The Coffee Grind · Brass Rat Capital FinTech
Persistence, Not Power
Two months into Operation Epic Fury, the pairs book stands at twenty-four positions across four tranches. The Tranche 1 trades that have been working since September continue to work; Tranche 4 was added at Monday's open with five additional structural infrastructure long legs paired against five of the highest-multiple AI, chip, and discount names. The trades still pending judgment will be answered tomorrow at the open.
The Strait of Hormuz has been effectively closed since February 28. That is two months tomorrow. In the first weeks of Operation Epic Fury ("OEF"), the temptation was to position for a fast resolution — a diplomatic settlement, a tanker convoy, a one-week ceasefire that would unwind the energy and water-infrastructure trades into a relief rally. That temptation was correctly resisted. The pairs portfolio entered the period long structural shifts and short legacy concentration. It has been held without subtraction. Where positions have been added — Tranche 2 on March 31, Tranche 3 on April 13, Tranche 4 at the April 27 open — the additions have extended the same thesis at higher entry points, not pivoted away from it. The thesis has not required power to vindicate itself. It has required persistence.
The same observation applies to several other long-running BRC commitments that move slowly through this column. The Standard Business Report Model ("SBRM") and Financial Data Transparency Act ("FDTA") Section 5821 advocacy is in its third year. The dispute with EDMA and OMG over Voluntary Consensus Standards Body ("VCSB") status under Office of Management and Budget ("OMB") Circular A-119 is in its seventh month, with five unanswered written requests on file. None of these have required force. They have required showing up every day.
Tomorrow, April 29, the persistence thesis faces its sharpest test of the year so far. Corning ("GLW"), Microsoft ("MSFT"), Generac Holdings ("GNRC"), and Alphabet ("GOOGL") all report fiscal first-quarter results in the same session. GLW and MSFT are the long and short legs of Pair 1, Pair 13, and Pair 21. GNRC is the long leg of Pair 2, Pair 7, Pair 14, and Pair 20. Alphabet (GOOGL Class A) is the long leg of Pair 24. Seven of the twenty-four pairs in the book have one or both legs reporting on the same day. That is not a hedge. That is a referendum.
Phase 0 · NYSE Calendar Gate
PASS. Tuesday, April 28, 2026 is a normal U.S. equity trading session. Regular hours. No scheduled holiday or early close. Easter (April 5) is past; the next NYSE closure is Memorial Day, May 25.
Market Dashboard
APR 28 SETTLED CLOSE. All values reflect April 28, 2026 settled NY close, confirmed via Yahoo Finance closing prints, CNBC Treasury quote pages, and yfinance batch pull. The intraday snapshot used in this morning's edition has been refreshed to settled values for archival accuracy.
| Instrument | Apr 27 Close | Apr 28 Close | Δ |
|---|---|---|---|
| S&P 500 (SPX) | 7,173.91 | 7,138.80 | −35.11 (−0.49%) |
| Dow Jones (INDU) | 49,167.79 | 49,141.93 | −25.86 (−0.05%) |
| Nasdaq Composite | 24,887.10 | 24,663.80 | −223.30 (−0.90%) |
| Russell 2000 (RTY) | 2,788.19 | 2,756.05 | −32.14 (−1.15%) |
| VIX | 18.02 | 17.83 | −0.19 (−1.05%) |
| MOVE | 68.42 | 68.68 | +0.26 (+0.38%) |
| OVX (Oil VIX) | 73.11 | 70.50 | −2.61 (−3.57%) |
| DXY | 98.480 | 98.620 | +0.140 (+0.14%) |
| 10Y UST yield (%) | 4.336 | 4.354 | +1.8 bp |
| 30Y UST yield (%) | 4.942 | 4.944 | +0.2 bp |
| 2Y UST yield (%) | 3.805 | 3.830 | +2.5 bp |
| WTI front-month ($/bbl) | 96.37 | 99.93 | +3.56 (+3.69%) |
| Brent front-month ($/bbl) | 108.23 | 111.26 | +3.03 (+2.80%) |
| Henry Hub NG ($/MMBtu) | 2.55 | 2.56 | +0.01 (+0.39%) |
| Gold spot (XAU, $/oz) | 4,675.40 | 4,591.60 | −83.80 (−1.79%) |
| Silver spot (XAG, $/oz) | 75.003 | 73.160 | −1.843 (−2.46%) |
Brent backwardation note. Front-month Brent settled at $111.26, up 2.80% on the day, with spot Dated assessments trading slightly above the front-month contract. Persistent backwardation against the rolling forward curve is the textbook signature of a physically tight market caused by the Strait of Hormuz closure. WTI’s $99.93 settle, up 3.69%, retests the OEF-period highs set in March. The Bull Shit Detection ("BSD") algorithm flags any narrative of supply normalization at these prices.
The volatility complex settled with internal divergence intact. VIX closed lower at 17.83 (-1.05%) despite the headline-driven open. MOVE Index essentially flat at 68.68 confirms Treasury markets remained uninterested in the equity-side news flow. OVX (oil volatility) eased 3.57% to 70.50 even as oil prices firmed, a market consensus that the price level is sustainable rather than acutely spiking. The intraday narrative was driven by the Wall Street Journal report on OpenAI revenue and user-growth shortfalls; the close repaired roughly half of that early Nasdaq damage. Tranche 4 short legs (AVGO, INTC, MU, CRWV, JBLU) ended down on the day, supporting the mean-reversion thesis on day two. The April 29 earnings storm will deliver the next material data point.
The Dow finished essentially flat (-0.05%) versus the S&P’s -0.49% and Nasdaq’s -0.90% loss. Coca-Cola’s 5%+ gain after a Q1 beat anchored the Dow against the broader tech selling. Russell 2000 was the day’s underperformer at -1.15%, signaling that the de-risking extended past mega-cap technology into smaller caps and lower-quality balance sheets — the specific names where private credit gating and refinancing risk are concentrated. Pair-portfolio exposure to that internal rotation is direct and favorable: long legs in Pair 1, Pair 13, and Pair 21 (Corning) and Pair 22 (Chevron) are not in the names being sold, while Tranche 3 and Tranche 4 short legs are precisely the names being sold.
Pairs Portfolio · 24 Pairs Active
The book holds 24 pairs across four tranches. Tranche 1 (Pairs 1–2, baseline September 29, 2025, $100,000 per leg) is the original Long Corning / Short Microsoft and Long Generac / Short NVIDIA structure. Tranche 2 (Pairs 3–14, initiated March 31, 2026, $100,000 per leg) added water infrastructure, private credit alternatives, and three additional Corning and Generac legs at higher entry points. Tranche 3 (Pairs 15–19, initiated April 13, 2026) added Freeport-McMoRan / Aptiv, Alcoa / Boeing, Sibanye Stillwater / Honda Motor (the platinum-group metals "PGM squeeze" thesis), GitLab / Atlassian, and AMD / iShares MSCI South Korea (the last leg restated from HYMTF after that ticker's delisting).
Tranche 4 (Pairs 20–24, initiated April 27, 2026 at the open) extends the same structural framework deliberately into AI, semiconductor, AI-infrastructure, and discount-airline concentration. The five new long legs are Generac, Corning, Chevron, Delta Air Lines, and Alphabet (GOOGL Class A voting common). The five new short legs are Micron Technology ("MU"), Intel ("INTC"), Broadcom ("AVGO"), CoreWeave ("CRWV"), and JetBlue ("JBLU"). The Tranche 4 view: over the next three to six months, the chip rally pauses while the AI infrastructure build-out re-emerges as the price-setting narrative. Chevron continues to benefit from the persistent OEF risk premium in oil. Delta benefits from real-economy travel demand against frothy AI infrastructure pure-play exposure (CoreWeave). Alphabet's cash-flow generation through cloud, advertising, and AI investment is paired against the structural challenges facing the discount-carrier business model.
The structural read across the full book: Tranche 1 is the persistence pair par excellence — both Corning and Generac have rallied substantially against their early-September entries while their short counterparties (Microsoft, NVIDIA) have not kept pace. Tranche 2 is mid-development; the water infrastructure pairs (PHO, XYL, ERII) have moved positively but their short legs have moved with them, leaving net P&L modest. Tranche 3 is two weeks old. SBSW (Sibanye) is up roughly 33% from a $9.50 entry to $12.50 on April 27 against a flat short in HMC; the PGM squeeze thesis is the early winner of that tranche. AMD against EWY (the South Korea ETF) is the Tranche 3 surprise, with AMD running from a $241.40 entry to $334.63 in two weeks. Tranche 4 is two trading days old.
APRIL 28 SETTLED P&L: +$134,788 across the 24 pairs, down from +$186,828 at April 27 close — a $52,040 single-day drawdown. The mechanism was almost entirely Corning ("GLW") trading from $168.01 to $153.05 intraday (-8.9%) in a pre-earnings de-risking move that hit Pairs 1, 6, 13, and 21 simultaneously. The single-stock concentration risk documented in the puke-scenario analysis from earlier this session is now visible in the live tape. When GLW reports tomorrow morning, the four affected pairs will move in lockstep regardless of how Microsoft (the short leg of two of them) prints. The book’s aggregate exposure remains net long structural infrastructure; the day’s loss does not change that posture.
| # | Long Leg | Entry | Apr 28 | Short Leg | Entry | Apr 28 | Note |
|---|---|---|---|---|---|---|---|
| 1 | GLW | $80.26 | $153.05 | MSFT | $514.60 | $429.25 | ★ Runaway · both report Apr 29 |
| 2 | GNRC | $165.82 | $217.12 | NVDA | $181.85 | $213.17 | GNRC reports Apr 29; NVDA May 20 |
| 3 | PHO | $66.86 | $68.47 | BEDZ | $31.03 | $33.53 | Water infra vs leisure |
| 4 | XYL | $119.56 | $117.91 | RONB | $22.87 | $23.31 | Water tech vs short basket |
| 5 | ERII | $10.07 | $11.09 | MCR | $6.06 | $5.99 | Desalination PX vs BDC |
| 6 | GLW | $135.97 | $153.05 | META | $549.86 | $671.34 | Short META lagging thesis |
| 7 | GNRC | $195.33 | $217.12 | MU | $460.83 | $504.29 | Entry back-solved from share count |
| 8 | APO | $108.42 | $123.57 | GSIB | $48.93 | $55.49 | Themes GSIB ETF |
| 9 | BX | $108.07 | $121.52 | KBWB | $85.76 | $86.87 | Alts vs banks ETF |
| 10 | BLK | $934.06 | $1,049.76 | XLF | $49.05 | $51.85 | Asset mgmt vs financials |
| 11 | BRK-B | $474.66 | $478.16 | MURGY | $12.24 | $12.73 | Annual mtg May 3 Omaha |
| 12 | MET | $67.73 | $78.28 | CVS | $70.08 | $80.98 | Liability-driven inv. thesis |
| 13 | GLW | $128.55 | $153.05 | MSFT | $358.96 | $429.25 | Both report Apr 29 |
| 14 | GNRC | $185.45 | $217.12 | NVDA | $165.17 | $213.17 | GNRC reports Apr 29 |
| 15 | FCX | $66.65 | $58.21 | APTV | $58.61 | $59.12 | Tranche 3, copper vs autos |
| 16 | AA | $70.20 | $63.17 | BA | $218.00 | $230.72 | Tranche 3, alum vs aerospace |
| 17 | SBSW | $9.50 | $11.93 | HMC | $24.04 | $24.20 | Tranche 3 leader · PGM squeeze |
| 18 | GTLB | $18.73 | $22.44 | TEAM | $57.47 | $69.71 | Tranche 3, devops |
| 19 | AMD | $241.40 | $323.21 | EWY | $141.23 | $154.37 | Tranche 3 (restated from HYMTF) |
| 20 | GNRC | $225.22 | $217.12 | MU | $510.56 | $504.29 | Tranche 4 · chip rally vs grid power |
| 21 | GLW | $176.30 | $153.05 | INTC | $83.67 | $84.52 | Tranche 4 · optical vs legacy CPU |
| 22 | CVX | $186.19 | $188.36 | AVGO | $420.48 | $399.83 | Tranche 4 · OEF energy vs custom AI chips |
| 23 | DAL | $68.75 | $67.22 | CRWV | $109.87 | $105.53 | Tranche 4 · real-economy travel vs AI infra pure-play |
| 24 | GOOGL | $345.98 | $349.78 | JBLU | $5.27 | $5.00 | Tranche 4 · megacap cash flows vs discount carrier; GOOGL reports Apr 29 |
Note on Pair 7 short leg. The source pricing workbook recorded the Micron short entry as "$88.89" against a documented 217-share execution. Those numbers do not reconcile to the $100,000 per-leg notional convention (217 × $88.89 = $19,289). The entry price has been back-solved from the share count to $460.83, which is consistent with where MU was trading at the March 31 inception date. The original blotter remains the source of truth; the workbook description requires correction. The pair is approximately flat after this reconciliation, not the catastrophic loser the un-reconciled numbers would have implied.
Operation Epic Fury · Scenario Assessment
The standing internal probability assignment holds at 60% Scenario C (extended closure, gradual partial reopening with permanent capacity loss). Scenario A (rapid diplomatic resolution and full reopening) remains discounted; Scenario B (phased partial reopening within 90 days) absorbs most of the residual probability mass.
Two months in, three observations support the Scenario C base case:
- Volatility compression has not been a peace signal. The April 27 and 28 readings show VIX rising from 18.02 to 18.92, MOVE roughly flat at 67-68, and OVX climbing from 73.11 to 74.43. Treasury volatility is steady-state; equity and oil volatility are firming. WTI is up 2.9% intraday at $99.18 and gold has given back 1.8% to $4,591.60. A genuine Scenario A unwind would compress oil and gold simultaneously. That has not occurred — gold's pullback is consistent with a positioning unwind, not with peace, while oil is moving the wrong direction for an OEF resolution narrative.
- The desalination capital cycle has begun. Energy Recovery ("ERII") and Flowserve ("FLS") order books are turning. Air Products ("APD") received a J.P. Morgan upgrade on March 20 explicitly tied to the helium and industrial-gas supply implications of the Qatar Liquefied Natural Gas ("LNG") shut-in. Twelve-to-twenty-four-month equipment lead times are not consistent with a market priced for a near-term reopening.
- Private credit gating is accelerating. Carlyle Tactical Private Credit Fund posted 15.7% redemption requests against a 5% gating cap in the most recent monthly window — the third such fund this calendar month. Apollo Debt Solutions BDC, Ares Strategic Income, BlackRock HPS Lending, Blackstone BCRED, and Blue Owl OCIC are all on the BSD Castle Bravo cascade watchlist. Total gated capital across monitored vehicles is now in excess of $100 billion.
The persistence frame is an analytical posture as much as a meditation. Scenario C does not require a single dramatic event to be vindicated. It requires the absence of a dramatic resolution event — day after day, week after week. The river does not announce its work.
Tomorrow · April 29 Earnings Storm
Wednesday, April 29 carries the heaviest concentration of pairs-relevant earnings of any single session this year. Four reporters land before or after the bell:
- Corning ("GLW") — long leg of Pairs 1, 6, 13, and 21. Display, optical communications, and life sciences segments. The pair-portfolio thesis rides on continued data-center optical demand against any guidance softening. April 27 close $168.01 (down 4.7% intraday from a $176.30 open); intraday all-time high $176.75 was set April 10. Investor relations: investor.corning.com.
- Microsoft ("MSFT") — short leg of Pairs 1 and 13. Azure capacity, Copilot monetization, and capital expenditure guidance. April 27 close $424.82. Investor relations: microsoft.com/en-us/investor.
- Generac Holdings ("GNRC") — long leg of Pairs 2, 7, 14, and 20. Residential and commercial-industrial generator demand, grid-edge software, and any commentary on data-center backup capacity. April 27 close $220.30. Investor relations: investors.generac.com.
- Alphabet ("GOOGL") — long leg of Pair 24 (Class A voting common, $345.98 entry, 289 shares). Magnificent Seven anchor, advertising and Google Cloud read, AI infrastructure capex commentary. April 27 close $350.34. Investor relations: abc.xyz/investor.
The pairs-portfolio implication: four of the seven pairs in the book have at least one leg reporting April 29. A constructive print from GLW combined with a guidance-driven softening in MSFT would mark Pairs 1 and 13 to materially higher P&L. A constructive GNRC print would do the same for Pairs 2, 7, and 14. The asymmetric outcome the portfolio is built for — structural infrastructure beneficiaries beating concentrated software incumbents — gets a single-day verdict.
Apple ("AAPL") and Amazon ("AMZN") report Thursday April 30, with Apple's iPhone services mix and Amazon Web Services capital-expenditure cadence as the day's pivot points.
Castle Bravo Watch · Private Credit Cascade
The Castle Bravo excluded-variable framework continues to flag private credit as the highest-priority systemic monitor. The pattern is consistent: a monthly redemption window opens, demand exceeds the contractual gating cap (typically 5% of net asset value per quarter), the gate triggers, and the fund proceeds to mark assets at sponsor-supplied valuations rather than secondary-market clearing prices.
The most recent additions to the gated list:
- Carlyle Tactical Private Credit Fund — 15.7% redemption requests, 5% cap, third gated fund this month (flagged April 8).
- Blue Owl OCIC — 29.1% redemption requests against the 5% cap.
- Apollo Debt Solutions BDC — 5% cap reached in the Q1 2026 window.
- Blackstone BCRED — $400 million emergency capital injection from the sponsor.
- BlackRock HPS Lending — $26 billion in gated or restricted capital.
The "Paper versus Physical Divergence" ("PvP") framing applies directly. The marks are paper. The redemption demand is physical. The longer the gates hold, the wider the gap between the two becomes — until a forced-seller event at any one of these vehicles establishes a clearing print that the others must reckon with.
BRC long positions in this cluster (Apollo Pair 8 long, Blackstone Pair 9 long, BlackRock Pair 10 long) are matched against short positions in regional and large-bank ETF baskets. The structural view is not that the alternative asset managers are themselves at risk — their fee streams are durable and locked-up — but that the regulated-bank counterparts face the same credit-quality reality at higher disclosure burden. The pair structure is designed to capture the disclosure asymmetry, not to take outright direction.
Persistence as Strategy
This column has, over many editions, observed three things about long-running positions of any kind — whether portfolio positions, regulatory advocacy positions, or institutional governance positions:
The first observation is that conviction without scaling is meaningless. A trade that one is right about but cannot sit in is the same as a trade one was wrong about. The Tranche 1 GLW and GNRC long legs were correct in late September; they were also correct in October, November, December, January, February, and March. They were correct on every single day of that period, including the days when the screen was red. Persistence is the mechanism by which conviction becomes return.
The second observation is that institutions test persistence by silence. Five written requests to OMG Technical Steering Committee principals on the EDMA acquisition and VCSB-status question remain unanswered. The silence is not random. It is a tactic that succeeds only against parties who lose interest. The BRC standing commitment is to file one paragraph per Coffee Grind edition referencing the dispute through the OMG Q2 Technical Committee meeting in Chicago, June 1–5. The substantive return to the governance section is calendared for that week. The pause is the persistence.
The third observation is that persistence is asymmetric in its effect. A river that floods once and recedes leaves a watermark. A river that returns every spring for ten thousand years leaves a canyon. The pairs portfolio is currently holding 24 positions across four tranches. None of them require a flood. All of them require return.
Governance
The EDMA / OMG / VCSB / FDTA governance section is intentionally paused through the OMG Q2 Technical Committee meeting in Chicago, June 1–5, 2026. Five written requests to OMG Technology Steering Committee principals remain unanswered as of this edition. Substantive return to this section is calendared for the first edition published after the Chicago meeting concludes.