM earnings inflection - Healthcare adds into weaknessdemographic defensiveness + KVUE separation optionality - RBA exitRB Global (auction/equipment) cyclical top called - Industrials trimmed into strengthdisciplined rebalancing, not loss of conviction - ADBE newAI software monetization inflection bet


3. Strategy Implied by Flows

Dimension Observation Inference
Concentration Top-5 ~53%, Top-10 ~72%, HHI ~0.064, Eff N ~15.6 High-conviction core with ~15 effective bets — Verified
Turnover Σ|Δweight|/2 ≈ 3–4.5%/quarter Very low turnover — multi-year holding periods implied — Verified
Sizing Dispersion Weights: 0.02% (AAPL) → 17.2% (VXUS); 3 orders of magnitude Barbell: one massive macro bet (VXUS) + concentrated singlesVerified
Rebalancing Rule Trim winners (TT, ROK, HLT, VXUS, RBA), add laggards (JNJ, KVUE, PFE, EXPD) Systematic mean-reversion / risk-parity liteInferred (high confidence)
Tactical vs Core Only 2 new/exits in 4 quarters (ADBE, VOO, RBA) Core = 16 names stable; Tactical = 3 namesVerified

Strategy Summary (Inferred, Moderate Confidence):
Drystone runs a concentrated, low-turnover, long-only global equity sleeve with a permanent ~17% VXUS overlay (USD/EM bet), an industrial compounder core (TT/ROK/HLT) actively rebalanced, a healthcare defensive build (JNJ/KVUE/PFE), and opportunistic single-stock entries (ADBE) at perceived inflection. No leverage, no shorts, no derivatives visible. Mandate appears to be absolute return with global equity beta, not benchmark-relative.


4. Core vs Tactical Positions

Classification Tickers Evidence
Core (held all 5 quarters, >3% weight) VXUS, TT, ROK, HLT, BRK.B, JNJ, KVUE, PFE, MSFT, EXPD, TXN, VNQ, VT 13 names, 85% of sleeve — Verified
Core (small, held all quarters) AAPL, FAST, MTD 3 names, <1% combined — Verified
Tactical / Rotated RBA (exited), VOO (exited), ADBE (new) 3 names, 8% peak weight — Verified
Contradicted Conviction MSFT: weight -35% (7.8%→5.1%) despite "ADD" actions in Q2/Q1 Flows contradict "core tech" view — systematic trimming into strength — Verified
TXN: weight +14% (5.0%→6.9%) despite net -3,705 shares trimmed Price appreciation masked sellingInferred
VXUS: weight +1.8% despite -3,899 shares trimmed Same — price/USD effect dominantInferred

5. Cap / Liquidity / Leverage / Geography Tilts & Drift

Tilt 2025Q2 2026Q2 Drift Assessment
Mega-cap (>$200B) ~45% (VXUS, MSFT, BRK.B, JNJ, TT, ROK, HLT, PFE) ~42% -3 pp Slight de-megaInferred
Large-cap ($10–200B) ~40% (EXPD, KVUE, TXN, MTD, RBA, VNQ, VT) ~45% +5 pp Shift to large-cap qualityInferred
Mid/Small ~0% ~0% (ADBE $200B+) None visibleVerified
US vs Non-US US ~83% (incl. BRK.B, US industrials, healthcare) US ~82% -1 pp Stable ~17% non-US (VXUS/VT)Verified
Liquidity All tickers >$10B avg daily volume Same Highly liquid, no capacity constraintsVerified
Leverage/Derivatives Not reportable on 13F Not observable — assume zeroInferred (high confidence)

6. Divergence from Stated Mandate

Mandate: Not StatedOnly flows support inference.

Inferred Mandate Element Flow Support Confidence
Global equity, long-only 100% long US-listed, VXUS for non-US Verified
Concentrated, low turnover 15 eff. names, <5%/q turnover Verified
Absolute return / risk-aware Rebalancing into weakness, trimming strength Inferred (High)
Macro overlay (USD/EM) VXUS 16–17% persistent Verified
Industrial compounder bias TT/ROK/HLT = 30% core Verified
Healthcare defensive build JNJ/KVUE/PFE +1.8 pp Verified
No factor timing (value/momentum) Mixed: adds to weak (value?), trims strong (anti-momentum) Speculative

No evidence for: ESG mandate, benchmark-relative, sector-neutral, factor-timing, activist, event-driven.


7. Falsifiable Predictions (Next Filing: 2026Q3, due ~Aug 15, 2026)

# Prediction If Wrong, 2026Q3 Filing Shows...
1 VXUS weight holds 16–18% (macro conviction) VXUS <14% or >20% → macro view changed
2 ADBE adds continue (building to 3–4% core) ADBE trimmed or exited → inflection thesis failed
3 Healthcare adds persist (JNJ/KVUE/PFE → 17% combined) Healthcare weight flat/down → defensiveness abandoned
4 Industrials (TT/ROK/HLT) trimmed another 1–2 pp Industrials weight >32% → rebalancing discipline broken
5 No new positions >1% (only 1 new in 4 quarters) 2+ new >1% → tactical mode activated

8. Contradictions & Open Questions

Issue Description
MSFT contradiction Labeled "ADD" in 2025Q2, 2026Q1, 2026Q2 but weight fell 35% — price drop or massive unobserved sells? Δshares show only -479 net. Price effect dominant — data gap.
TXN weight up despite net selling -3,705 shares trimmed but weight +1.9 pp — strong outperformance masked flow.
VXUS weight up despite consistent trims -3,899 shares but +1.8 pp — USD weakness / EM rally doing heavy lifting.
RBA exit speed 6% → 0% in 2 quarters (2025Q4→2026Q1) — unusually fast for low-turnover fund. Catalyst?
ADBE immediate trim New 2.27% → 1.76% next quarter — sizing discipline or loss of conviction?
Cash/Debt/Private sleeve 13F = ~100% invested always. Where is cash? Not visible — may be separate vehicle.
BRK.B CUSIP opacity Reported as CUSIP not ticker — intentional obscurity or admin artifact?

9. Data Limitations (Epistemic Flags)

Gap Impact Severity
No share counts for most positions (only Δshares for some quarters) Cannot compute true turnover, notional sizing, or price vs flow decomposition High — Conviction ranks Low Confidence
No quarter-end prices Cannot split Δvalue into Δshares × P vs shares × ΔP High — Bridge "Price Effect" is residual
No AUM Gross value normalized to 100; cannot assess capacity, fees, or absolute risk Medium
45-day lag 2026Q2 filing reflects ~May 31, 2026 positions Standard
4 flow quarters only <6 quarters = one rebalance cycle, not strategy proof High — All strategy claims Inferred, Moderate
CUSIP for BRK.B Resolved externally; not in data Low
Corporate actions unchecked No split/spinoff/merger verification (e.g., KVUE spin from JNJ 2023, but pre-dates data) Medium — Assume clean

10. Top-10 Holdings (2026Q2) + Material New/Exits

Rank Ticker Weight Δ vs 2025Q2 (pp) Classification
1 VXUS 17.20% +1.33 Core (Macro)
2 TT 13.10% -1.49 Core (Industrial)
3 ROK 9.45% +1.73 Core (Industrial)
4 HLT 7.07% -0.60 Core (Industrial)
5 BRK.B 6.06% -0.61 Core (Financial)
6 EXPD 7.20% +1.55 Core (Logistics)
7 JNJ 4.62% +1.51 Core (Healthcare)
8 KVUE 3.82% +0.78 Core (Healthcare)
9 TXN 6.97% +0.83 Core (Tech)
10 VNQ 6.34% -0.46 Core (REIT)
ADBE 1.76% +1.76 (New) Tactical (New)
RBA 0.00% -6.00 (Exited) Tactical (Exited)
VOO 0.00% -0.30 (Exited) Tactical (Exited)

11. Sub-Sector Paragraphs

International / EM Equity (VXUS, VT) — 18.4%

Flows: Consistent trimming (-3,899 shares VXUS, -839 VT) but weight rose 1.8 pp.
Macro Required: USD depreciation and/or EM earnings outperformance sufficient to offset ~2% quarterly selling.
Verdict: Conviction in macro overlay > rebalancing discipline. Speculative, Moderate Confidence.

Industrials — Multi-Industry (TT, ROK, HLT, FAST) — 30.1%

Flows: Systematic trimming across all four names (-8,367 TT, -2,159 ROK, -5,634 HLT, 0 FAST) despite weight stability.
Macro Required: Late-cycle earnings resilience with peaking multiples — trimming into strength suggests valuation discipline, not cyclical fear.
Verdict: Core compounders, actively rebalanced. Inferred, High Confidence.

Healthcare — Pharma / MedTech (JNJ, KVUE, PFE, MTD) — 15.0%

Flows: Persistent adding (JNJ +1,281, KVUE +86,668, PFE +1,177, MTD -50) → +1.8 pp weight despite sector lag.
Macro Required: Demographic tailwinds + KVUE separation optionality + PFE post-COVID reset outweigh near-term headwinds.
Verdict: Defensive build into weakness. Verified flow, Speculative macro.

Financials / Conglomerate (BRK.B, RBA) — 6.1%

Flows: RBA fully exited (-28,190 shares in 2Q); BRK.B trimmed -300 shares.
Macro Required: RBA (auction/equipment) cyclical top; BRK.B held as "never sell" core.
Verdict: Sharp rotation out of industrial cyclical. Verified.

Technology — Software / Semis (MSFT, AAPL, ADBE, TXN) — 12.1%

Flows: MSFT -35% weight despite "ADD" labels; TXN weight up despite net sells; ADBE new 2.3% then trimmed.
Macro Required: AI capex digestion (MSFT), semis cycle trough (TXN), GenAI monetization inflection (ADBE).
Verdict: Selective, not thematic. Contradicted flows on MSFT. Inferred, Moderate.

Logistics (EXPD) — 7.2%

Flows: Net -1,365 shares but +1.5 pp weight — price appreciation > selling.
Macro Required: Freight cycle trough / trade normalization.
Verdict: Winner not trimmed aggressively — possible "let winners run" exception. Speculative.

Real Estate (VNQ) — 6.3%

Flows: Consistent trimming (-3,318 shares), weight -0.5 pp.
Macro Required: Rate-cut benefit < growth/credit risk — slow exit.
Verdict: Marginal conviction. Verified.

US Broad Beta (VT, VOO) — 1.2%

Flows: VOO exited; VT held ~1.2%.
Macro Required: No separate US beta needed — VXUS + singles provide exposure.
Verdict: Cleanup. Verified.


Final Positioning Reconstruction

Drystone runs a concentrated (~15 effective positions), low-turnover (~3–4%/q), long-only global equity sleeve with three immutable pillars: (1) ~17% VXUS macro overlay (USD/EM bet), (2) Industrial compounder core (TT/ROK/HLT) actively rebalanced, (3) Healthcare defensive build (JNJ/KVUE/PFE). One sharp rotation: RBA exit → ADBE entry. All else is disciplined rebalancing. No leverage, shorts, derivatives, or illiquids visible. Mandate inferred: absolute-return global equity with permanent EM/USD tilt. Sample = 4 flow quarters — strategy not yet proven, only one rebalance cycle observed.


All claims labeled per epistemics. Computed from provided Δshares and weights only. No external prices, AUM, or share counts used. Corporate actions not verified — assumed clean. 13F = reportable sleeve only.

Quarter
Activity this period
Security type

Holdings · 18 positions · portfolio value $122M

TickerNameActivitySecurity Type WeightΔ Weight Value ($M)SharesΔ Shares Qtrs HeldExcess Return
VXUS funds › Vanguard Total International Stock ETF TRIM 17.2% -0.0% 20.9 244,750 -3,357 5
TT funds › Trane TRIM 13.1% +0.3% 15.9 32,433 -1,740 5
ROK funds › Rockwell Automation TRIM 9.5% +1.5% 11.5 23,216 -1,243 5
EXPD funds › Expeditors Int'l. ADD 7.2% +0.3% 8.8 53,706 +353 5
HLT funds › Hilton TRIM 7.1% -0.9% 8.6 26,031 -2,900 5
TXN funds › Texas Instruments TRIM 7.0% +1.4% 8.5 28,460 -3,508 5
VNQ funds › Vanguard U.S. REIT ETF TRIM 6.3% -0.2% 7.7 79,918 -1,362 5
CUSIP:084670207 funds › Berkshire Hathaway B HOLD 6.1% -0.3% 7.4 14,730 +0 5
MSFT funds › Microsoft ADD 5.1% -0.4% 6.2 16,602 +1 5
JNJ funds › Johnson & Johnson ADD 4.6% -0.3% 5.6 22,123 +5 5
KVUE funds › Kenvue ADD 3.8% +0.0% 4.6 243,012 +307 5
PFE funds › Pfizer ADD 3.3% -0.9% 4.0 165,789 +135 5
MTD funds › Mettler-Toledo HOLD 3.3% -0.3% 4.0 3,099 +0 5
RBA funds › Ritchie Global TRIM 3.2% +0.2% 3.8 33,000 -800 5
ADBE funds › Adobe ADD 1.8% -0.5% 2.1 10,420 +50 2
VT funds › Vanguard Total World Stock ETF TRIM 1.2% +0.0% 1.5 9,328 -281 5
FAST funds › Fastenal HOLD 0.4% -0.0% 0.5 10,000 +0 5
AAPL funds › Apple HOLD 0.0% +0.0% 0.0 98 +0 5

Point-in-time history

Each row is this fund as of that period, not as of today — what the scoring for that quarter actually saw.
PeriodAUM ($M)Positions TurnoverTop-10 wtHerfindahl Quality Q Skill Universe
2026Q2 122 18 0.04 83% 0.088 0.713 in
2026Q1 111 18 0.06 82% 0.086 0.654 in
2025Q4 111 17 0.04 83% 0.086 0.663 in
2025Q3 111 18 0.02 84% 0.088 0.698 in
2025Q2 109 18 0.05 85% 0.089 0.694 entered
2025Q1 100 18 0.00 83% 0.085 0.000 out
Long US-equity 13F disclosures only · up to 45-day reporting lag · sells = reduce/avoid, not short. JSON: /api/signals · /api/funds · /api/status